Analysis of Mutations in the Yeast mRNA Decapping Enzyme

Corresponding author: Roy Parker, Departments of Molecular and Cellular Biology and Biochemistry and the Howard Hughes Medical Institute, Life Sciences South Building, East Lowell St., P.O. Box 210 106, University of Arizona, Tucson, AZ 85721-0106., rrparker{at}u.arizona.edu (E-mail)

Communicating editor: F. WINSTON

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

A major mechanism of mRNA decay in yeast is initiated by deadenylation, followed by mRNA decapping, which exposes the transcript to 5' to 3' exonucleolytic degradation. The decapping enzyme that removes the 5' cap structure is encoded by the DCP1 gene. To understand the function of the decapping enzyme, we used alanine scanning mutagenesis to create 31 mutant versions of the enzyme, and we examined the effects of the mutations both in vivo and in vitro. Two types of mutations that affected mRNA decapping in vivo were identified, including a temperature-sensitive allele. First, two mutants produced decapping enzymes that were defective for decapping in vitro, suggesting that these mutated residues are required for enzymatic activity. In contrast, several mutants that moderately affected mRNA decapping in vivo yielded decapping enzymes that had at least the same specific activity as the wild-type enzyme in vitro. Combination of alleles within this group yielded decapping enzymes that showed a strong loss of function in vivo, but that still produced fully active enzymes in vitro. This suggested that interactions of the decapping enzyme with other factors may be required for efficient decapping in vivo, and that these particular mutations may be disrupting such interactions. Interestingly, partial loss of decapping activity in vivo led to a defect in normal deadenylation-dependent decapping, but it did not affect the rapid deadenylation-independent decapping triggered by early nonsense codons. This observation suggested that these two types of mRNA decapping differ in their requirements for the decapping enzyme.

THE turnover of mRNA is an important control point in the regulation of gene expression. Though mRNA decay mechanisms vary with organisms and the nature of the mRNA, several commonalities do exist among the various eukaryotes (ROSS 1995 ; JACOBSON and PELTZ 1996 ; THARUN and PARKER 1997 ). For example, in many eukaryotes, shortening of the poly(A) tail forms an initial step that is followed by the degradation of the body of the mRNA (BEELMAN and PARKER 1995 ). In Saccharomyces cerevisiae, mRNA deadenylation leads to decapping of the mRNA, which is followed by rapid 5' to 3' exonucleolytic degradation of the mRNA (MUHLRAD and PARKER 1992 ; DECKER and PARKER 1993 ; HSU and STEVENS 1993 ; MUHLRAD et al. 1994 , MUHLRAD et al. 1995 ). Several observations suggest that decapping and 5' to 3' decay may also occur in other eukaryotes. For example, full-length mRNAs that are devoid of both the cap and most of the poly(A) tail have been detected from murine liver cells (COUTTET et al. 1997 ). Similarly, mRNA decay intermediates that are shortened at their 5' ends have been identified in both plant and animal cells (LIM and MAQUAT 1992 ; HIGGS and COLBERT 1994 ; GERA and BAKER 1998 ). Furthermore, homologs of the yeast 5' to 3' exonuclease Xrn1p that degrades mRNA after decapping have been identified in several organisms (HEYER et al. 1995 ; BASHKIROV et al. 1997 ).

Decapping is a key step in the turnover of yeast mRNAs, because variation in decapping rates accounts for part of the differences in decay rates of specific mRNAs in yeast (MUHLRAD et al. 1994 , MUHLRAD et al. 1995 ). Decapping also occurs in a process termed mRNA surveillance, whereby mRNAs containing a premature translational stop codon are rapidly decapped and degraded by a 5' to 3' exonucleolytic process. In this case, however, decapping occurs independently of deadenylation (MUHLRAD and PARKER 1994 ). Thus, given the central role of decapping in mRNA decay, resolution of the mechanisms by which mRNA decapping occurs and is controlled will be critical for understanding mRNA turnover.

Earlier studies have identified the yeast DCP1 gene as encoding a decapping enzyme that is required for mRNA decay in vivo and sufficient for decapping activity in vitro (BEELMAN et al. 1996 ; LAGRANDEUR and PARKER 1998 ). This decapping enzyme cleaves capped mRNAs within the cap 5'–5' triphosphate linkage to release m⁷GDP and a 5'-monophosphate mRNA (BEELMAN et al. 1996 ). In strains deleted for the DCP1 gene, many mRNAs are stabilized as a result of a block in mRNA decapping (BEELMAN et al. 1996 ). These mRNAs include stable and unstable mRNAs that undergo deadenylation-dependent decapping, as well as mRNAs with early nonsense codons that are degraded by deadenylation-independent decapping. These results indicate that the decapping enzyme Dcp1p is required for most, if not all, mRNA decapping in vivo. Elucidation of the function and regulation of Dcp1p is, therefore, important for understanding decapping.

In the present work, we used alanine-scanning mutagenesis to create 31 different mutant versions of the decapping enzyme, and we examined the effects of the mutations both in vivo and in vitro. Two types of mutations that affected mRNA decapping were identified. One set of mutations resulted in strong loss of function both in vivo and in vitro, indicating that they affected specific amino acid residues important for enzymatic function. The second set of mutations resulted in moderate, or in some allelic combinations, strong loss of function in vivo, but they failed to cause any loss of activity of the protein in vitro. This suggests that the amino acid residues that were changed in these mutants may be required for some functionally important in vivo interactions of Dcp1p with other proteins that were not present in the purified in vitro system. Our studies also revealed that mutants with partial loss of function in vivo were not defective for the rapid deadenylation-independent decapping triggered by early nonsense codons, and that they were defective only for the normal deadenylation-dependent decapping in vivo, suggesting a difference between these two processes with regard to their requirement for the decapping enzyme.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Oligonucleotide-directed mutagenesis:
All mutants of DCP1, except dcp1-41 to -44, were generated by this procedure. For performing site-directed mutagenesis, the DCP1 gene was cloned into the yeast shuttle vector pUN45 (ELLEDGE and DAVIS 1988 ), which contains the M13 origin of replication. pUN45 is a CEN vector with the yeast TRP1 gene as an auxotrophic marker. The entire DCP1 gene present in the plasmid p424DCP1 (BEELMAN et al. 1996 ) was excised as an 1.7-kb fragment by digesting p424DCP1 with enzymes ApaI and NotI, and it was then inserted into the pUN45 vector digested with the same two enzymes. This resulted in the plasmid pRP783, which was used for making all the mutants. The single-stranded form of this plasmid was made after transforming it into Escherichia coli XL1 Blue cells and infecting them with M13K07 helper phage. Site-directed mutagenesis was then performed following standard methods by annealing mutagenic oligonucleotide to the single-stranded form of pRP783, extending it with Klenow to complete the second strand, and closing the second strand with ligase. After this, a portion of the mutagenesis reaction mixture was transformed into E. coli TB1 cells, and plasmid minipreps made from several transformants were screened by restriction analysis (using pRP783 as control) to find those bearing the mutated DCP1 gene. This yielded plasmids pRP872–pRP898, which bore the various primary mutant forms of DCP1 (Table 1). These plasmids were then transformed into dcp1 strain yRP1071 (BEELMAN et al. 1996 ), and the transformants were studied for their RNA degradation-related phenotypes to assess the degree of complementation provided by the plasmid-borne mutant dcp1.

