Inhibition of protein synthesis by an efficiently ... - BioMedSearch

686-693 Nucleic Acids Research, 1994, Vol. 22, No. 4

,../ 1994 Oxford University Press

Inhibition of protein synthesis by an efficiently expressed mutation in the yeast 5.8S ribosomal RNA Sherif Abou Elela, Liam Good, Yuri F.Melekhovets and Ross N.Nazar* Department of Molecular Biology and Genetics, University of Guelph, Guelph, Ontario NiG 2W1, Canada Received July 7, 1993; Revised and Accepted January 3, 1994

ABSTRACT Recent studies on the inhibition of protein synthesis by specific anti 5.8S rRNA oligonucleotides strongly suggested that this RNA plays an important role in eukaryotic ribosome function. To evaluate this possibility further, a ribosomal DNA transcription unit from Schizosaccharomyces pombe was cloned into yeast shuttle vectors with copy numbers ranging from 2 to approximately 90 per cell; to allow direct detection of expressed RNA and to disrupt the function of the 5.8S rRNA molecule, a five base insertion was made in a universally conserved GAAC sequence. The altered mobility of the mutant RNA was readily detected by gel electrophoresis and analyses indicated that mutant RNA transcription reflected the ratio of plasmid to endogenous rDNA. The highest copy number plasmid resulted in about 40- 50% mutant RNA. This mutant RNA was readily integrated into the ribosome structure resulting in an in vivo ribosome population which was also about 40- 50% mutant; the rates of growth and protein synthesis were equally reduced by approximately 40%. A comparable level of inhibition in protein synthesis was demonstrated in vitro and polyribosomal profiles revealed a consistent increase in size. Subsequent RNA analyses indicated a normal distribution of mutant RNA in both monoribosomes and polyribosomes, but elevated tRNA levels in mutant polyribosomes. Additional mutations in alternate GAAC sequences revealed similar but cumulative effects on both protein synthesis and polyribosome profiles. Taken together, these results suggest little or no effect on initiation but provide in vivo evidence of a functional role for the 5.8S rRNA in protein elongation.

INTRODUCTION Despite decades of study and

very

substantial

progress, many

aspects of ribosome biogenesis, structure and function remain

unclear. Recently, the discovery by Cech and others (see ref. 1) that RNA can act as a biological catalyst has significantly altered the 'protein bias' with respect to ribosome function and *To whom correspondence should be addressed

led to new speculation regarding the function of the ribosomal RNA in protein synthesis (e.g. refs. 2-4), a possibility which now has been extended with direct experimental evidence (5, 6). Furthermore, technical advances with cell transformation and sitedirected mutagenesis have provided new approaches to questions regarding both rRNA structure and function. For example, Kunkel (7) has developed a highly efficient mutagenesis system based on the M13 phage cloning vectors and intermediate and high copy plasmid vectors in yeast can produce a range of copy numbers which exceeds 100 (refs. 8-10). In studies on eukaryotic ribosomes, the high genomic rDNA copy number has generally been considered a serious limitation because endogenous copies cannot be readily replaced and ribosomes are essential for cell viability (11). To overcome this problem, Planta and co-workers have successfully tagged the mature 17 and 26S rDNAs of yeast with insertions which can be detected even though only a few percent of the molecules are mutant (11, 12). The availability of high copy vectors for yeast has removed even this need for sensitive probes. For example, studies in our laboratory on the expression of mutant yeast 5S rRNA genes in vivo have shown that specific base substitutions can easily be detected by gel electrophoresis and despite the high genomic copy number for the 5S rRNA gene, under optimized conditions more than 80% of the cellular 5S RNA can be mutant (13, 14).

In the cytoplasmic ribosomes of eukaryotes, the 5.8S rRNA is present as a separate molecule (15) forming an RNA:RNA complex with the 25 -28S rRNA (16, 17). A recent study on the function of the 5.8S rRNA demonstrated that oligonucleotides which were complementary to specific exposed regions in the ribosome associated molecules could significantly inhibit protein synthesis in a sequence specific manner (18), strongly suggesting an important role in ribosome function. To examine this possibility further, we have begun in vivo studies to test the effects of specific mutations. The present results show that under optimized conditions, mutant 5.8S rRNA can be efficiently expressed and readily detected with a substantial portion of the cellular 5.8S rRNA population being mutant. Furthermore, mutations in the putative tRNA binding sites are observed to significantly inhibit cell growth as well as protein synthesis, in vitro. sequence

