The poly(A) binding protein is required for poly(A) shortening and 60S ribosomal subunit-dependent translation initiation

Identification of ribosomal protein S1 as a poly(A) binding protein in Escherichia coli

To elucidate the metabolic function of mRNA polyadenylation in Escherichia coli. we searched for a polyadenylate-binding protein as a potential mediator of the function of the poly(A) moiety. Using a nitrocellulose filter-binding assay and a Northwestern blot technique, a protein in the ribosomal supernatant fraction of E coli was identified and purified to homogeneity. N-terminal sequence analysis yielded a 25-residue sequence which corresponded to the 25 N-terminal amino acids of protein S1, one of the proteins of the E coli 30S ribosomal subunit. Poly(A) binding to S1 protein was inhibited by Mg2+ and Mn2+ and by ATP and stimulated 8-fold by 100 mM KCl. The binding of S1 to poly(A) occurred with an association constant of 3 x 10(6) M-1 and seemed to be only mildly cooperative. Competition studies of the binding of poly(A) and poly(C) to purified S1 protein were consistent with the presence of two polynucleotide binding sites, of which one binds poly(A) five times more strongly than poly(C), whereas the other binds poly(C) 50 times more strongly than poly(A). Poly(A) bound to 30S ribosomal subunits but not to 50S ribosomes. To study possible association of S1 with the poly(A) tracts of E coli mRNA in the process of translation, poly(A) RNA was isolated from polysomes by oligo(dT) cellulose chromatography and the poly(A) RNA with bound protein was eluted either directly or after digestion with RNase T1 and A. When subjected to Western blot analysis with antibody to S1, both poly(A) RNA and isolated poly(A) tracts revealed bound S1 protein. The implications of these results for the possible interaction of poly(A) tracts of mRNA and the translational machinery of E coli are discussed.

Qsr1p, a 60S ribosomal subunit protein, is required for joining of 40S and 60S subunits

QSR1 is a recently discovered, essential Saccharomyces cerevisiae gene, which encodes a 60S ribosomal subunit protein. Thirty-one unique temperature-sensitive alleles of QSR1 were generated by regional codon randomization within a conserved 20-amino-acid sequence of the QSR1-encoded protein. The temperature-sensitive mutants arrest as viable, large, unbudded cells 24 to 48 h after a shift to 37 degrees C. Polysome and ribosomal subunit analysis by velocity gradient centrifugation of lysates from temperature-sensitive qsr1 mutants and from cells in which Qsr1p was depleted by down regulation of an inducible promoter revealed the presence of half-mer polysomes and a large pool of free 60S subunits that lack Qsr1p. In vitro subunit-joining assays and analysis of a mutant conditional for the synthesis of Qsr1p demonstrate that 60S subunits devoid of Qsr1p are unable to join with 40S subunits whereas 60S subunits that contain either wild-type or mutant forms of the protein are capable of subunit joining. The defective 60S subunits result from a reduced association of mutant Qsr1p with 60S subunits. These results indicate that Qsr1p is required for ribosomal subunit joining.

Translation initiation factor eIF4G mediates in vitro poly(A) tail-dependent translation

The yeast translation factor eIF4G associates with both the cap-binding protein eIF4E and the poly(A)-binding protein Pab1p. Here we report that the two yeast eIF4G homologs, Tif4631p and Tif4632p, share a conserved Pab1p-binding site. This site is required for Pab1p and poly(A) tails to stimulate the in vitro translation of uncapped polyadenylylated mRNA, and the region encompassing it is required for the cap and the poly(A) tail to synergistically stimulate translation. This region on Tif4631p becomes essential for cell growth when the eIF4E binding site on Tif4631p is mutated. Pab1p mutations also show synthetic lethal interactions with eIF4E mutations. These data suggest that eIF4G mediates poly(A) tail stimulated translation in vitro, and that Pab1p and the domain encompassing the Pab1p-binding site on eIF4G can compensate for partial loss of eIF4E function in vivo.

Involvement of a 50-kDa mRNP protein from Saccharomyces cerevisiae in mRNA binding to ribosomes

A yeast 50-kDa mRNA-binding protein (50mRNP) is found selectively associated with the 48S and 80S initiation complexes. This protein is structurally related to the translational elongation factor EF-1alpha. The protein reacts with antibodies directed against EF-1alpha and, similarly, EF-1alpha recognizes antibodies against the 50mRNP protein. This is evidence that they share at least one epitope which allows a similar antigenic behavior. In addition, both proteins show similar cleavage patterns upon treatment with the endoproteinase Lys-C. A murine antibody raised against 50mRNP inhibits both 48S and 80S initiation complex formation. The inhibitory effect is relieved by preincubating anti-50mRNP with EF-1alpha. Antibody to EF-1alpha manifests a similar inhibitory pattern for the formation of 48S and 80S complexes. These data strongly suggest that 50mRNP is an EF-1alpha-like polypeptide essential for the formation of the above complexes.