Mutants dcp1-41 to -44 were made by combining mutations of allele dcp1-17 or dcp1-4 with mutations of dcp1-19 or dcp1-25. The mutations of dcp1-17 and dcp1-4 occur close to the N terminus of Dcp1p, while the dcp1-19 and dcp1-25 mutations affect sites closer to the C terminus of the Dcp1p sequence. Therefore, the combinations were made by exchanging the region coding for the N-terminal portion of Dcp1p in vectors pRP886 and pRP889 with those derived from pRP885 and pRP874 by restriction digestion and ligation.

RNA preparation and analysis:
For steady-state RNA analysis, cells were grown to midlog phase in SC–Trp minimal medium containing sucrose and galactose as carbon sources. For mRNA half-life determination by transcriptional repression, cells were grown as described above to midlog and then shifted to SC–Trp minimal medium with glucose as a carbon source. RNA preparation and analysis were done as described previously (CAPONIGRO et al. 1993 ). Levels of full-length species and poly(G)3' end fragments of MFA2pG and PGK1pG mRNAs were determined by probing Northern blots with 5'-³²P end-labeled oligonucleotides [capable of hybridizing to both full-length species and poly(G)3' end fragments] specific for the respective mRNAs (CAPONIGRO and PARKER 1995 ) and by quantitating the signal with a PhosphorImager (Molecular Dynamics). To determine the levels of CYH2 mRNA and CYH2 pre-mRNA, a random-primed cDNA probe capable of specifically hybridizing to both of them was used, and quantitation was done using a PhosphorImager, as described above. MFA2pG mRNA half-life measurements were performed by the method of transcriptional repression from GAL1 UAS by glucose, as described before (PARKER et al. 1991 ). For each time point, the time of freezing (in dry ice) the cell pellet was noted and used for drawing the decay curve.

Construction of the FLAG-Dcp1p overexpression vector:
The FLAG-fused chimeras of mutant DCP1 genes were created by PCR amplification of the mutant DCP1 sequences from their corresponding CEN vector clones (pRP872–pRP898, see above and Table 1) using the oligonucleotide GCAGCACCGGATCCATGGACTACAAGGACGACGATGACAAGATGACCGGAGCAGCAAC (oRP311), which places the FLAG-coding sequence 5' of the DCP1 gene, as well as the oligonucleotide GCAGCACCGTCGACTTCTCACTTGGGCATCTC (oRP312). This PCR fragment was digested with BamHI and SalI, and it was ligated into the yeast expression vector pG-1 (SCHENA et al. 1991 ) between the BamHI and SalI sites. This way, FLAG fusion chimeras of the alleles dcp1-2 (pRP900), dcp1-4 (pRP901), dcp1-7 (pRP902), dcp1-17 (pRP903), dcp1-19 (pRP904), dcp1-25 (pRP905), and dcp1-31 (pRP906) were cloned. These 2µ plasmids (with the TRP1 marker) express FLAG-Dcp1p from the constitutive glyceraldehyde-3-phosphate dehydrogenase promoter.

Purification of FLAG-Dcp1p:
Wild-type and mutant FLAG-Dcp1p proteins were purified from dcp1 cells (yRP1071) carrying the plasmid pRP801 (LAGRANDEUR and PARKER 1998 ) and the 2µ clones of the mutant DCP1 genes (see above), respectively. A 200-ml culture of the yRP1071 cells harboring the appropriate 2µ plasmid was grown to late-log phase in SC–Trp minimal medium containing 2% glucose, and it was harvested by centrifugation. Cells were washed in buffer A (10 mM Tris-HCl, pH 7.6, 100 mM potassium acetate, 2 mM magnesium acetate, and 2 mM 2-mercaptoethanol), spun down, and suspended in 1.5 ml/g cell pellet weight of the same buffer containing COMPLETE (Boehringer Mannheim, Indianapolis, IN) protease inhibitors. An equal volume of acid-washed glass beads was added to the suspension, and cells were lysed by six cycles of vortexing for 15 and 45 sec of cooling in ice water. The lysate was centrifuged at 18,000 x g for 15 min at 4°, and the supernatant was chromatographed at 4° by gravity flow through a 0.5-ml anti-FLAG M2 monoclonal antibody immuno-affinity column (Kodak) equilibrated with buffer A. The flowthrough was collected and reapplied to the column. The column was sequentially washed with 15 bed volumes each of buffer AN (buffer A with 0.05% NP-40), buffer ANS (buffer AN with 0.7 M potassium acetate), and buffer AN. FLAG-Dcp1p was eluted with 5 bed volumes of buffer AN containing 200 µg/ml FLAG peptide. The FLAG peptide present in the FLAG-Dcp1p preparation was removed by dialysis, and the FLAG-Dcp1p was finally stored in 20 mM Tris, pH 7.5, with 5 mM DTT and 20% (v/v) glycerol.

The FLAG-Dcp1p preparation was analyzed by standard SDS-PAGE methods on a 10% gel (LAEMMLI 1970 ). Protein size markers were purchased from GIBCO BRL (Gaithersburg, MD). Silver staining of SDS-PAGE gels was done using the Silver Stain Plus kit from Bio-Rad (Richmond, CA) following the manufacturer's protocol.

Preparation of cap-labeled substrate for in vitro decapping assay:
Uncapped mRNAs lacking poly(A) tails were synthesized in vitro by T7 RNA polymerase runoff transcriptions. Full-length MFA2 mRNA was transcribed from plasmid pRP802 (LAGRANDEUR and PARKER 1998 ) that was linearized with EcoRI to produce a 343-nucleotide mRNA consisting of 325 nucleotides of MFA2 sequence, followed by 18 nucleotides encoded by the plasmid polylinker.

T7 transcriptions were done in 100-µl reactions containing 1–2 µg of template DNA, 5 mM NTPs, 40 mM Tris-HCl, pH 8.0, 1 mM spermidine, 5 mM DTT, 50 µg/ml BSA, 0.01% Triton X-100, 20 mM MgCl₂, 5 units yeast inorganic pyrophosphatase (Sigma, St. Louis, MO) and 40 units T7 RNA polymerase (Boehringer Mannheim) at 37° for ~15 hr (HARRIS et al. 1994 ). The uncapped transcripts were purified by polyacrylamide electrophoresis.

7-Methyl caps were added to in vitro-synthesized MFA2 transcripts in reactions typically containing 23 pmol RNA, 45 pmol [-³²P]GTP, 40 units RNasin (Promega, Madison, WI), 0.67 mM S-adenosylmethionine, 50 mM Tris-HCl, pH 7.6, 2 mM MgCl₂, 6 mM KCl, 1 mM DTT, and 25 units guanylyltransferase (GIBCO-BRL) at 37° for 2 hr (BEELMAN et al. 1996 ). As a result of the qualitative nature of capping by guanylyltransferase, these conditions typically capped only 35–40% of the RNA in the reaction (SHUMAN and MOSS 1990 ). Cap-labeled RNAs were separated from unincorporated label by Sepharose-CL6B (Pharmacia, Piscataway, NJ) spin chromatography.