Nucleic Acids Research, 1994, Vol. 22, No. 4 687

MATERIALS AND METHODS Construction and expression of mutant rRNA

genes

S.pombe genomic DNA was prepared as described by Cryer et al. (19) except the cell wall digestion condition was modified (20) to include additional glucanase activities (21). A genomic library of Hindm endonuclease digestion fragments was prepared in the lambda Charon 39A vector (22) and recombinants containing complete rDNA transcriptional units were selected by plaque hybridization (23). Site-specific mutations were introduced into the Sc.pombe rDNA using T7 DNA polymerase as described by Kunkel (7) or by 2-step PCR-based targeted mutagenesis (24). Initially, alO.4 Kbp HindH digestion fragment containing one complete rDNA transcriptional unit (25) was subcloned into the pRS316 shuttle vector (10) from which uridinecontaining template DNA was prepared in E. coli strain CJ236. After primer extension and ligation, E. coli strain DHa5' (26) was transformed with the resulting double-stranded DNA; replicated DNA was prepared from transformants and mutations were confirmed by DNA sequencing using the dideoxy method of Sanger and co-workers (27). As indicated in Fig. 1, the mutant rDNA was excised by HindIII or PvulI endonuclease digestion and subcloned into the yeast shuttle vectors using standard recombinant DNA techniques (28) and E. coli strain C490 as the bacterial host. Replicated recombinant plasmid was purified as previously described (29) and used to transform Sc.pombe strain h- leul-32 ura4-D18 by the method of Okazaki and co-workers (30). PCR-based mutagenesis experiments utilized the shuttle vector recombinants for templates. Growth rates for the transformants were determined using the absorbancy of cultures at 550nm and the plasmid copy number was determined by DNA blot hybridation as previously described (13). Preparation and analysis of cellular RNAs For all analyses, 5.8S rRNA was isolated from ribosomes, polyribosomes or whole cells by sodium dodecyl sulphate/phenol extraction as previously described (31). For in vivo labeled RNA, cells were grown in 200 ml of minimal medium (0.67 % nitrogen base containing 2 % dextrose, 80 tig/ml of leucine and 200 dig/ml of asparagine) which was pretreated to remove inorganic phosphate (32). Typically, cells were incubated with 300 tiCi of inorganic 32P-orthophosphate for 12-18 hours prior to extraction. Unlabeled, in vivo labeled or RNA labeled at the 3' end with cytidine [3', 5'-32p] bisphosphate (33) was fractionated on 8 or 12% polyacrylamide gels containing 8M urea, using autoradiography or methylene blue staining (34) to detect the separated RNA components. For quantitative analyses the fractionated bands were scanned using a model 620 CCD densitometer (Bio-Rad Laboratories, Richmond, CA) or counted using a scintillation counter. For RNA sequence analysis the labeled RNA was chemically degraded as described by Peattie (33) and fractionated on 12% sequencing gels. In vitro translation assay Yeast transformants were grown with constant shaking at 30°C in minimal medium broth (35) containing 2% dextrose, 80 14g/ml of leucine and 200 ,ug/ml of asparagine, to an absorbency of 0.6-0.8 at 550 nm. The cells were harvested by centrifugation (3,000 xg) for 5 min. at 4°C and washed with water followed with extraction buffer (0. IM NH4C1, 2 mM Mg acetate, 2 mM dithiothreitol, 20% glycerol in 20 mM HEPES/KOH, pH 7.4). The washed peflet aga'm was resuspended in cold extraction buffer

(1 ml per gram wet weight) and the cells were broken by vortexing with an equal volume of glass beads (0.5 mm) using ten 30 second pulses at maximum speed. The lysate was chilled for 30 seconds on ice between pulses; by microscope, typically the breakage was estimated to be greater than 80%. The homogenate was removed by pipette, the glass beads were further rinsed with an equal volume of buffer and the pooled extract was cleared of cell debris by centrifugation at 15,000 x g for 10 min. (4°C). The supernatant was removed avoiding the top lipid layer and further cleared by centrifugation at 15,000xg for 15 min (4°C). Again, the top lipid layer was avoided and the supernatant was diluted with homogenization buffer to 50 A260 units per ml. The assay protocol was that of Picard and Wegnez (36) with modification. Cell free extract (40 td) was added to 60 Al of reaction mix (125 mM NH4Cl, 3 mM Mg acetate, 0.5 mM ATP, 0.1 mM GTP, 25 mM creatine phosphate, 20 jd/ml creatine phosphokinase, 10% glycerol, 20 mM HEPES/KOH, pH 7.4) containing 40 yM of each amino acid except leucine and 2 /Ci 3H-leucine (1.46 AM) and incubated at room temperature. Aliquots (20 1l) were removed at appropriate times, mixed with an equal volume of cold 10% TCA and heated for 3 min. at 900C. One-half ml of a 3 % Casamino acids solution was added to each sample which were then placed on ice for 10 min. before being applied to Whatman GF/C glass filters (Maidstone, England); the filters were washed twice with 10% TCA, followedonce with ethanol. After the filters were dried at 800C, the radioactivity was determined by liquid scintillation counting. To normalize the profiles for small differences in ribosome concentration, RNA was extracted from aliquots of each extract with SDS/phenol, fractionated on 8% polyacrylamide gel and a relative ribosome concentration was deduced assuming 1 molecule of 5 and 5.8S rRNA per ribosome.

Preparation and analysis of cellular protein Cellular proteins were extracted essentially as described by Gorenstein and Warner (37). Cells were harvested by centrifugation and washed twice with cold 1.2 M sorbitol before being incubated for 90 minutes at 300C with 5 mg/ml of Novozym234 (Sigma Chemical Co., St. Louis, MO). The protoplasts were lysed in distilled water followed immediately by the addition of 0.4 volumes of 1M MgCl2 and 2 volumes of glacial acetic acid. After mixing on ice for 30 minutes, the suspension was cleared by centrifugation at 20,000xg for 15 minutes and the supernatant was dialyzed against 1 % acetic acid and lyophilized. An equal weight was dissolved in loading buffer and aliquots were fractionated on 10% SDS/polyacrylamide gels as described by Laemmli et al. (38).