The major yeast poly(A)-binding protein is associated with cleavage factor IA and functions in premessenger RNA 3'-end formation

Polyadenylation of premessenger RNAs occurs posttranscriptionally in the nucleus of eukaryotic cells by cleavage of the precursor and polymerization of adenosine residues. In the yeast Saccharomyces cerevisiae, the mature poly(A) tail ranges from 60 to 70 nucleotides. 3'-end processing can be reproduced in vitro with purified factors. The cleavage reaction requires cleavage factors I and II (CF I and CF II), whereas polyadenylation involves CF I, polyadenylation factor I (PFI), and poly(A) polymerase (Pap1p). CF I has recently been separated into two factors, CF IA and CF IB. We have independently purified CF IA and found that five polypeptides cofractionate with the activity. They include Rna14p, Rna15p, Pcf11p, a new protein called Clp1p, and remarkably, the major poly(A)-binding protein Pab1p. Extracts from strains where the PAB1 gene is mutated or deleted are active for cleavage but generate transcripts bearing abnormally long poly(A) tracts. Complementation with recombinant Pab1p not only restores the length of the poly(A) tails to normal, but also triggers a poly(A) shortening activity. In addition, a monoclonal Pab1p antibody prevents the formation of poly(A) tails in extracts or in a reconstituted system. Our data support the notion that Pab1p is involved in the length control of the poly(A) tails of yeast mRNAs and define a new essential function for Pab1p in the formation of mature mRNAs.

Characterization of translation initiation factor 5 (eIF5) from Saccharomyces cerevisiae. Functional homology with mammalian eIF5 and the effect of depletion of eIF5 on protein synthesis in vivo and in vitro

Eukaryotic translation initiation factor 5 (eIF5) interacts in vitro with the 40 S initiation complex (40 S.AUG.Met-tRNAf.eIF2.GTP) to mediate the hydrolysis of ribosome-bound GTP. In Saccharomyces cerevisiae, eIF5 is encoded by a single copy essential gene, TIF5, that encodes a protein of 45,346 daltons. To understand the function of eIF5 in vivo, we constructed a conditional mutant yeast strain in which a functional but a rapidly degradable form of eIF5 fusion protein was synthesized from the repressible GAL promoter. Depletion of eIF5 from this mutant yeast strain resulted in inhibition of both cell growth and the rate of in vivo protein synthesis. Analysis of the polysome profiles of eIF5-depleted cells showed greatly diminished polysomes with simultaneous increase in free ribosomes. Furthermore, lysates of cells depleted of eIF5 were dependent on exogenously added yeast eIF5 for efficient translation of mRNAs in vitro. This is the first demonstration that the TIF5 gene encodes a protein involved in initiation of translation in eukaryotic cells. Additionally, we show that rat eIF5 can functionally substitute yeast eIF5 in translation of mRNAs in vitro as well as in complementing in vivo a genetic disruption in the chromosomal copy of TIF5.

Yeast Pab1 interacts with Rna15 and participates in the control of the poly(A) tail length in vitro

In Saccharomyces cerevisiae, the single poly(A) binding protein, Pab1, is the major ribonucleoprotein associated with the poly(A) tails of mRNAs in both the nucleus and the cytoplasm. We found that Pab1 interacts with Rna15 in two-hybrid assays and in coimmunoprecipitation experiments. Overexpression of PAB1 partially but specifically suppressed the rna15-2 mutation in vivo. RNA15 codes for a component of the cleavage and polyadenylation factor CF I, one of the four factors needed for pre-mRNA 3'-end processing. We show that Pab1 and CF I copurify in anion-exchange chromatography. These data suggest that Pab1 is physically associated with CF I. Extracts from a thermosensitive pab1 mutant and from a wild-type strain immunoneutralized for Pab1 showed normal cleavage activity but a large increase in poly(A) tail length. A normal tail length was restored by adding recombinant Pab1 to the mutant extract. The longer poly(A) tails were not due to an inhibition of exonuclease activities. Pab1 has previously been implicated in the regulation of translation initiation and in cytoplasmic mRNA stability. Our data indicate that Pab1 is also a part of the 3'-end RNA-processing complex and thus participates in the control of the poly(A) tail lengths during the polyadenylation reaction.

Visualization of poly(A)-binding protein complex formation with poly(A) RNA using atomic force microscopy

Poly(A)-binding protein (PABP) is an RNA-binding protein that binds specifically to the poly(A) tail of messenger RNAs in eukaryotes. The PABP/poly(A) tail complex has been implicated as being important in promoting the efficient initiation of translation as well as in maintaining the integrity of the mRNA. PABP binds poly(A) cooperatively with a packing density of one PABP molecule per 25 adenosine residues. We have investigated the complexes formed between purified PABP and poly(A) RNA using atomic force microscopy (AFM). PABP alone was observed to be primarily in a monomer form with a height of 1.0 +/- 0.2 nm. Following binding to poly(A), PABP appeared to be present in variable size complexes that bound lengthwise along the RNA. This size of the PABP/poly(A) complex appeared to be maximal, suggesting that PABP binding to poly(A) may be self-limiting. Poly(A) RNA alone appeared to contain a knob-like structure that largely disappeared once PABP was bound. The use of AFM has therefore provided potential new insights into the complexes formed by this RNA-binding protein.