Decapping assays:
Decapping assays used in this study were similar to those described previously (BEELMAN et al. 1996 ). m7G[³²P]pppMFA2 mRNA was incubated with Dcp1p preparation in 50 mM HEPES, pH 7.0, containing 1 mM MgCl₂, 0.05% NP-40, and 1 mM DTT in a total reaction volume of 15 µl at 30°. Reactions were stopped by addition of EDTA to a final 50-mM concentration and shifting the sample to ice. To follow the reaction time course, aliquots of the reaction mixture were drawn at different time points, mixed with EDTA, and placed on ice. The product of the reaction (-³²P-m7GDP) was separated from the unreacted substrate by PEI-cellulose thin-layer chromatography developed in 0.45 M (NH₄)₂SO₄. Activity at any time t is calculated by dividing the amount of product formed at time t by the total amount of substrate taken for the reaction (the sum of the amount of product formed and the amount of unreacted substrate at time t). The activity values were then normalized for the amount of Dcp1p protein in the sample to determine the specific activity values, which were used to plot graphs for the time courses. Quantitations of product formed and unreacted substrate left after the reaction were done using a Molecular Dynamics PhosphorImager. To determine the amount of Dcp1p protein in the purified preparations, Western blot of purified samples was probed with anti-Dcp1p antibodies, and the Dcp1p band intensity in the autoradiograph was then quantitated using the IP Lab Gel program. In addition, for careful comparison, protein concentrations of the purified enzymes were adjusted and then the activity was reassayed where equal amounts of purified Dcp1p could be compared directly (see Figure 6 for example). Whenever a set of purified Dcp1p samples were compared for specific activity, they were all assayed together with wild-type Dcp1p using the same substrate preparation. Assays done with different substrate preparations were not compared.

View larger version (39K): In this window In a new window Download PPT slide   Figure 1. Positions (on the amino acid sequence) of the various primary mutations of DCP1 generated in this study. The amino acid sequence of Dcp1p is shown. Underlines indicate the parts of Dcp1p sequence that are targeted in different primary mutants, and the actual residues mutated are in boldface letters. The allele number of each mutant (preceded by #) is given below the corresponding underlines. All amino acid changes are to alanines, except where indicated (two cases, dcp1-28 and dcp1-35). The numbers shown above the sequence are the amino acid residue numbers starting from the N terminus. The allele dcp1-1 (Gly156 to Asp, shown in italics) was isolated earlier (HATFIELD et al. 1996 ) by screening mutants generated by ethyl methanesulfonate mutagenesis.

View larger version (74K): In this window In a new window Download PPT slide   Figure 2. Half-life of MFA2pG mRNA in different dcp1 primary mutants. Shown in the figure are MFA2pG mRNA half-life measurements in different dcp1 mutants (grown at 30°) on Northern blots after transcriptional shutoff by shifting galactose-induced cells to glucose medium. The time points at which cells were collected for RNA isolation are indicated on top of the gel pictures. The name of the dcp1 allele and the MFA2pG half-life determined are shown on the right of each gel picture. Northern analysis using an oligonucleotide probe specific for MFA2pG mRNA and half-life determination was done as described in MATERIALS AND METHODS.

View larger version (26K): In this window In a new window Download PPT slide   Figure 3. Temperature-sensitive RNA decay phenotype of alleles dcp1-2 and dcp1-34. Wild-type, dcp1, dcp1-2, and dcp1-34 cells were grown at the temperatures indicated above the lanes in galactose-containing minimal medium, and RNA made from them was analyzed by Northern analysis using an oligonucleotide probe specific for MFA2pG mRNA, as described in MATERIALS AND METHODS.

View larger version (47K): In this window In a new window Download PPT slide   Figure 4. Half-life of MFA2pG mRNA at different temperatures in the allele dcp1-2. Shown is the MFA2pG mRNA half-life measurement in wild-type and dcp1-2 cells grown at 18° and 36° and in dcp1 cells (carrying only the pUN45 vector without insert) grown at 36° on Northern blots after transcriptional shutoff by shifting galactose-induced cells to glucose medium. The time points at which cells were collected for RNA isolation are indicated on top of the gel pictures. Allele name, growth temperature, and the MFA2pG half-life determined are shown on the right of each gel picture. Northern analysis using an oligonucleotide probe specific for MFA2pG mRNA and half-life determination were done as described in MATERIALS AND METHODS.

View larger version (31K): In this window In a new window Download PPT slide   Figure 5. SDS-PAGE analysis of various purified FLAG-Dcp1p samples: FLAG-tagged decapping enzyme was purified from wild-type cells and dcp1 mutant cells using anti-FLAG antibody affinity column, subjected to SDS-PAGE, and silver stained as described in MATERIALS AND METHODS. The dcp1 allele, from which the protein was purified, is shown on top of each lane. The Dcp1p band is indicated by an arrow on the left. The positions of molecular weight markers are shown on right. The slight differences seen in the mobilities of some of the mutant proteins are not reproducible.

View larger version (96K): In this window In a new window Download PPT slide   Figure 6. In vitro decapping activity of FLAG-tagged forms of wild-type and mutant Dcp1p proteins. Mutant and wild-type Dcp1p proteins were expressed as FLAG fusion proteins from the constitutive GPD promoter in a 2µ vector (see MATERIALS AND METHODS) in a dcp1 background and were purified using FLAG antibody column, as described in MATERIALS AND METHODS. Purified proteins were assayed in vitro for decapping activity, as described in MATERIALS AND METHODS. (A) Time course of decapping reaction carried out with wild-type and mutant FLAG-tagged decapping enzymes. Aliquots of the reaction mixture were drawn at 7, 14, and 25 min, resolved by TLC after stopping the reaction, and activity was quantitated and normalized for protein amounts, as described in MATERIALS AND METHODS. (B) Phosphorimage of a typical TLC run. Decapping reactions were carried out with the purified wild-type and mutant FLAG-Dcp1p samples for 20 min, after which the reaction was stopped and TLC was performed. The position of the product of the reaction, m7GDP and the origin where the unreacted substrate stays in the TLC plate are indicated by arrows on the right side of the phosphorimage. Shown below the TLC picture is a Western analysis of the FLAG-Dcp1p samples used in the assay (slight differences seen in the mobilities of some of the mutant proteins are not reproducible). The FLAG-Dcp1p samples used for the Western analysis and the assay are indicated below the lanes of the Western blot. For every FLAG enzyme sample, the amounts used for the assay shown in the TLC picture and the Western analysis are equal.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Mutagenic strategy:
For the analysis of Dcp1p, we used site-directed mutagenesis to change residues that are expected to be important for function. Two approaches were used. First, we used the strategy of charged-to-alanine scanning mutagenesis. In this procedure, clusters of charged amino acids in the primary sequence, which are likely to be on the surface of the folded protein, were identified and systematically changed to alanines (WELLS 1991 ). In the second approach, we targeted some residues that might be important on the basis of comparison to other proteins that interact with the cap site. For example, because the cap structure is often recognized by a stacking interaction between aromatic residues (HODEL et al. 1997 ; MARCOTRIGIANO et al. 1997 ), we targeted the aromatic residues Trp56, Trp204, and Tyr47 of DCP1. All mutants except dcp1-41 to -44 were made by standard oligomutagenesis on a copy of the DCP1 gene cloned in a centromere plasmid under its own promoter (see MATERIALS AND METHODS). Mutants dcp1-41 to -44 were made by different pairwise combinations of the mutations of dcp1-4, -17, -19, and -25 by restriction digestion and ligation (see MATERIALS AND METHODS and Table 1). In most (26 out of 31) of the mutants, 3 residues were changed. Five residues each were affected in dcp1-41 and -42, and four each were changed in dcp1-40, -43, and -44 (Table 1). In each case, it was confirmed by sequencing that the mutant DCP1 gene contained only the targeted mutations. The positions of all mutations on the primary sequence are shown in Figure 1.