Preparation and analysis of polyribosomes and ribosomal subunits Yeast transformants were grown with constant shaking at 300C in minimal medium broth to an absorbency of 0.4-0.6 at 550 nm. The cells were collected by centrifugation, resuspended in an equal volume of enriched medium (0.5% yeast extract) containing 3 % dextrose and grown further for 60 min. To prepare polyribosomes, the cells were disrupted by vortexing with glass beads or with Trichoderma lysing enzymes (Sigma Chemical Co., St. Louis, MO). When broken with glass beads, the cells again were harvested by centrifugation, resuspended (1 ml per gm wet weight) in lysis buffer (0.1 M NaCl, 30 mM MgC12, 10 mM Tris/HCl, pH 7.4) and broken by vortexing with an equal volume of glass beads, using eight 20 sec. pulses with 20 sec. pauses

688 Nucleic Acids Research, 1994, Vol. 22, No. 4 on ice (Van Ryk et al., 1992). Triton X-100 was added to a 1% final concentration and the lysate was cleared of cellular debris by centrifugation at 15,000xg for 5 min. (4°C). The supematant was diluted to 75-80 A260,k, units per ml and 0.5 ml aliquots were fractionated at 4°C on 5-40% sucrose gradients in lysis buffer for 2 hours at 27,000 RPM in a Beckman (Palo Alto, Ca) SW41Ti rotor. When enzymatically disrupted, the cells were incubated for 30-60 minutes in enriched medium containing 0.9M sorbitol and 1 mg per ml of Trichoderma lysing enzymes. The sphaeroplasts were then washed once with medium containing sorbitol and resuspended in lysing buffer (1OOmM KCI, 30mM MgCl2, 1mM DTT, 20mM Tris/HCl, pH 7.4) containing 0.2% of Triton X-100 and deoxycholate. Lysates were again fractionated on 5-40% sucrose gradients in lysis buffer. To dissociate the ribosomes for total subunit analyses, the cells were broken in buffer containing 0.8M KCI, 20 mM mercaptoethanol, 12 mM magnesium acetate and 50 mM Tris/HCl, pH 7.7 (39) and fractionated on 5-35% sucrose gradients at 35,000 RPM for 5 hours. All gradients were analyzed at 254 nm using an ISCO (Lincoln, NE) model 640 density gradient fractionator.

RESULTS The mutagenesis and expression strategy which was used in the initial studies is summarized in Fig. 1. A complete rDNA transcription unit was isolated from a library of HindIH restriction endonuclease cleavage fragments of Schizosaccharoyces pombe genomic DNA in the Charon 39A vector (22). The library was screened with a labeled 5.8S rRNA probe and the 10.4 Kbp fragment, characterized earlier by Schaak and co-workers (25), was subcloned into the HindU site in phagemid vector pRS316

Mutant

aOgnudeod. Prknr

u

pRS316 U+ T-DNA

T4

DMA U.1g

T4

Kinetic analysis of protein synthesis Based on the analytical approach of Dintzis (40), yeast transformants were grown with constant shaking at 30°C in minimal medium broth to an absorbancy of 0.4 at 550nm, labeled for 30 min. with 5 /Ci of L-[3,4,5-3H(N)]-leucine and transferred to medium supplemented with 80 mg/litre of L-leucine. Aliquots were rapidly frozen on an equal volume of pre-cooled (-85°C) glass beads, the cells were disrupted by vortexing, diluted with one volume of lysis buffer and Triton X-100 was added to 1 % final concentration. The lysate was cleared at 15,000Xg for 10 min. (4°C); the supematant was layered on an equal volume of 30% sucrose and the polyribosomes were collected by centrifugation at 50,000 RPM in a Beckman (Palo Alto, CA) 70. 1Ti rotor for 2 hours. The pellet was resuspended in 0.5 ml of water and a 100 1d aliquot was extracted with SDS/phenol for RNA quantification. The remainder was diluted with 2 volumes of 10% TCA, heated for 3 min. at 90°C and the protein was precipitated on ice. After filtration and washing with 10% TCA the remaining radioactivity was determined by liquid scintillation counting.

(10). Mutagenesis was based on the in vitro method of Kunkel (7). After mutagenesis, the mutation was confirmed by DNA sequencing, excised with appropriate endonucleases and cloned in a variety of autonomously replicating yeast shutfle vectors with differing in vivo gene copy numbers as indicated in Fig. 1. The

u

DMA Polymerae

pRS316 U+ RF-DNA

GAUCI_

E. coli DH5aF'

II H:EMua or

utcD

U 100

Mutant rDNA

0U U PAAACUCAGCAAAUUCUUCUCUCCAU ACCUAU 10 C QC 30C.^ A C 130 u

CUG

G.c

TCT A-U pFL20

pSPI

pW102

A

C

U-A U-C 2o0U-A

pDB248

4J90 A A

50CGA a Sc. pombe h- leul32 u4-D18

Figure 1. Expression of mutant 5.8S rRNA in Sc.pombe. Mutations were introduced into a pRS316/Sc.p. rDNA recombinant using the site-specific methods of Kunkel (7). Mutations were confirmed by DNA sequencing and the Hindff (H) or Pvull (P) restriction fragment which contained the mutated gene was subcloned into different copy number autonomously replicating yeast shutde vectors prior to yeast cell transformation. Transformants were selected by the Leu+ or Ura+ phenotype. Shuttle vectors pFL20, pSPI, pW102 and pDB248 are those described by Heyer et al. (8), Cottarel (50), Heyer et al. (8) and Beach and Nurse (52), respectively. U+T-DNA, uracil containing template DNA; U+RF-DNA, uracil-containing replicative form (double stranded); H and P are HindIm and Pvull restriction endonuclease cleavage sites, respectively.