Ssd1p of Saccharomyces cerevisiae associates with RNA

The SSD1 gene has been isolated as a single copy suppressor of many mutants, such as sit4, slk1/bck1, pde2, and rpc31, in the yeast Saccharomyces cerevisiae. Ssd1p has domains showing weak but significant homology with RNase II-related proteins, Cyt4p, Dss1p, VacB, and RNase II, which are involved in the modification of RNA. We found that Ssd1p had the ability to bind RNA, preferably poly(rA), as well as single-stranded DNA. Interestingly, the most conserved domain among the RNase II-related proteins was not necessary for interaction with RNA. Indirect immunofluorescence staining with anti-Ssd1p antibody revealed that Ssd1p was detected mainly in the cytoplasm. Furthermore, sucrose gradient sedimentation analysis demonstrated that Ssd1p was not cofractionated with polyribosomes, suggesting that Ssd1p is not particularly bound to a translationally active subpopulation of mRNA in the cytoplasm.

A comparison of mammalian and yeast pre-mRNA 3'-end processing

Many components of the mammalian and yeast pre-mRNA 3'-end-processing machinery have recently been purified and cDNAs or genes coding for these factors have been cloned. Most of the factors consist of multiple subunits, some of which serve to bind the RNA substrate, others of which are involved in forming a complex network of protein-protein interactions. Most of the mammalian 3'-end-processing factors are similar in their amino acid sequence to the yeast factors, indicating that they have a common evolutionary history.

Differential effects of aromatic and charged residue substitutions in the RNA binding domains of the yeast poly(A)-binding protein

The yeast poly(A)-binding protein (Pab1p) contains four RNA recognition motifs (RRMs). Site-directed mutations were introduced into each of these RRMs in order to investigate their relative contributions to specific and non-specific RNA binding, and to determine the consequences of these mutations on the ability of Pab1p to support viability. Specifically, a charged and an aromatic residue that were predicted to be involved in RNA binding were mutated in each RRM. These mutations revealed that the second RRM is primarily responsible for poly(A) binding, while the fourth RRM is primarily responsible for non-specific polypyrimidine RNA binding. The mutated aromatic residues in each RRM contributed to both modes of binding whereas the mutated charged residues contributed primarily to non-specific RNA binding. RNA binding in vivo correlated with the in vitro binding measurements. Furthermore, RNA binding, but not high-affinity poly(A) binding, correlated with the ability of Pab1p to sustain yeast cell viability. These data suggest that a single aromatic substitution in Pab1p can significantly reduce its RNA binding ability, that the capacity of Pab1p to bind poly(A) as well as other RNAs is mediated by distinct residues within different RRMs, and that Pab1p does not require high affinity poly(A) tail binding to perform its essential function.

Poly(A) tail shortening by a mammalian poly(A)-specific 3'-exoribonuclease

3'-Exonucleolytic removal of the poly(A) tail is the first and often rate-limiting step in the decay of many eucaryotic mRNAs. In a cytoplasmic extract from HeLa cells, the poly(A) tail of mRNA was degraded from the 3'-end. In agreement with earlier in vivo observations, prominent decay intermediates differed in length by about 30 nucleotides. The Mg2+-dependent, poly(A)-specific 3'-exoribonuclease responsible for this poly(A) shortening activity was purified from calf thymus. A polypeptide of 74 kDa copurified with the activity. The deadenylating nuclease (DAN) required a free 3'-OH group, released solely 5'-AMP, degraded RNA in a distributive fashion, and preferred poly(A) as a substrate. At low salt concentration, the activity of purified DAN was strongly dependent on spermidine or other, yet unidentified factors. Under these reaction conditions, DAN was also stimulated by the cytoplasmic poly(A)-binding protein I (PAB I). At physiological salt concentration, the stimulatory effect of spermidine was weak and PAB I was inhibitory. At either salt concentration DAN and PAB I reconstituted poly(A) shortening with the same pattern of intermediates seen in cytoplasmic extract. The properties of DAN suggest that the enzyme might be involved in the deadenylation of mRNA in vivo.

Life and death in the cytoplasm: messages from the 3' end

The cytoplasmic life of an mRNA revolves around the regulation of its localization, translation and stability. Interactions between the two ends of the mRNA may integrate translation and mRNA turnover. Regulatory elements in the region between the termination codon and poly(A) tail - the 3' untranslated region - have been identified in a wide variety of systems, as have been some of the key players with which these elements interact.