Effects of the mutations on deadenylation-dependent decapping in vivo:
To study the effect of the mutations in the DCP1 gene, we determined if they affected mRNA decapping in vivo. Loss of decapping enzyme function (as in dcp1) leads to a block in mRNA decay in vivo (BEELMAN et al. 1996 ) because the 5'3' exonucleolytic digestion of mRNA in vivo (by Xrn1p) is dependent on the removal of mRNA cap by the DCP1 gene product. The dcp1 mutant alleles were expressed from a centromere plasmid in the dcp1 yeast strain yRP1071 (BEELMAN et al. 1996 ). The ability of the different dcp1 mutants to degrade mRNAs was assessed by estimating the decay of both the unstable MFA2pG and stable PGK1pG mRNAs in vivo. Both of these mRNAs are known to be degraded by deadenylation-dependent decapping, leading to 5' to 3' digestion by the exonuclease Xrn1p (MUHLRAD et al. 1994 , MUHLRAD et al. 1995 ). In the yRP1071 strain, the MFA2pG and PGK1pG genes (driven by the GAL promoter) are expressed from a chromosomal location, and they contain a poly(G) tract insertion in their 3' untranslated regions. The poly(G) tract does not affect the decay of these transcripts, and it serves to block 5' to 3' exonucleolytic degradation in cis (DECKER and PARKER 1993 ). As a result, when the decapped mRNA is degraded by Xrn1p, it leads to the accumulation of a fragment of the mRNA which lacks the part of mRNA on the 5' side of the poly(G) stretch as a degradation intermediate, termed the poly(G)3' end fragment. In the absence of any decapping, however, the action of Xrn1p and, hence, the formation of a mRNA fragment resulting from the poly(G) tract is blocked (BEELMAN et al. 1996 ). Thus, the relative levels of the full-length species and poly(G)3' end fragment of the mRNA can be used as a first approximation of the efficiency of decapping and 5' to 3' exonucleolytic digestion (see HATFIELD et al. 1996 ). Given this, we examined the full-length to poly(G)3' end fragment ratios for the MFA2pG and PGK1pG transcripts in all the dcp1 mutants.

The results of these studies (summarized in Table 1) are as follows. First, 16 of the dcp1 primary mutations (covering 27 residues) did not have any significant effect on the ratio of poly(G)3' end fragment to full-length mRNA and, hence, were not pursued further. These alleles include dcp1-3, -5, -6, -10, to -16, -21, -24, -26 to -28 and -40. Second, the mutant dcp1-7 showed a large effect on the ratio. In this mutant, Arg70 and Asp71 were both converted to alanines (Figure 1). Subsequent studies indicated that the severity of the phenotype of this mutant could be almost fully attributed to the mutation of Arg70. This was revealed by the mutant dcp1-34, wherein only Arg70 is mutated to alanine. Third, 9 mutants, dcp1-2, -4, -17, -19, -25, -31, -32, -33, and -35, showed a partial change in the poly(G)3' end fragment to full-length mRNA ratio, suggesting that these alleles represented partial loss-of-function mutations. Importantly, it should be noted that the dcp1 mutations affected both MFA2pG and PGK1pG mRNAs similarly in each case, indicating that mRNA-specific rates of decapping cannot be attributed to the decapping enzyme per se (see DISCUSSION).

To confirm that the changes in poly(G)3' end fragment to full-length mRNA ratios observed in the mutant strains reflected a change in mRNA turnover rates, we directly measured the half-life of MFA2pG mRNA in the different mutants. For this purpose, we took advantage of the fact that the MFA2pG mRNA is expressed from the GAL1 UAS, whose transcription can be inhibited by the addition of glucose, thus allowing a simple determination of mRNA decay rates (see MATERIALS AND METHODS). The results of these analyses (Figure 2) demonstrated that there was a change in mRNA decay rate in the mutants with an altered ratio of full-length to poly(G)3' end fragment. As expected, mutants with a large change in the ratio (e.g., dcp1-34) showed a large change in t_1/2, and the partial loss-of-function mutants showed a more modest effect. Thus, these observations indicate that we have identified several mutations in the decapping enzyme that affect mRNA turnover in vivo.

The dcp1-2 allele is thermolabile for function:
To determine if any of the mutants were cold or heat sensitive, the ratios of poly(G)3' end fragment to full-length mRNA were also compared at 18° and 36°. Strikingly, the dcp1-2 mutant showed a severe defect at 36°, but a lesser defect at lower temperatures (Figure 3). This difference was also seen in decay rates where at 18°, both dcp1-2 and wild type showed the same rates of decay, while at 30° and 36°, there was a substantial difference (Figure 2 and Figure 4). In fact, at 36°, MFA2pG mRNA decay in dcp1-2 cells was as slow as in dcp1 cells grown at the same temperature (Figure 4). The dcp1-34 allele also showed small differences in function in response to the temperature of growth. This allele was essentially like a null allele at 30° and higher, but it produced a very small amount of the poly(G)3' end fragment at 18° (Figure 3). These thermosensitive alleles, especially dcp1-2, should be useful for the analysis of mRNA decay (e.g., see JACOBS ANDERSON and PARKER 1998 ).

In vitro analysis of the mutant decapping enzymes:
The above studies identified a number of mutations in Dcp1p that led to a defect in decapping in vivo. In principle, these defects could arise by several means, including destabilizing the protein, altering the enzymatic properties of the decapping enzyme, or preventing interactions with other factors required for proper decapping. To distinguish these possibilities, we first determined if the mutant proteins were being expressed at levels comparable to the wild-type enzyme. Western analysis of crude extracts made from the various mutant dcp1 strains performed with anti-Dcp1p antibodies showed that the level of the Dcp1p protein was not substantially different in the mutants compared to wild-type strain, with the exception of the dcp1-25 mutant, which showed reduced levels compared to wild type (data not shown). This observation indicated that the loss-of-function phenotypes of the mutants, with the exception of dcp1-25, were not caused by changes in protein stability and were, therefore, likely to be caused by the mutations' effects on either the enzymatic function of the decapping enzyme or its ability to interact with other proteins required for decapping in vivo.

To determine if the mutant Dcp1ps expressed in these alleles were defective in enzymatic activity, we purified them and assayed their ability to decap a capped mRNA in vitro. For this purpose, we expressed FLAG epitope-tagged versions of wild-type and mutant dcp1 genes from 2µ vectors (see MATERIALS AND METHODS) in dcp1 cells (strain yRP1071). FLAG-tagged proteins encoded by the following alleles were purified (see MATERIALS AND METHODS): wild type; five of the mutant alleles that lead to partial defects in mRNA decapping in vivo (dcp1-17, dcp1-4, dcp1-19, dcp1-25, and dcp1-31); the allele dcp1-7, which is completely defective in RNA decay in vivo; and the allele dcp1-2, which has a temperature-sensitive RNA decay phenotype in vivo. Analysis of the purified protein samples on SDS polyacrylamide gel revealed a clean ~30-kD protein (Figure 5) that was verified to be Dcp1p by Western analysis (Figure 6B). The purified enzymes were then assayed in vitro for decapping activity using in vitro-synthesized MFA2 mRNA labeled with ³²P in the cap. Upon decapping, this substrate releases radiolabeled m7 GDP, which is separated from it by TLC (see MATERIALS AND METHODS). These assays were done under limiting concentrations of substrate so that they would be sensitive to changes in the affinity of the enzyme for the substrate. Figure 6A shows time course curves of decapping reactions performed with wild-type and mutant FLAG-Dcp1p samples drawn by plotting decapping activity values that were normalized for the amount of Dcp1p protein present in the respective samples, as described in MATERIALS AND METHODS. Assays were also done after adjusting the protein concentrations of the individual preparations so that equal amounts of dcp1 protein were compared directly (Figure 6B). The results are summarized in Table 2.