C.G

U-A

A

a oQcQ A .A

LG

U.A

a..C

QG.C 130

a CA

Figure 2. Estirmate of the structure for the ribosomial SpS.8A44iS mutant 5.8S RNA domain. The estimate for the secondary structure of the 5.8S rRNA is taken from Nazar et al. (54) as adapted to a generalized model for the large subunit RNA (55). The 25S rRNA and the two potential tRNA binding sites are indicated by the shaded line and circles, respectively.

Nucleic Acids Research, 1994, Vol. 22, No. 4 689 Table 1. Effect of gene copy number on mutant RNA expression

Transforming plasmid pFL20/Sp rDNA pFL20/Sp5.8A44i5 pSPI/Sp5.8A44i5 pWI02/Sp5.8A44i5 pDB248/Sp5.8A44i5 a

Plasmid copy numbe?'

Growth rateb

Mutant RNAC

90 90 30 15 8

4.1 6.1 5.6 5.2 4.5

0% 46.5 21.6 13.1 6.7

A + + +

2.3 1.3 1.5 0.9

Published values for transfonring vector H/doubling for logarithmically growing cells in selective medium as

determined by the absorbancy at c Percentage of the total 5.8S rRNA fraction; values are the average of 2-4 determinations, A the standard

b

550nm.

deviation.

recombinants were then used to transform yeast and to express the mutant gene in vivo. A normal gene was also cloned and expressed to assess the effect of plasmid associated rDNA. To permit the efficient detection of mutant RNA in the yeast, the first mutation was a five base insertion which would allow the resulting RNA to be resolved simply by electrophoretic mobility. The insertion was made in a universally conserved GAAC sequence (Fig. 2), a sequence which has previously been suggested to be a putative tRNA-binding site (41). As shown in Fig. 3, this approach was successful in each instance with substantial amounts of mutant RNA being produced and easily detected by gel electrophoresis; the mutant RNA was clearly evident as a slower migrating species. Furthermore, as anticipated, the amount of mutant DNA basically reflected the copy numbers which have been reported for the various shuttle vectors that were used (Table 1). To ensure that the new band was actually mutant 5.8S rRNA, the band was eluted and the RNA was sequenced by chemical degradation (33). As shown in Fig. 4, the five nucleotide insertion (CTAGG) was clearly present disrupting the putative GAAC binding sequence motif. As illustrated with a second isolate (lane e' in Fig. 3), a high level of expression was always present, although some variation in the initial amount of mutant RNA (42-49%)) was observed between different isolates. In each instance, with subculturing the amount stabilized at about 42%. To evaluate the effect of this mutation, the transformants were first characterized with respect to their growth rates. As also indicated in Table 1, the presence of mutated 5.8S rDNA substantially reduced the growth rate. When cells transformed with pFL2OSp5.8A44i5 are compared with cells transformed with the normal 5.8S rRNA sequence, the growth rate in selective medium was reduced by approximately 40% to 6.2 hours/ doubling. To ensure that the reduction in growth rate reflected a real inhibition in protein synthesis, the transformants were also characterized with respect to their protein labeling kinetics in vivo, their ability to support protein synthesis in vitro and the type of protein which is synthesized. As shown in Fig. 5 (left), when proteins were labeled briefly in vivo and chased from ribosomes with unlabeled amino acid, the completion and release of nascent peptides in pFL2OSp5.8A44i5 was entirely consistent with a defect in protein synthesis. These analyses, based on the approach of Dintzis (40) clearly demonstrated a substantial reduction in the rate at which ribosome-associated label was released to the supernatant. Whole cellular protein was also examined in the event that unusual or truncated proteins might be formed. Despite

a

b

c

d

e

e

Figure 3. Electrophoretic fractionation of the Sp5.8A44i5 mutant 5.8S rRNA. Cell RNA was exted from Sc.pombe cells transformed with yeast shuttle vectors containing Sp5.8A44i5 rDNA. The RNA was fractionated on an 8% polyacrylamide gel slab; the 5.8S rRNA fraction was labeled at the 3' end with cytidine [3',5' 32p] bisphosphate and further fractionated on a 40 cm 12% polyacrylamide gel containing 8M urea. After autoradiography, the radioactivity in each band was determined by scintillation counting as indicated in Table 1. RNA was extracted from nontransformed yeast (lane a) or yeast transformed with pDB248/Sp5.8A44i5 (lane b), pWl02/Sp5.8A44i5 (lane c), pSPI/Sp5.8A44i5 (lane d) or pFL20/Sp5.8A44i5 (lane e); RNA from a second isolate with pFL20/Sp5.8A44i5 is shown in lane e'.