Identification of a competitive translation determinant in the 3' untranslated region of alfalfa mosaic virus coat protein mRNA

We report that the competitive translational activity of alfalfa mosaic virus coat protein mRNA (CP RNA), a nonadenylated mRNA, is determined in part by the 3' untranslated region (UTR). Competitive translation was characterized both in vitro, with cotranslation assays, and in vivo, with microinjected Xenopus laevis oocytes. In wheat germ extracts, coat protein synthesis was constant when a fixed amount of full-length CP RNA was cotranslated with increasing concentrations of competitor globin mRNA. However, translation of CP RNA lacking the 3' UTR decreased significantly under competitive conditions. RNA stabilities were equivalent. In X. laevis oocytes, which are translationally saturated and are an inherently competitive translational environment, full-length CP RNA assembled into large polysomes and coat protein synthesis was readily detectable. Alternatively, CP RNA lacking the 3' UTR sedimented as small polysomes, and little coat protein was detected. Again, RNA stabilities were equivalent. Site-directed mutagenesis was used to localize RNA sequences or structures required for competitive translation. Since the CP RNA 3' UTR has an unusually large number of AUG nucleotide triplets, two AUG-containing sites were altered in full-length RNA prior to oocyte injections. Nucleotide substitutions at the sequence GAUG, 20 nucleotides downstream of the coat protein termination codon, specifically reduced full-length CP RNA translation, while similar substitutions at the next AUG triplet had little effect on translation. The competitive influence of the 3' UTR could be explained by RNA-protein interactions that affect translation initiation or by ribosome reinitiation at downstream AUG codons, which would increase the number of ribosomes committed to coat protein synthesis.

alpha-thalassemia caused by a 16 bp deletion in the 3' untranslated region of the alpha 2-globin gene including the first nucleotide of the poly A signal sequence

We have identified a 16 bp deletion in the 3' untranslated region of the alpha 2-globin gene, including the first nucleotide of the polyadenylation signal sequence. The propositus, her mother and one of her brothers, all heterozygotes for the above deletion, have mild microcytic anemia. This is the first description of a deletion in the alpha gene involving both the 3' untranslated region and the polyadenylation signal sequence, causing alpha-thalassemia.

Ribosome concentration contributes to discrimination against poly(A)- mRNA during translation initiation in Saccharomyces cerevisiae

Inactivation of Saccharomyces cerevisiae poly(A) polymerase in a strain bearing the temperature-sensitive lethal pap1-1 mutation results in the synthesis of poly(A)- mRNAs that initiate translation with surprising efficiency. Translation of poly(A)- mRNAs after polyadenylation shut-off might result from an increase in the ratio of ribosomes and associated translation factors to mRNA, caused by the inability of poly(A)- mRNAs to accumulate to normal levels. To test this hypothesis, we used ribosomal subunit protein gene mutations to decrease either 40 or 60 S ribosomal subunit concentrations in strains carrying the pap1-1 mutation. Polyadenylation shut-off in such cells results in a nearly normal ratio of ribosomes to mRNA as revealed by polyribosome sedimentation analysis. Ribonuclease protection and Northern blot analyses showed that a significant percentage of poly(A)-deficient and poly(A)- mRNA associate with smaller polyribosomes compared with cells with normal ribosome levels. Analysis of the ratio of poly(A)-deficient and poly(A)- forms of a specific mRNA showed relatively more poly(A)- mRNA sedimenting with 20-60 S complexes than do poly(A)+ forms, suggesting a block in an early step of the translation initiation of the poly(A)- transcripts. These findings support models featuring the poly(A) tail as an enhancer of translation and suggest that the full effect of a poly(A) tail on the initiation strength of a mRNA may require competition for a limited number of free ribosomes or translation factors.

Preferential binding of poly(A)-binding protein 1 to an inhibitory RNA element in the human immunodeficiency virus type 1 gag mRNA

Human immunodeficiency virus type 1 (HIV-1) mRNAs encoding structural proteins contain multiple inhibitory/instability elements (INS), which decrease the efficiency of viral protein expression. We have previously identified a strong INS element (INS-1) within the p17(gag) coding region. Here we show that poly(A)-binding protein 1 (PABP1) binds preferentially to INS-1 within the p17(gag) mRNA, but not to a mutated mRNA in which INS-1 function is eliminated. Competition experiments performed in the presence of different nucleic acids and homoribopolymers demonstrated preferential binding of PABP1 to the INS-1-containing RNA. In contrast to HeLa cells and several lymphoid cell lines, certain human glioma cell lines exhibit high levels of gag expression in the absence of Rev upon transient transfection with wild type gag expression vectors. We analyzed extracts of different cell lines and found that the binding of PABP1 to INS-1 RNA is significantly diminished in glial cell extracts. The expression levels of gag correlate with the absence of binding of PABP1 to the INS-1 RNA in cellular extracts. These results suggest a role for PABP1 in the inhibition of gag expression mediated through INS-1.