Two classes of mutant proteins were identified. First, the alleles dcp1-2 and dcp1-7, which caused a severe loss of mRNA decay function in vivo, yielded FLAG fusion proteins that had in vitro-specific activities at least five times less than that of wild-type FLAG-Dcp1p (Table 2 and Figure 6). We interpreted this observation to indicate that the dcp1-2 and dcp1-7 mutations alter residues that are important for the enzymatic function of Dcp1p.

The second class consisted of six mutants that showed a partial loss of function in vivo, dcp1-4, dcp1-17, dcp1-19, dcp1-25, and dcp1-31. Unlike their in vivo phenotypes, the in vitro-specific activities of the decapping enzymes made by these mutants did not show any loss of function compared to the wild-type protein. As seen in Table 2 and Figure 6, all the mutant proteins had specific activities equal to or greater than that of wild-type Dcp1p. This result was confirmed in multiple preparations of the proteins. The higher specific activity (compared to that of wild-type protein) of some of the mutant proteins may result from the fact that when Dcp1p is overexpressed, only a fraction of the enzyme is active (LAGRANDEUR and PARKER 1998 ), and it is possible that these mutants alter the percentage of enzyme that is active, possibly as a result of the defect in the ability to decap mRNAs. Nevertheless, the clear and important result is that these proteins are active decapping enzymes in isolation, yet they show defects in decapping in vivo.

One possible explanation for the mutations that affect decapping in vivo but not in vitro is that they target residues required for interactions in vivo that are absent in the in vitro experiment. Alternatively, because these partial loss-of-function mutants have only a weak RNA decay phenotype (moderate stabilization of MFA2pG mRNA) in vivo, another explanation could be that the effects of these mutations on the in vitro enzymatic efficiency of the protein is too small to be detected. One way to distinguish between these two possibilities would be to combine the mutations borne by two or more of these partial loss-of-function mutants to make new mutants that show a more severe loss of function in vivo and to study the specific activity of the decapping enzyme made by such mutants. If the second possibility were true, then one would predict these proteins to have significantly less specific activity than the wild-type protein. On the other hand, if these mutants affect important in vivo interactions with other factors, then combination of the lesions should yield proteins with strong defects in vivo, but still active in vitro when the enzyme is purified.

To this end, mutations of dcp1-17 and dcp1-4 alleles were individually combined by restriction digestion and ligation with mutations of dcp1-19 or dcp1-25. This resulted in four new mutants, dcp1-41 (combination of dcp1-17 and dcp1-19 mutations), dcp1-42 (combination of dcp1-17 and dcp1-25 mutations), dcp1-43 (combination of dcp1-4 and dcp1-19 mutations), and dcp1-44 (combination of dcp1-4 and dcp1-25 mutations). The plasmids were then individually transformed into dcp1 strain to analyze their phenotypes.

MFA2pG mRNA half-life was measured in the mutant strains dcp1-41 to -44. As shown in Figure 7, the alleles dcp1-41 and -42 (combination of mutations of dcp1-17 with those of dcp1-19 and dcp1-25, respectively) do not cause any significantly stronger loss of function than the primary partial loss-of-function mutants and, hence, were not studied further. On the other hand, the other two alleles, dcp1-43 and dcp1-44 (combination of mutations of dcp1-4 with those of dcp1-19 and dcp1-25, respectively) clearly resulted in a much higher stabilization of MFA2pG mRNA than any of the primary partial loss-of-function mutants. In addition, Western analysis showed that all these mutant proteins were expressed at approximately the same levels as the wild-type protein (data not shown).

View larger version (47K): In this window In a new window Download PPT slide   Figure 7. Half-life of MFA2pG mRNA in mutants dcp1-41 to -44. Shown are MFA2pG mRNA half-life measurements in alleles dcp1-41 to -44 (grown at 30°) on Northern blots after transcriptional shutoff by shifting galactose-induced cells to glucose medium. The time points at which cells were collected for RNA isolation are indicated on top of the gel pictures. The name of the dcp1 allele and the MFA2pG half-life determined are shown on the right of each gel picture. Northern analysis using an oligonucleotide probe specific for MFA2pG mRNA and half-life determination was done as described in MATERIALS AND METHODS.

To find if the mutant Dcp1p made in the dcp1-43 and dcp1-44 mutants was active in vitro, we made FLAG epitope-tagged versions of the two mutant dcp1 genes and purified the FLAG-tagged enzymes encoded by them as described earlier. In vitro decapping assays performed with these mutant proteins revealed that both of these mutant proteins had specific activities similar to those of the wild-type protein (Table 2 and Figure 6). This strongly supports the idea that the amino acid residues changed in these mutants are not likely to be important for the enzymatic activity of Dcp1p, but, rather, are important for some functionally important interactions of Dcp1p with other proteins in vivo.

Dcp1p mutants partially defective in MFA2pG and PGK1pG mRNA decay are not defective in mRNA surveillance:
The decapping enzyme Dcp1p functions in both the normal deadenylation-dependent decay pathway and in the deadenylation-independent decapping triggered by early nonsense codons (BEELMAN et al. 1996 ). The experiments described above examined the effects of dcp1 mutations on the former process (because MFA2pG and PGK1pG mRNAs decay by this pathway). Therefore, we wanted to examine the effects of the dcp1 mutations on the deadenylation-independent decapping process. For this purpose, we took advantage of the observation that the CYH2 pre-mRNA is a poorly spliced RNA; some unspliced precursor enters the cytoplasm in wild-type cells (HE et al. 1993 ). Because the intronic sequence leads to an in-frame premature termination codon, this pre-mRNA forms a natural substrate for the deadenylation-independent decapping pathway, which requires Dcp1p (HE et al. 1993 ; BEELMAN et al. 1996 ). Thus, the accumulation of CYH2 premRNA in vivo can be used as a measure of how efficiently the deadenylation-independent decapping mechanism operates in the cells.

Examination of the extent of CYH2 pre-mRNA accumulation by Northern analysis in the various dcp1 mutants is shown in Figure 8. As expected, in wild-type cells, very little CYH2 pre-mRNA accumulated in relation to the CYH2 mRNA. In contrast, the amount of the CYH2 pre-mRNA accumulated in the dcp1 strain and in all the strong loss-of-function mutants, except for dcp1-43 (i.e., in dcp1-2, -7, -34, and -44), was substantially higher than in wild-type cells. Consistent with their MFA2pG mRNA decay phenotype in vivo, dcp1-7 and -34 showed higher CYH2 pre-mRNA accumulation than did dcp1-2 and -44. The allele dcp1-43 caused only a moderate increase in CYH2 premRNA accumulation compared to wild-type cells. Nevertheless, the dcp1 alleles that had a partial loss of function in deadenylation-dependent decay (i.e., MFA2pG mRNA decay) did not show any significant increase in the accumulation of CYH2 pre-mRNA compared to the wild-type control. It is important to note here that as seen in Figure 8, the levels of CYH2 pre-mRNA are very low in wild-type cells, and they are increased about fourfold in dcp1 cells, suggesting that ~75% of the pre-mRNA pool in wild-type cells decays in a DCP1-dependent manner.