the inhibition in protein synthesis, however, the overall profile of cellular proteins appears to be normal (Fig. 5, right). While smaller differences cannot be eliminated, a large increase in truncated or degraded protein clearly was not present. Finally, an inhibition in protein synthesis was confirmed in vitro. Cell free systems were prepared from disrupted cells based on the method of Gassior and co-workers (42) and as illustrated by the example shown in Fig. 6, when normalized for the ribosome concentration, repeated assays consistently demonstrated a significant reduction in protein synthetic activity when mutant extracts were compared with cells transformed with the normal 5.8S rRNA sequence. It may be important to note further that in these studies the degree of inhibition was similar to that which was observed in the growth rate (Table 1). To begin a more detailed characterization of the mechanism of inhibition, polyribosomal and ribosomal subunit profiles were compared both in cells grown in selective medium and with cells in which the demand for protein synthesis was elevated by a short incubation in rich medium. As indicated by the examples shown in Fig. 7, no difference in the subunits was observed when the polysomes were dissociated and the subunits examined (Fig. 7,

690 Nucleic Acids Research, 1994, Vol. 22, No. 4 2000

CtU G

1600

0.. 1200

-

c 0

&ow

p.00 0

4.P

A44

400

I

0

0

15

10

5

Time(MIn)

Figure 6. Effect of 5.8S rRNA mutant Sp5.8A44i5 on protein synthesis in vitro. Cell free extracts were prepared from transformed yeast and assayed for protein synthetic activity as described under 'Materials and Methods'. Incorporation profiles for pFL20/Sp rDNA transformed cell extract, pFL20/Sp rDNA transformed cell extract + 50 ug/ml puromycin and pFL20/Sp5.8A44i5 transformed cell extract which represent the averages of three experiments are indicated by the open circles, close circles and closed squares, respectively; profiles were normalized to reflect equal ribosome concentrations.

REP..

ow

Figure 4. Nucleotide sequence of 5.8S rRNA mutant Sp5.8A44i5. Sc.pombe cells were transformed with pFL20/Sp5.8A44i5 and cellular RNA was extracted, labeled and fractionated as described in Fig. 3. The mutant RNA was eluted and its nucleotide sequence was determined by the chemical degradation methods of Peattie (33). Untreated RNA and the RNA treated with the four base specific reactions are identified as Cd, G, C, A and T, respectively.

2.0

2.0

F 608

A

1.5

I

1.5

L

E

a .0 1.

a A0

Ct,' M ut

0 0.5

0.5 gL I.0

100

80

0

10

20

Fraction c

30

0

15

30

15

30

Fraction

60

...... W

0

co

0

40 0 c

:: 20

0

2

4

6

Time(Min)

8 10

-g

~~Dye

Figure 5. Effect of 5.8S rRNA mutant Sp5.8A44i5 on protein synthesis in vivo. The completion and release of nascent peptides (left) from actively translating ribosomes was determined after brief labeling with 3H-leucine and an unlabeled amino acid chase based on the analyses of Dintzis (40). The polyribosomeassociated radioactivity is indicated for pFL20/SprDNA (open circles) and pFL20/Sp5.8A44i5 (closed circles) cells after the addition of unlabeled leucine. Total cellular protein (right) was prepared from pFL20/SprDNA (Ctl) and pFL2OSp5.8A44i5 (Mut) cells and fractionated on a 10% polyacrylamide as described under 'Materials and Methods'.

Figure 7. Effect of 5.8S rRNA mutant Sp5.8A44i5 on ribosomal profiles. Polyribosomes were prepared from logorithmically growing transformants of Sc.pombe h-leul-32ura4-D18 using glass beads to disrupt the cells as described under 'Materials and Methods' and 0.5 ml aliquots (75-80 A260 units per ml) were fractionated direcdy on 5-40% sucrose gradients for 2h (eft) or dissociated into constituent subunits with 0.8M KCI (39) and fractionated on 5-35% sucrose gradients for 5h. Gradients were subsequently analyzed at 254nm using a Model 640 density gradient fractionator; the fractions indicated by arrows were verified by gel electrophoretic analyses of the extracted constituent RNA. Extracts from pFL20/Sp rDNA and pFL20/Sp5.8A44i5 transformed cells are indicated by the broken and solid profiles, respectively.

right) but the mutant polyribosomal profile reflected an increase in the average size of the polyribosomes (Fig. 7, left), an observation which was consistently made in four replicate experiments. To characterize the distribution of the mutant RNA, total RNA was extracted from the polyribosome fractions and the 5.8S rRNA was quantified in each extract by gel electrophoresis. As indicated in Table 2, essentially an equal

Nucleic Acids Research, 1994, Vol. 22, No. 4 691 Table 2. RNA constituents of polyribosomes from Spombe transformed with normal and mutant 5.8S rDNA

Fraction

Monoribosomes Small polyribosomes Intermediate polyribosomes Large polyribosomes

Mutanta 5.8S rRNA

tRNAb pFL20/SprDNA

pFL20/Sp5.8A44i5

45.5 41.3 42.4 42.1

1.8 + 0.2 1.5 A 0.07 1.2 + 0.10 1.1 0.09

12.6 2.0 2.3 1.9

X + + +

2.3 3.9 4.4 2.1

A + i A

0.15 0.32 0.30 0.15

Polyribosomes were prepared from in vivo labeled cells and fractionated on a sucrose gradient as described in Fig. 7. Individual fractions were pooled as monoribosomes, (fractions 10- 12) and small (fractions 13- 16) intermediate (fractions 17-20) or large (fractions 21-24) polyribosomes; the RNA was extracted with SDS phenol, fractionated on polyacrylamide gels and the radioactivity of each fraction was determined by scintillation counting. Results shown are average for 3 replicate experiments. a Percentage of the total 5.8S rRNA fraction, the standard deviation. b Number of tRNA molecules per ribosome as calculated from the radioactivity in the 5S, 5.8S and tRNA fractionations assuming 1 molecule of 5S and 5.8S rRNA per ribosome, X the standard deviation.