The wheat poly(A)-binding protein functionally complements pab1 in yeast

Poly(A)-binding protein (PAB) binds to the poly(A) tail of most eukaryotic mRNAs and influences its translational efficiency as well as its stability. Although the primary structure of PAB is well conserved in eukaryotes, its functional conservation across species has not been extensively investigated. In order to determine whether PAB from a monocot plant species could function in yeast, a protein characterized as having PAB activity was purified from wheat and a cDNA encoding for PAB was isolated from a wheat seedling expression library. Wheat PAB (72 kDa as estimated by SDS/PAGE and a theoretical mass of 70 823 Da as determined from the cDNA) was present in multiple isoforms and exhibited binding characteristics similar to that determined for yeast PAB. Comparison of the wheat PAB protein sequence with PABs from yeast and other species revealed that wheat PAB contained the characteristic features of all PABs, including four RNA binding domains each of which contained the conserved RNP1 and RNP2 sequence motifs. The wheat PAB cDNA functionally complemented a pab1 mutant in yeast suggesting that, although the amino acid sequence of wheat PAB is only 47% conserved from that of yeast PAB, this monocot protein can function in yeast.

Identification and structure of activated-platelet protein-1, a protein with RNA-binding domain motifs that is expressed by activated platelets

Beyond their critical role in thrombosis, platelets perform important functions in vascular remodeling, inflammation, and wound repair. Many of these functions are executed by molecules expressed by activated platelets. A novel molecule, activated-platelet protein-1 (APP-1), was identified by a monoclonal antibody against activated rabbit platelets. When platelets were stimulated by thrombin, A23187 or ADP, APP-I was expressed on the platelet surface. APP-1 was also detected in whole cell lysates of platelets, but not on the external surfaces of resting platelets. With maximal activation by thrombin, 15 900 +/- 2800 molecules APP-1 were expressed/platelet. A 2.3-kb cDNA fragment containing a partial coding sequence for APP-1 was isolated from a rabbit bone marrow library by expression cloning with the anti-APP-1 monoclonal antibody. When expressed as a recombinant fusion protein in bacteria, APP-1 bound specifically to poly(A)-Sepharose. The full-length cDNA coding for human APP-1, obtained by DNA hybridization techniques, showed 98.7% amino acid sequence identity with the rabbit protein. Northern analysis with human APP-1 identified a 3.7-kb mRNA transcript in megakaryocytic lines that express transcripts for platelet proteins. Human APP-1 has four ribonucleotide binding domains with ribonucleoprotein 1 and 2 motifs. By virtue of its ribonucleotide binding domains, APP-1 is structurally related to polyadenylate-binding protein, which regulates translation initiation and polyadenylate shortening, and to nucleolysin, a specific effector molecule found in the granules of cytotoxic T lymphocytes.

The phosphorylation state of translation initiation factors is regulated developmentally and following heat shock in wheat

Several translation initiation factors in mammals and yeast are regulated by phosphorylation. The phosphorylation state of these factors is subject to alteration during development, environmental stress (heat shock, starvation, or heme deprivation), or viral infection. The phosphorylation state and the effect of changes in phosphorylation of the translation initiation factors of higher plants have not been previously investigated. We have determined the isoelectric states for the wheat translation initiation factors eIF-4A, eIF-4B, eIF-4F, eIF-iso4F, and eIF-2 and the poly(A)-binding protein in the seed, during germination, and following heat shock of wheat seedlings using two-dimensional gel electrophoresis and Western analysis. We found that the developmentally induced changes in isoelectric state observed during germination or the stress-induced changes were consistent with changes in phosphorylation. Treatment of the phosphorylated forms of the factors with phosphatases confirmed that the nature of the modification was due to phosphorylation. The isoelectric states of eIF-4B, eIF-4F (eIF-4E, p26), eIF-iso4F (eIF-iso4E, p28), and eIF-2alpha (p42) were altered during germination, suggesting that phosphorylation of these factors is developmentally regulated and correlates with the resumption of protein synthesis that occurs during germination. The phosphorylation of eIF-2beta (p38) or poly(A)-binding protein did not change either during germination or following a thermal stress. Only the phosphorylation state of two factors, eIF-4A and eIF-4B, changed following a heat shock, suggesting that plants may differ significantly from animals in the way in which their translational machinery is modified in response to a thermal stress.

Two different RNA binding activities for the AU-rich element and the poly(A) sequence of the mouse neuronal protein mHuC

HuC is one of the RNA binding proteins which are suggested to play important roles in neuronal differentiation and maintenance. We cloned and sequenced cDNAs encoding a mouse protein which is homologous to human HuC (hHuC). The longest cDNA encodes a 367 amino acid protein with three RNA recognition motifs (RRMs) and displays 96% identity to hHuC. Northern blot analysis showed that two different mRNAs, of 5.3 and 4.3 kb, for mouse HuC (mHuC) are expressed specifically in brain tissue. Comparison of cDNA sequences with the corresponding genomic sequence revealed that alternative 3' splice site selection generates two closely related mHuC isoforms. Iterative in vitro RNA selection and binding analyses showed that both HuC isoforms can bind with almost identical specificity to sequences similar to the AU-rich element (ARE), which is involved in the regulation of mRNA stability. Functional domain mapping using mHuC deletion mutants showed that the first RRM binds to ARE, that the second RRM has no RNA binding activity by itself, but facilitates ARE binding by the first RRM and that the third RRM has specific binding activity for the poly(A) sequence.