View larger version (86K): In this window In a new window Download PPT slide   Figure 8. Effect of the different dcp1 mutations on the degradation of CYH2 pre-mRNA in vivo. RNA made from wild-type and mutant cells grown to midlog at 30° in minimal medium was subjected to Northern analysis using a random-primed CYH2 cDNA probe (which hybridizes to both CYH2 pre-mRNA and mRNA), as described in MATERIALS AND METHODS. (A) Phosphorimages of Northern blots. The dcp1 allele from which RNA was prepared is indicated above each lane, and positions of the CYH2 pre-mRNA and mRNA are shown on the left side of gel picture. (B) Ratio of the level of CYH2 pre-mRNA to that of CYH2 mRNA in different dcp1 alleles determined by quantitation of the corresponding bands in the phosphorimages shown in A.

The observation that the partial loss-of-function mutations did not affect deadenylation-independent decapping could be explained in two ways. First, these mutations could affect some specific structural feature(s) of Dcp1p that is required only for deadenylation-dependent decapping. This possibility seemed unlikely because similar results were seen with all the partial loss-of-function alleles. Alternatively, the rate of decapping of the nonsense-containing mRNAs (which undergo deadenylation-independent decapping) could be less sensitive (than normal mRNAs that undergo deadenylation-dependent decapping) to perturbations in the levels of decapping activity in vivo. To test the second possibility, we took advantage of the temperature sensitivity of the dcp1-2 allele. The logic was to determine if there would be a differential effect on nonsense-mediated vs. regular mRNA decay as we raised the temperature and, thereby, decreased the proportion of functional decapping enzyme in vivo.

For this purpose, we prepared RNA from the dcp1-2 mutant strain grown at 18°, 20°, 22°, 24°, 27°, 30°, 33°, and 36°, and from wild-type cells grown at 18°, 30°, and 36°, and we determined the effects on normal decapping (by examination of the full-length to fragment ratios for the MFA2pG transcript) and deadenylation-independent decapping (by examination of the CYH2 pre-mRNA accumulation relative to CYH2 mRNA accumulation). Comparison of these data showed that decay of the CYH2 pre-mRNA was less sensitive to decreases in the levels of decapping activity. The critical observation was that in dcp1-2 mutants, the decay of MFA2pG mRNA was significantly defective at temperatures where the CYH2 pre-mRNA decay remains normal (Figure 9A). For example, at 30°, there was a substantial increase in the amount of full-length MFA2pG transcript relative to the poly(G)3' end fragment, whereas there was little change in the amount of CYH2 pre-mRNA accumulation. This observation was consistent with the explanation that in vivo, the decapping activity is limiting for normal mRNA substrates but not for nonsense mRNAs, and, therefore, a partial decrease in decapping activity affects only normal mRNA decay function. This idea was further supported by studying the CYH2 pre-mRNA accumulation in the other mutant with temperature-sensitive MFA2pG mRNA decay phenotype, dcp1-34. In this mutant, the accumulation of CYH2 pre-mRNA is almost as low as in the wild-type control at 18° (Figure 9B), while there is a strong defect in the decapping of the MFA2pG transcript in vivo at the same temperature (as shown by very low levels of poly(G) fragment in relation to full-length MFA2pG mRNA in Figure 3). Importantly, Figure 9 shows that at 36°, the difference in CYH2 pre-mRNA/mRNA ratio between wild-type and dcp1-2 cells is approximately fourfold, which is comparable to the approximately fourfold difference observed in that ratio between wild-type and dcp1 cells grown at 30° (see Figure 8). Because dcp1-2 cells are almost as defective in decapping as dcp1 cells at 36° (see Figure 4), this indicates that the size of the CYH2 pre-mRNA pool degrading in a DCP1-dependent fashion is not drastically altered at higher temperatures. We have also observed a similar fold difference in this ratio when wild-type cells and dcp1 cells grown at 36° were compared (data not shown).

View larger version (33K): In this window In a new window Download PPT slide   Figure 9. Decay of nonsense mRNA substrate in vivo is not sensitive to modest losses of Dcp1p function. (A) Effect of growth temperature on the steady-state ratio of the levels of full-length (FL) species and poly(G)3' end fragment of MFA2pG mRNA and of pre-mRNA and mRNA of CYH2 in a dcp1-2 mutant in vivo. RNA was made from wild-type cells (filled circles) grown at 18°, 30°, and 36° and from dcp1-2 cells (open circles) grown at 18°, 20°, 22°, 24°, 27°, 30°, 33°, and 36° and was analyzed by Northern using an end-labeled oligonucleotide probe specific for MFA2pG. Blots were then stripped and reprobed with a random-primed CYH2 cDNA probe. Steady-state levels of full-length species and poly(G)3' end fragments of MFA2pG mRNA and pre-mRNA and mRNA of CYH2 were determined as described in MATERIALS AND METHODS, and ratios were calculated. (B) Accumulation of CYH2 pre-mRNA at steady state at 18°, 30°, and 36° in vivo in a dcp1-34 strain, which shows severe loss of function at 18° and complete loss of function at 30° and 36° for MFA2pG mRNA decay in vivo. Wild-type and dcp1-34 cells were grown at the temperatures indicated above the lanes to midlog in minimal medium, and RNA made from them was analyzed by Northern analysis using CYH2 cDNA probe (which hybridizes to both CYH2 pre-mRNA and mRNA) as described in MATERIALS AND METHODS.

As mentioned earlier (Figure 8), the strong loss-of-function allele, dcp1-43, caused only a slight increase in the accumulation of CYH2 pre-mRNA compared to the wild type, unlike all the other strong loss-of-function alleles, dcp1-7, -34, -2, and -44, which resulted in a substantial increase in accumulation of the CYH2 pre-mRNA. At the present time, we do not know if this is because the mutations borne by this allele (dcp1-43) affect features of the decapping enzyme that are specifically required for deadenylation-dependent decapping (normal mRNA decay). Further work will be needed to resolve this issue.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Identification of residues important for Dcp1p activity:
In this work, we have identified several residues that are important for the ability of the decapping enzyme to function. The strongest effects were seen with the mutation of arginine at position 70 (dcp1-34) and with the dcp1-2 mutant, in which Arg29 and Asp31 were both changed to alanine residues. The dcp1-34 mutant was defective in decapping in vivo at all temperatures tested, with only a very small amount of mRNA turnover observed at 18°. The dcp1-2 mutant was more strikingly temperature sensitive in its function and showed little difference from the wild type at a low temperature, yet it was phenotypically similar to a null mutant at 36°. In addition to their strong in vivo phenotypes, both dcp1-2 and dcp1-7 produced proteins that were defective in enzymatic activity when purified. This observation indicated that these residues were critical for the enzyme's function. Interestingly, both Arg70 and Asp31 are conserved among a family of related ORFs in the database with homology to the DCP1 sequence (S. MIAN and R. PARKER, unpublished data). One potential role for these conserved arginine and aspartic residues is to be involved in recognition of the substrate. This possibility is suggested by studies on the cytoplasmic cap-binding protein and the viral cap recognition protein, which show that Arg and Asp residues are involved in recognizing the cap moiety. For example, in both proteins, acidic amino acid residues form hydrogen bonds with the N1 and N2 positions of the 7-methylguanine through their side chain carboxyl oxygens (HODEL et al. 1997 ; MARCOTRIGIANO et al. 1997 ). Similarly, the - and ß-phosphate oxygens of m7GDP interact with an Arg and a Lys residue of eIF4E (MARCOTRIGIANO et al. 1997 ). Thus, in Dcp1p, Arg29, Asp31, and Arg70 could have analogous roles. Alternatively, these two residues could play an important but not absolutely required role in the catalytic mechanism of the decapping enzyme.