Table 3. Effect of 5.8S rRNA mutations on protein synthesis

Transforming plasmid pFL20/SprDNA pFL20/SpS.8A44i5 pFL20/SpS.8T40C41T42G43T44 pFL20/SpS.8A44i5G104C105 pFL20/SpS.8Al05i4

Plasmid copy numbera

rate'

In vitro protein synthesisc

88 84 87 87 55

4.1 6.2 5.6 9.2 6.3

100 48 70 29 55

Growth

Relative copy number based on a published rRNA gene copy number of 120 copies per genome (25). H/doubling for logarithmically growing cells in selective medium as determined by the absorbancy at 550nm. c Incorporation rate relative to extracts from cells transformed with pFL20/SprDNA were assayed as described in Fig. 5; values are averages of 2-3 determinations. a

b

amount of mutant 5.8S rRNA was present in each fraction with only a very slight increase in the monoribosome peak when compared with the polyribosome fractions. Because previous studies have suggested a potential role in tRNA binding, levels of bound tRNA also were assessed in each fraction. As shown in Table 2, repeated analyses indicated that the polyribosomes of mutant cells contained elevated amounts of tRNA which on a molar basis were approximately inversely proportional to the observed reduction in growth rate or in vitro protein synthesis. Neither the elevation in polyribosome size nor in bound tRNA were observed with cells transformed with normal rDNA (Fig. 8 and Table 2), eliminating the possibility that these effects were simply the result of an elevated rDNA copy number. To further confirm a direct effect by 5.8S rRNA mutation and to evaluate contributions by both of the conserved GAAC sequences (Fig. 2) additional mutations were introduced, either individually or collectively using a 2-step PCR-based targeted mutagenesis (24). The basic effects of these mutations are summarized in Table 3. As indicated, substitutions (pFL20/ Sp5.8T40C41T42G43T44) and insertions (pFL20/Sp5 .8A44i5) in the GAAC site 1 or site 2 (e.g. pFL/Sp5.8AlO5i4) had inhibitory effects on both the growth rate and protein synthesis. Furthermore, when a double mutation was prepared in which both sites were modified (pFL20/Sp5.8A44i5G104C105), the effects were cumulative with a further significant inhibition in growth rate and protein synthesis. The cumulative effect with a double mutant permitted a more critical evaluation of the modest increase in the polyribosome

2.0

80S

I.

Ec N

1.5O

U

.0 j) 0.5

0

10

20

30

Fraction

Figure 8. Polyribosomes in Sc.pombe transformed with pFL20/Sp5.8A44i5G104C105. Polyribosomes were prepared from logarithmically growing Sc.pombe h-leul-32ura4-D18, transformed with pFL20/Sp5.8A44i5G104C105 (profile), pFL20/Sp5.8A44i5 (lightly shaded profile) or pFL20/SprDNA (darkly shaded profile) using Tnichodenna lysing enzymes to disrupt the cells as described under 'Materials and Methods' and 0.5 ml aliquots were fractionated on 5-40% sucrose gradients as described in Fig. 7. Gradients were subsequently analyzed at 254nm using a Model 640 density gradient fractionator; the ribosome fraction was verified by gel electrophoretic analysis of the extracted constituent RNA. Mutant cells were grown using a glucose carbon source while pFL20/SprRNA transformed cells were grown using glycerol.

692 Nucleic Acids Research, 1994, Vol. 22, No. 4 size with pFl2OSpS.8A44i5 (Fig. 7). As shown in Fig. 8, when polyribosomes were prepared from the double mutant (pFL20/Sp5.8A44i5G104C105) and the profile was compared with the previous mutant (pFL20/Sp5.8A44i5), the size was clearly further elevated. These differences are even more apparent when the profiles are compared with polyribosomes from a normal transformant (pFL20/SprDNA) grown at a more comparable growth rate (7.2H/doubling) using glycerol as a carbon source. In this case, even the single mutation clearly contains a substantially elevated polyribosome size.