Development of an in vitro mRNA decay system for Escherichia coli: poly(A) polymerase I is necessary to trigger degradation

Using a novel Escherichia coli in vitro decay system in which polysomes are the source of both enzymes and mRNA, we demonstrate a requirement for poly(A) polymerase I (PAP I) in mRNA turnover. The in vitro decay of two different mRNAs (trxA and lpp) is triggered by the addition of ATP only when polysomes are prepared from s strain carrying the wild-type gene for PAP I (pcnB+). The relative decay rates of these two messages are similar in vitro and in vivo. Poly(A) tails are formed on both mRNAs, but no poly(A) are detected on the 3' end of mature 23S rRNA. The size distribution of poly(A) tails generated in vitro, averaging 50 nt in length, is comparable to that previously reported in vivo. PAP I activity is associated exclusively with the polysomes. Exogenously added PAP I does not restore mRNA decay to PAP I-polysomes, suggesting that, in vivo, PAP I may be part of a multiprotein complex. The potential of this in vitro system for analyzing mRNA decay in E. coli is discussed.

Mutations in trans-acting factors affecting mRNA decapping in Saccharomyces cerevisiae

The decay of several yeast mRNAs occurs by a mechanism in which deadenylation precedes decapping and subsequent 5'-to-3' exonucleolytic decay. In order to identify gene products required for this process of mRNA turnover, we screened a library of temperature-sensitive strains for mutants with altered mRNA degradation. We identified seven mutations in four genes that inhibited mRNA turnover. Two mutations were alleles of the XRN1 5'-to-3' exoribonuclease known to degrade mRNAs following decapping. One mutation defined a new gene, termed DCP1, which in subsequent work was demonstrated to encode a decapping enzyme or a necessary component of a decapping complex. The other mutations defined two additional genes, termed MRT1 and MRT3 (for mRNA turnover). Mutations in the MRT1 and MRT3 genes slow the rate of deadenylation-dependent decapping, show transcript-specific effects on mRNA decay rates, and do not affect the rapid turnover of an mRNA containing an early nonsense codon, which is degraded by a deadenylation-independent decapping mechanism. Importantly, cell extracts from mrt1 and mrt3 strains contain normal levels of the decapping activity required for mRNA decay. These observations suggest that the products of the MRT1 and MRT3 genes function to modulate the rates of decapping that occur following deadenylation.

Translation in plants--rules and exceptions

Translation processes in plants are very similar to those in other eukaryotic organisms and can in general be explained with the scanning model. Particularly among plant viruses, unconventional mRNAs are frequent, which use modulated translation processes for their expression: leaky scanning, translational stop codon readthrough or frameshifting, and transactivation by virus-encoded proteins are used to translate polycistronic mRNAs; leader and trailer sequences confer (cap-independent) efficient ribosome binding, usually in an end-dependent mechanism, but true internal ribosome entry may occur as well; in a ribosome shunt, sequences within an RNA can be bypassed by scanning ribosomes. Translation in plant cells is regulated under conditions of stress and during development, but the underlying molecular mechanisms have not yet been determined. Only a small number of plant mRNAs, whose structure suggests that they might require some unusual translation mechanisms, have been described.

PAN3 encodes a subunit of the Pab1p-dependent poly(A) nuclease in Saccharomyces cerevisiae

The Pab1p-dependent poly(A) nuclease (PAN) from Saccharomyces cerevisiae copurifies with polypeptides of approximately 127 and 76 kDa. Previously, it was demonstrated that the 127-kDa Pan2 protein is required for PAN activity (R. Boeck, S. Tarun, M. Reiger, J. Deardorff, S. Müller-Auer, and A.B. Sachs, J. Biol. Chem. 271:432-438, 1996). Here we demonstrate that the 76-kDa protein, encoded by the nonessential PAN3 gene, is also required for enzymatic activity. Deletion of PAN3 resulted in the loss of PAN activity in yeast extracts, and immunodepletion of Pan3p from purified PAN fractions abolished enzymatic activity. We show by coimmunoprecipitation and directed two-hybrid studies that the Pan2 and Pan3 proteins physically interact. In addition, we demonstrate that a deletion of PAN2, PAN3, or both resulted in similar increases in mRNA poly(A) tail lengths in vivo. These data strongly suggest that both Pan2p and Pan3p are required subunits of the PAN enzyme and that PAN functions in vivo to shorten mRNA poly(A) tails.