A second set of interesting residues are the aromatic residues, Trp56, Trp204, and Tyr47, each of which causes a partial loss of function when changed to alanine. These residues are also conserved among the family of related ORFs in the database with homology to DCP1. Furthermore, the importance of such aromatic residues is underscored by the fact that both the cap-binding protein eIF4E and the vaccinia virus cap-specific methyltransferase have been shown to interact with the cap structure by stacking the aromatic ring of 7-methyl guanine between the ring structures of their conserved aromatic amino acid residues, and that the methyl group specificity has at least partly been attributed to such a stacking interaction (HODEL et al. 1997 ; MARCOTRIGIANO et al. 1997 ; MATSUO et al. 1997 ). This raises the possibility that these aromatic residues in yeast Dcp1p may have similar functions.

In the case of Tyr47, its importance for the protein's function may arise either by virtue of its aromatic ring or its hydroxyl, as in RNase-A, where a conserved Tyr residue has been shown to be required for stabilizing the active site structure (EBERHARDT et al. 1996 ) by hydrogen bond interactions through its hydroxyl. Therefore, we tested the effect of removing only the hydroxyl moiety of the Tyr47 in the mutant dcp1-35, where Tyr47 is replaced with Phe. The fact that this mutation, like dcp1-32, also causes a partial loss of function is consistent with the idea of the hydroxyl of Tyr47 being important (Figure 2). This suggests that the Tyr47 hydroxyl is likely to be engaged in an important hydrogen bond interaction.

An interesting observation was that the Dcp1p purified from the partial loss-of-function mutants (dcp1-17, dcp1-4, dcp1-19, dcp1-25, and dcp1-31) had wild-type-specific activity in vitro. This was also true for the mutants dcp1-43 and dcp1-44, which contained combinations of alleles that led to a strong defect in decapping in vivo, yet produced a functional decapping enzyme in vitro. These observations indicated that these lesions alter a property of the decapping enzyme that is not assayable in the current in vitro system. One formal possibility is that these mutations alter the enzyme's structure/folding in such a manner that its catalytic efficiency is lost in vivo but not in vitro. This possibility seems unlikely, considering that this effect was seen with seven different mutants. An alternate and simpler explanation is that these lesions disrupt activation (or recruitment onto the substrate) of Dcp1p by some other factors that are required for efficient decapping in vivo rather than by affecting Dcp1p's own catalytic effeciency. Such activation may involve the direct interaction of such factors with Dcp1p. The importance of the requirement of other gene products for the functioning of Dcp1p in vivo has also been suggested by earlier work in which mutations in other genes that affect decapping in vivo without altering the levels of the decapping enzyme were identified (HATFIELD et al. 1996 ; BOECK et al. 1998 ). Moreover, the specific activity of Dcp1p was found to be significantly higher when it is a part of the crude lysate rather than in purified form (T. E. LAGRANDEUR and R. PARKER, unpublished observations).

Interestingly, among the four mutants that were made by combining mutations borne by different pairs of primary partial loss-of-function mutants, only two (dcp1-43 and dcp1-44) exhibited a clearly more severe (than the primary partial loss-of-function mutants) loss-of-function phenotype in vivo, while the other two (dcp1-41 and dcp1-42) were defective roughly to the same extent as the primary partial loss-of-function mutants. In other words, only two combinations (out of four) of the primary mutations resulted in aggravating the loss-of-function phenotype of the primary mutations while the other two did not. This could happen if the interaction of Dcp1p with other factor(s) involves several sites of contact on the Dcp1p molecule and some of these contacts are redundant with each other. In that case, in any given primary partial loss-of-function mutant—where a given site(s) is already mutated—introducing additional mutations in sites that are redundant (for interaction) with that will fail to aggravate the phenotype any further (for discussion of a similar situation see HOLTZMAN et al. 1994 ). Further work will be required to determine the proteins that interact with Dcp1p and if those interactions are indeed affected by these lesions.

The effects of Dcp1p mutations on differential mRNA decapping rates:
An important question is how the different rates of mRNA decapping are specified on individual mRNAs. In principle, there could be specific interactions with the decapping enzyme that recruit the enzyme to individual mRNAs at different rates. From this perspective, it would be expected that specific alleles of the Dcp1p would affect mRNAs differentially. In this light, we initially compared the effects of the mutations on the decay of the PGK1, a stable mRNA, and MFA2, an unstable mRNA. Both of these mRNAs degrade by the deadenylation-dependent decapping pathway, and the difference in their decay rates at least partly results from the difference in their decapping rates (DECKER et al. 1993; MUHLRAD et al. 1994 , MUHLRAD et al. 1995 ). The fact that none of our mutations show a differential effect on the decay of these mRNAs suggests that DCP1 per se is unlikely to distinguish between these two transcripts. This observation is also consistent with more recent work arguing that the status of the translation initiation complex per se plays an important role in modulating decapping rates (LAGRANDEUR and PARKER 1999 ; D. SCHWARTZ and R. PARKER, unpublished results; D. MUHLRAD and R. PARKER, unpublished results).

Our experiments show that moderate losses of the decapping enzyme's function specifically affected only deadenylation-dependent decapping and did not have any detectable effect on nonsense-mediated decay in vivo. This conclusion was supported by the following observations, in which conditions that led to a partial defect in normal (MFA2pG) mRNA decay resulted in no defect in nonsense-mediated mRNA decay. First, in all the partial loss-of-function mutants studied, there was no increase in the accumulation of CYH2 pre-mRNA compared to wild type (Figure 8). Second, with increasing growth temperature of dcp1-2 cells, MFA2pG mRNA decay becomes defective at a lower temperature than does the CYH2 premRNA decay (Figure 9A). Third, dcp1-34 cells showed a strong defect in MFA2pG mRNA decay at 18° (Figure 3), but they showed no effect on CYH2 premRNA accumulation at this temperature (Figure 9B).

The above observations argue that nonsense-mediated decapping is less sensitive to perturbations in the decapping enzyme's function than normal deadenylation-dependent decapping. This suggests that the nonsense codon-containing substrate may be a more easily accessible substrate for decapping than the normal mRNA substrate because of the manner in which decapping is triggered in response to a nonsense codon. This view has two implications. First, other trans-acting mutations that have been described as being specific for deadenylation-dependent decapping (HATFIELD et al. 1996 ; BOECK et al. 1998 ) may in fact simply lead to partial loss-of-function phenotypes for decapping activity, with the ensuing effect only on normal mRNAs. Thus, the MRT1, MRT3, and SPB8 gene products may not truly be specific for deadenylation-dependent decapping per se. The second interesting implication is that by limiting decapping, either in cis or in trans, a situation could be created where a nonsense-mediated decay will still occur, but deadenylated mRNAs will be stable. Strikingly, this situation describes the state in Xenopus oocytes wherein normal mRNAs are stable as deadenylated species, yet nonsense-containing mRNAs are still rapidly degraded (WHITFIELD et al. 1994 ). This implies that modulation of the levels of the decapping activity, either in cis or in trans, may be important in early development and in other biological situations.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

BASHKIROV, V. I., H. SCHERTHAN, J. A. SOLINGER, J.-M. BUERSTEDDE, and W.-D. HEYER, 1997  A mouse cytoplasmic exoribonuclease (mXRN1p) with preference for G4 tetraplex substrates. J. Cell Biol. 136:761-773[Abstract/Free Full Text].