DISCUSSION The mutagenesis strategy illustrated in this study has a broad application both for the investigation of 5.8S rDNA expression, structure and function and for studies on the eukaryotic rDNA, in general. The efficient expression of a clearly identifiable product in vivo which subsequently can be easily prepared for studies in vitro has obvious advantages. Previous studies on rRNA processing (refs. 11, 12) have successfully utilized low copy vectors but conclusions must be qualified when mutants are studied under very non-stoichiometric conditions and a very low product yield does not permit the isolation of key intermediates. As observed with 5S rDNA genes (13), the present study shows that plasmid integrated rDNA also can compete very effectively with the chromosomal copies. Under appropriate conditions a significant portion of the cellular RNA can be mutant and when integrated into ribosomes the ribosomal population also can be up to 50% mutant. In addition, the use of alternate copy number shuttle vectors and differing growth conditions permits the experimental adjustment of mutant RNA concentrations (e.g. Table 1). The specific mutation which was initially characterized as an easily identifiable marker for gene expression and DNA processing immediately raises important questions about 5.8S rRNA function; the inhibition in cell growth and protein synthesis (e.g. Fig. 5) again supports an important function. Almost two decades ago Nishikawa and Takemura (41) first noted GAAC sequences in the eukaryotic 5.8S rRNA which they speculated could be candidates for a tRNA binding site that interacts with the highly conserved T*CG arm of tRNA. Indeed, subsequent accessibility studies (43, 44) suggested that the 5.8S rRNA could be localized in the ribosomal interface. Other studies utilizing temperature denaturation (45) or affinity chromatography revealed ternary (46) or quaternary (47) complexes of the SS and 5.8S rRNAs and ribosomal proteins or tRNAs, and a probe of the ribosome structure utilizing chemical cross-linking (48) has shown that some 5.8S rRNA binding proteins are proximal to the A-site. More recent studies illustrating the inhibition of protein synthesis by anti 5.8S rRNA oligonucleotides have provided additional evidence for an important role with a specific effect on a conserved GAAC sequence (18). As a result, the inhibition of growth and protein synthesis by insertions or mutations in the GAAC sequences in the present study provide complementary and more direct evidence for this functional role although the elevated levels in tRNA (Table 2) make a direct interaction less likely. The analyses of polyribosomes and associated 5.8S and tRNAs in yeast cells transformed with pFL2OSp5.8A44i5 begin to more precisely define the ribosomal function with which the mutation interferes. An essentially equal distribution of mutant RNA in monosomes and polyribosomes indicates that ribosomes

containing mutant RNA can efficiently initiate protein synthesis. The elevated polyribosomal profile with one deficient GAAC site and the substantially more elevated profile with two defective sites strongly confirms a defect which occurs after initiation, but some continued protein synthesis indicates that elongation and termination are not blocked completely. This is confirmed by the kinetic analysis of peptide completion and release (Fig. 5, left). In this case, labeled nascent protein was clearly chased from ribosomes at a reduced rate consistent with a less efficient elongation cycle. An apparently normal profile of cellular protein after separation by gel electrophoresis without any sign of unusual short peptides or degradation products (Fig. 5, right) is consistent with the polypeptides not being prematurely released but the elevated ribosome-associated tRNA level (Table 3) also suggests an inefficient release of tRNA. All ribosomes appear to contain at least three tRNA binding sites identified as A, P and E (49), however, a recent report on eukaryotic ribosomes (51) has provided strong evidence for a fourth site (S), a site that may allosterically modulate the function of the ribosome. In the course of normal protein synthesis, the binding of a new A-site ligand triggers the release of the deacyl tRNA bound to the E site (4-53) tRNA. The elevated tRNA level, therefore, could reflect a defect in the trigger or release mechanism which results in slower release kinetics for the deacylated tRNA and an overall elevation. To resolve these or other possibilities specific assays are being undertaken to precisely determine the actual step which is affected as well as structural studies to determine if the inhibitory effects is the result of a direct interaction with a GAAC sequence, a more global rearrangement of the 5.8S rRNA structure or even a longer range ribosomal readjustment. Further mutations also are being introduced to determine other regions of the 5.8S rRNA which may affect protein synthesis. In the interim, this study clearly illustrates the potential effectiveness of targeted mutagenesis in studies on eukaryotic rRNA function and provides new in vivo evidence of a functional role for the eukaryotic 5.8S ribosomal RNAs.

ACKNOWLEDGEMENTS The authors are greatly indebted to Dr. E. Morgan (Roswell Park Institute) for discussing his unpublished results on rDNA expression in S.cerevisiae. This study was supported by grants from the Medical Research Council and the Natural Sciences and Engineering Research Council of Canada.

REFERENCES 1. 2. 3. 4. 5. 6. 7.

8. 9. 10. 11. 12.

Cech, T.R. and Bass, B.L. (1986) Ann. Rev. Biochem. 55, 599-629. Moazed, D. and Noller, H.F. (1989) Nature 342, 142-148. Dahlberg, A.E. (1989) Cell 57, 525-529. Noller, H.F. (1991) Ann. Rev. Biochem. 60, 191-200. Noller, H.F., Hoffarth, V. and Zimmink, L. (1992) Science 256, 1416-1419. Piccirilli, J.A., McConnell, T.S., Zan, A.J., Noller, H.F. and Cech, T.R. (1992) Science 256, 1420-1424. Kunkel, T.A., Roberts, J.D. and Zakour, R.A. (1987) Methods Enzymol. 154, 367-382. Heyer, W.D., Sipiczki, M. and Kohli, J. (1986) Molec. and Cell. Biol. 6, 80-89. Percival-Smith, A. and Segall, J. (1987) Mol. Cell. Biol. 7, 2484-2490. Sikorski, R.S. and Hieter, P. (1989) Genetics 122, 19-27. Musters, W., Venema, J., Van der Linden, G., Van Heerikhuizen, H., Klootwijk, J. and Planta, R.J. (1989) Mol. Cell. Biol. 9, 551-559. Van der Sande, C.A.F.M., Kwa, M., Van Nues, R.W., Heerikhuizen, H.V., Raue, H.A. and Planta, R.J. (1992) J. Mol. Biol. 223, 899-910.