Inducers of erythroleukemic differentiation cause messenger RNAs that lack poly(A)-binding protein to accumulate in translationally inactive, salt-labile 80 S ribosomal complexes

Translation has an established role in the regulation of cell growth. Posttranslational modification of translation initiation and elongation factors or regulation of mRNA polyadenylation represent common means of regulating translation in response to mitogenic or developmental signals. Induced differentiation of Friend virus-transformed erythroleukemia cells is accompanied by a rapid decrease in the translation rate of these cells. Although inducers do not alter initiation factor modifications, characterization of their effect on mRNA translation provides evidence that this is mediated by the poly(A)-binding protein (PABP). Inducer exposure results in an increase in the amount of mRNA that sediments at 80 S and a decrease in the amount in polysomes. Although these 80 S ribosomes have characteristics previously attributed to "vacant ribosomal couples," including lability in 500 mM KCl and an inability to incorporate amino acids into protein, we provide evidence that these 80 S complexes are not vacant but contain mRNA that is stably bound to the 40 S subunit, whereas the 60 S subunit is dissociated from the complex by high salt. The absence of eukaryotic initiation factor 2 from these complexes suggests that translation has proceeded through subunit joining. Immunoblotting demonstrates that the mRNAs in these 80 S ribosomal complexes do not contain bound PABP and that this protein is found to be almost exclusively associated with translating polysomes. These data suggest that the PABP plays a role in the accumulation of these 80 S ribosomal.mRNA complexes and may facilitate the formation of translationally active salt-stable ribosomes.

A conserved 90 nucleotide element mediates translational repression of nanos RNA

Correct formation of the Drosophila body plan requires restriction of nanos activity to the posterior of the embryo. Spatial regulation of nanos is achieved by a combination of RNA localization and localization-dependent translation such that only posteriorly localized nanos RNA is translated. Cis-acting sequences that mediate both RNA localization and translational regulation lie within the nanos 3′ untranslated region. We have identified a discrete translational control element within the nanos 3′ untranslated region that acts independently of the localization signal to mediate translational repression of unlocalized nanos RNA. Both the translational regulatory function of the nanos 3′UTR and the sequence of the translational control element are conserved between D. melanogaster and D. virilis. Furthermore, we show that the RNA helicase Vasa, which is required for nanos RNA localization, also plays a critical role in promoting nanos translation. Our results specifically exclude models for translational regulation of nanos that rely on changes in polyadenylation.

Developmental regulation of expression of rabbit C-reactive protein and serum amyloid A genes

Serum amyloid A (SAA) and C-reactive protein (CRP) are acute phase plasma proteins which increases 100- to 1000-fold after inflammatory stimuli. In this study pregnant rabbits were given lipopolysaccharide (LPS) or subjected to laparotomy with fetal injections of LPS at different stages of gestation. Newborn rabbits were given LPS or saline. SAA and CRP mRNA were studied using Northern blot analyses and scanning densitometry. In vitro transcribed RNAs were used as standards for quantitative mRNA analyses. A gradual increase in LPS-induced SAA and CRP mRNA levels was observed during development, but only SAA mRNA induction was seen at gestational day 19. Fetal SAA and CRP mRNA induction was not seen after maternal LPS stimulation. The constitutive level of SAA and CRP mRNA was significantly lower in fetal rabbits than in adults. The control level of SAA mRNA in one-day-old rabbits was higher than the normal adult level, while the neonatal CRP mRNA level was lower. SAA2 seemed to be the major acute phase reactant in both fetal, neonatal and adult rabbits, while relatively more SAA3 was found during early developmental stages. The study demonstrated that CRP and three SAA genes are differentially regulated during development.

Translation is enhanced after silent nucleotide substitutions in A+T- rich sequences of the coding region of CD46 cDNA

Specific sequences in the coding region of CD46 (membrane cofactor protein) transcripts have been shown to have a marked effect on translation. Two A+T-rich regions of CD46 cDNA were altered by mutation without changing the CD46 amino acid sequence (silent nucleotide substitution). In one region, the A+T content was reduced from 78% to 55% and in the other a putative polyadenylation addition sequence was disrupted. In each example, mutated sequences transfected into COS-7 cells produced significantly more soluble or cell surface protein (up to a 20-fold increase) than wild-type sequences. The amount of cellular plasmid DNA and CD46 mRNA was not increased, suggesting that the effect was not due to increased transfection efficiency, or transcript synthesis or stability. Biosynthetically labelled transfected cells showed an increase in translation rate but cell-free in vitro translation studies demonstrated that wild-type and mutated transcripts were translated with similar efficiency. The data show that translation of CD46 is affected by specific mRNA coding sequences, 400-540 bases from the initiation codon, and suggest that these sequences require the structural integrity of the cell to exert their effect.

Ribosomal association of poly(A)-binding protein in poly(A)-deficient Saccharomyces cerevisiae

Poly(A)-binding protein, the most abundant eukaryotic mRNP protein, is known primarily for its association with polyadenylate tails of mRNA. In the yeast, Saccharomyces cerevisiae, this protein (Pabp) was found to be essential for viability and has been implicated in models featuring roles in mRNA stability and as an enhancer of translation initiation. Although the mechanism of action is unknown, it is thought to require an activity to bind poly(A) tails and an additional capacity for an interaction with 60 S ribosomal subunits, perhaps via ribosomal protein L46 (Rpl46). We have found that a significant amount of Pabp in wild-type cells is not associated with polyribosome complexes. The remaining majority, which is found in these complexes, maintains its association even in yeast cells deficient in polyadenylated mRNA and/or Rpl46. These observations suggest that Pabp may not require interaction with poly(A) tails during translation. Further treatment of polyribosome lysates with agents known to differentially disrupt components of polyribosomes indicated that Pabp may require contact with some RNA component of the polyribosome, which could be either non-poly(A)-rich sequences of the translated mRNA or possibly a component of the ribosome. These findings suggest that Pabp may possess the ability to bind to ribosomes independently of its interaction with poly(A). We discuss these conclusions with respect to current models suggesting a multifunctional binding capacity of Pabp.