BEELMAN, C. A. and R. PARKER, 1995  Degradation of mRNA in eukaryotes. Cell 81:179-183[Medline].

BEELMAN, C. A., A. STEVENS, G. CAPONIGRO, T. E. LAGRANDEUR, and L. HATFIELD et al., 1996  An essential component of the decapping enzyme required for normal rates of mRNA turnover. Nature 382:642-646[Medline].

BOECK, R., B. LAPEYRE, C. E. BROWN, and A. B. SACHS, 1998  Capped mRNA degradation intermediates accumulate in the yeast spb8-2 mutant. Mol. Cell. Biol. 18:5062-5072[Abstract/Free Full Text].

CAPONIGRO, G. and R. PARKER, 1995  Multiple functions for the poly(A)-binding protein in mRNA decapping and deadenylation in yeast. Genes Dev. 9:2421-2432[Abstract/Free Full Text].

CAPONIGRO, G., D. MUHLRAD, and R. PARKER, 1993  A small segment of MAT1 transcript promotes mRNA decay in Saccharomyces cerevisiae: a stimulatory role for rare codons. Mol. Cell. Biol. 13:5141-5148[Abstract/Free Full Text].

COUTTET, P., M. FROMONT-RACINE, D. STEEL, R. PICTET, and T. GRANGE, 1997  Messenger RNA deadenylation precedes decapping in mammalian cells. Proc. Natl. Acad. Sci. USA 94:5628-5633[Abstract/Free Full Text].

DECKER, C. J. and R. PARKER, 1993  A turnover pathway for both stable and unstable mRNAs in yeast: evidence for a requirement for deadenylation. Genes Dev. 7:1632-1643[Abstract/Free Full Text].

EBERHARDT, E. S., P. K. WITTMAYER, B. M. TEMPLER, and R. T. RAINES, 1996  Contribution of a tyrosine side chain to ribonuclease A catalysis and stability. Prot. Sci. 5:1697-1703[Abstract].

ELLEDGE, S. J. and R. W. DAVIS, 1988  A family of versatile centromeric vectors designed for use in the sectoring-shuffle mutagenesis assay in Saccharomyces cerevisiae. Gene 70:303-312[Medline].

GERA, J. F. and E. J. BAKER, 1998  Deadenylation-dependent and -independent decay pathways for alpha1-tubulin mRNA in Chlamydomonas reinhardtii.. Mol. Cell. Biol. 18:1498-1505[Abstract/Free Full Text].

HARRIS, M. E., J. M. NOLAN, A. MALHOTRA, J. W. BROWN, and S. C. HARVEY et al., 1994  Use of photoaffinity crosslinking and molecular modeling to analyze the global architecture of ribonuclease P RNA. EMBO J. 13:3953-3963[Medline].

HATFIELD, L., C. A. BEELMAN, A. STEVENS, and R. PARKER, 1996  Mutations in trans-acting factors affecting mRNA decapping in Saccharomyces cerevisiae.. Mol. Cell. Biol. 16:5830-5838[Abstract].

HE, F., S. W. PELTZ, J. L. DONAHUE, M. ROSBASH, and A. JACOBSON, 1993  Stabilization and ribosome association of unspliced pre-mRNAs in a upf1 mutant. Proc. Natl. Acad. Sci. USA 90:7034-7038[Abstract/Free Full Text].

HEYER, W. D., A. W. JOHNSON, U. REINHART, and R. D. KOLODNER, 1995  Regulation and intracellular localization of Saccharomyces cerevisiae strand exchange protein 1 (Sep1/Xrn1/Kem1), a multifunctional exonuclease. Mol. Biol. Cell. 15:2728-2736.

HIGGS, D. C. and J. T. COLBERT, 1994  Oat phytochrome A mRNA degradation appears to occur via two distinct pathways. Plant Cell 6:1007-1019[Abstract].

HODEL, A., X. SHI, and P. D. GERSHON, 1997  Specific protein recognition of an mRNA cap through its alkylated base. Nat. Struct. Biol. 4:350-354[Medline].

HOLTZMAN, D. A., K. F. WERTMAN, and D. G. DRUBIN, 1994  Mapping actin surfaces required for functional interactions in vivo.. J. Cell Biol. 126:423-432[Abstract/Free Full Text].

HSU, C. L. and A. STEVENS, 1993  Yeast cells lacking 5'3' exoribonuclease 1 contain mRNA species that are poly (A) deficient and partially lack the 5' cap structure. Mol. Cell. Biol. 13:4826-4835[Abstract/Free Full Text].

JACOBS ANDERSON, J. S. and R. PARKER, 1998  The 3' to 5' degradation of yeast mRNAs is a general mechanism for mRNA turnover that requires the SKI2 DEVH box protein and 3' to 5' exonucleases of the exosome complex. EMBO J. 17:1497-1506[Medline].

JACOBSON, A. and S. W. PELTZ, 1996  Interrelationships of the pathways of mRNA decay and translation in eukaryotic cells. Annu. Rev. Biochem. 65:693-739[Medline].

LAEMMLI, U. K., 1970  Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685[Medline].

LAGRANDEUR, T. E. and R. PARKER, 1998  Isolation and characterization of Dcp1p, the yeast mRNA decapping enzyme. EMBO J. 17:101-110[Medline].

LAGRANDEUR, T. E. and R. PARKER, 1999  The cis-acting sequences responsible for the differential decay of the unstable MFA2 and stable PGK1 transcripts in yeast include the context of translational start codon. RNA in press.

LIM, S. K. and L. E. MAQUAT, 1992  Human ß-globin mRNAs that harbor a nonsense codon are degraded in murine erythroid tissues to intermediates lacking regions of exon I or exons I and II that have a cap-like structure at the 5' termini. EMBO J. 11:3271-3278[Medline].

MARCOTRIGIANO, J., A.-C. GINGRAS, N. SONENBERG, and S. K. BURLEY, 1997  Cocrystal structure of the messenger RNA 5' cap-binding protein (eIF-4E) bound to 7-methyl GDP. Cell 89:951-961[Medline].

MATSUO, H., H. LI, A. M. MCGUIRE, C. M. FLETCHER, and A. C. GINGRAS et al., 1997  Structure of translation factor eIF4E bound to m7GDP and interaction with 4E-binding protein. Nat. Struct. Biol. 4:717-724[Medline].

MUHLRAD, D. and R. PARKER, 1992  Mutations affecting stability and deadenylation of the yeast MFA2 transcript. Genes Dev. 6:2100-2111[Abstract/Free Full Text].

MUHLRAD, D. and R. PARKER, 1994  Premature translational termination triggers mRNA decapping. Nature 370:578-581[Medline].

MUHLRAD, D., C. J. DECKER, and R. PARKER, 1994  Deadenylation of the unstable mRNA encoded by the yeast MFA2 gene leads to decapping followed by 5' to 3' digestion of the transcript. Genes Dev. 8:855-866[Abstract/Free Full Text].

MUHLRAD, D., C. J. DECKER, and R. PARKER, 1995  Turnover mechanisms of the stable yeast PGK1 mRNA. Mol. Cell. Biol. 15:2145-2156