Nucleic Acids Research, 1994, Vol. 22, No. 4 693 13. Van Ryk, D.I., Lee, Y. and Nazar, R.N. (1990) J. Biol. Chem. 265, 8377-8389. 14. Van Ryk, D.I., Lee, Y. and Nazar, R.N. (1992) J. Biol. Chem., 267, 16177-16181. 15. Nazar. R.N. (1980) Cell Nucleus 11, 1-28. 16. Pene, J.J., Knight, E., Jr. and Darnell, J.E. Jr. (1968) J. Mol. Biol. 33, 609-623. 17. Weinberg, R.A. and Penman, S. (1968) J. Mol. Biol. 38, 289-304. 18. Walker, K., Abou Elela, S., Nazar, R.N. (1990) J. Biol. Chem. 265, 2428-2430. 19. Cryer, D.R., Eccleshall, R. and Marmuur, J. (1975) Methods Cell Biol. 12, 39-44. 20. Chuanpit, 0. (1987) Ph.D. Thesis. University of Guelph. 21. Koecka, M. (1975) Folia Microbiol. 20, 273-276. 22. Dunn, I.S. and Blattner, F.R. (1987) Nucleic Acids Res. 15, 2677-2698. 23. Benton, W.D. and Davis, R.W. (1977) Science 196, 180-182. 24. Good, L. and Nazar, R.N. (1992) Nucleic Acids Res. 20, 4934-4935. 25. Schaak, J., Mao, J. and Soll, D. (1982) Nucleic Acids Res. 10, 2851 -2863. 26. Hanahan, D. (1983) J. Mol. Biol. 166, 557-580. 27. Sanger, F., Nicklen, A. and Coulson, A.R. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 5463-5468. 28. Maniatis, T., Fritsch, E.F. and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. 29. Clewell, D.B. and Helinski, D.C. (1969) Proc. Nati. Acad. Sci. U.S.A. 62, 1159-1167. 30. Okazaki, K., Okazaki, N., Kune, K., Jinno, S., Tanala, K. and Okayama, H. (1990) Nucleic Acids Res. 18, 6485-6489. 31. Steele, W.J., Okaura, N. and Busch, H. (1965) J. Biol. Chem. 240, 1742- 1749. 32. Rubin, G.M. (1973) J. Biol. Chem. 248, 3860-3875. 33. Peattie, D. (1979) Proc. Nail. Acad. Sci. U.S.A. 76, 1760-1764. 34. Peacock, A.C. and Dingman, C.W. (1967) Biochemistry 7, 659-674. 35. Leupold, U. (1955) Arch. Julus Klous-Stift, Vererbungsforsch Sozialathropol, Rassenkyg 30, 506-511. 36. Picard, B. and Wegnez, M. (1979) Proc. Nadl. Acad. Sci. U.S.A. 76, 241 -245. 37. Gorenstein, C. and Warner, J.R. (1976) Proc. Nadl. Acad. Sci. U.S.A. 73, 1547-1551. 38. Laemmli, V.K. (1970) Nature 227, 680-685. 39. Torano, A., Sandoval, A. and Heredia, C.F. (1974) Methods Enzymol. 30, 254-261. 40. Dintzis, H.M. (1961) Proc. Natl. Acad. Sci. U.S.A. 47, 247-261. 41. Nishikawa, K. and Takemura, S. (1974) FEBS Lett. 40, 106-109. 42. Gassior, E., Hevena, F., Saduik, I., McLaughlin, C.S. and Moldaves, K. (1979) J. Biol. Chem. 254, 3965-3969. 43. Lo, A.C. and Nazar, R.N. (1982) J. Biol. Chem. 257, 3516-3524. 44. Liu, W., Lo, A.C. and Nazar, R.N. (1983) J. Mol. Biol. 171, 217-224. 45. Nazar, R.N. (1978) J. Biol. Chem. 253, 4505-4507. 46. Metspalu, A., Toots, I., Saarma, M. and Villems, R. (1980) FEBS Lett. 119, 81-84. 47. Ulkich, N., Chan, Y.L., Huber, P.W. and Wool, I.G. (1982) J. Biol. Chem. 257, 11353-11357. 48. Lee, J.C. and Trout, R.R. (1984) J. Biol. Chem. 259, 99871-9974. 49. Rheinberger, H.J., Stemback, H. and Nierhaus, K.H. (1981) Proc. Nail. Acad. Sci. U.S.A. 78, 5310-5314. 50. Cottarel, G. (1993) Curr. Genet. 23, 547-548. 51. Rodnina, M.V. and Wintermeyer, W. (1992) J. Mol. Biol. 228, 450-459. 52. Beach, D. and Nurse, P. (1981) Nature 290, 140-142. 53. Hausner, T.P., Geigenmililer, U. and Nierhaus, K.H. (1988) J. Biol. Chem. 263, 13103-13111. 54. Nazar, R.N., Sitz, T.O. and Busch, H. (1975) J. Biol. Chem. 250, 8591 -8597. 55. Gutell, R.R. and Fox, G.E. (1988) Nucleic Acids Research 16, r175-r269.