Translational control of poly(A)-binding protein expression

Poly(A)-binding protein (PABP) is important for translation of eukaryotic mRNA and may be involved in shortening of its poly(A) tract. In many eukaryotic cells, this mRNA is inefficiently translated. The 5' untranslated region (UTR) of PABP mRNA has several adenine-rich regions which may serve as the PABP-binding sites to control its translation by a feed-back mechanism. This postulate was tested by using in vitro transcribed PABP mRNA and a rabbit reticulocyte lysate cell-free system. Results of our studies show that removal of the putative PABP-binding sites from the 5' UTR of this mRNA enhances its translation in the rabbit reticulocyte cell-free system. Furthermore, in vitro translation of the full-length PABP mRNA was inhibited by addition of purified PABP to the cell-free system. In contrast, translation of truncated mRNA lacking the putative PABP-binding sites at the 5' UTR was not inhibited by exogenous PABP. We have also tested the ability of purified PABP to bind to the 5' UTR of PABP mRNA using ultraviolet-mediated covalent cross-linking of RNA and proteins in vitro. Our results show that exogenous PABP binds to the 5' UTR of its full-length mRNA. Furthermore, incubation of PABP mRNA in rabbit reticulocyte lysate also led to binding of the endogenous PABP within the first 223 nucleotides of the 5' UTR. The adenine-rich regions are located within this segment of PABP mRNA. Following incubation of PABP mRNA in the reticulocyte lysate cell-free system under conditions of mRNA translation, the polysomal and non-translated free mRNA fractions were separated by centrifugation. Analysis of free and polysomal mRNA-protein (mRNP) complexes following ultraviolet-induced cross-linking showed that the free mRNP population was preferentially enriched in PABP. Results of our studies, therefore, suggest that PABP mRNA translation may be repressed by a unique feed-back mechanism.

Initiation of protein synthesis in eukaryotic cells

It is becoming increasingly apparent that translational control plays an important role in the regulation of gene expression in eukaryotic cells. Most of the known physiological effects on translation are exerted at the level of polypeptide chain initiation. Research on initiation of translation over the past five years has yielded much new information, which can be divided into three main areas: (a) structure and function of initiation factors (including identification by sequencing studies of consensus domains and motifs) and investigation of protein-protein and protein-RNA interactions during initiation; (b) physiological regulation of initiation factor activities and (c) identification of features in the 5' and 3' untranslated regions of messenger RNA molecules that regulate the selection of these mRNAs for translation. This review aims to assess recent progress in these three areas and to explore their interrelationships.

RNA recognition and translational regulation by a homeodomain protein

In Drosophila, the primary determinant of anterior pattern is the gradient morphogen bicoid (bcd), a homeodomain protein that binds DNA and transcriptionally activates target genes at different threshold concentrations. Here we present evidence that bcd also binds RNA and acts as a translational repressor to generate an opposing gradient of the homeodomain protein caudal (cad). RNA binding by bcd seems to involve direct interactions between the bcd homeodomain and discrete target sequences within the 3' untranslated region of the cad messenger RNA and to block the initiation of cad translation.

Efficient extraction and partial purification of the polyribosome-associated stem-loop binding protein bound to the 3' end of histone mRNA

Replication-dependent histone mRNAs end in a highly conserved stem-loop sequence rather than a polyA sequence. A 45-kDa stem-loop binding protein (SLBP), which specifically binds the stem-loop of histone mRNA, is present in both polyribosomes and nuclei. An identical 45-kDa protein, as determined by partial protease digestion, is cross-linked to a 30 nt RNA containing the 3' stem-loop from both nuclei and polyribosomes. The SLBP can also be detected by a Northwestern blot procedure using the 30 nt RNA as a probe. As judged from the Northwestern assay, more than 90% of the SLBP in the cell is found in the polyribosomes with the remaining SLBP localized to the nucleus. Only 5-10% of the SLBP could be extracted from the polyribosomes with salt. Treatment of the polyribosomes with micrococcal nuclease prior to salt extraction solubilized 5-10 times more SLBP as an RNA-protein complex. The SLBP could be subsequently partially purified from this complex.

The yeast Pan2 protein is required for poly(A)-binding protein-stimulated poly(A)-nuclease activity

The removal of the mRNA poly(A) tail in the yeast Saccharomyces cerevisiae is stimulated by the poly(A)-binding protein (Pab1p). A large scale purification of the Pab1p-stimulated poly(A) ribonuclease (PAN) identifies a 76-kDa and two 135-Da polypeptides as candidate enzyme subunits. Antibodies against the Pan1p protein, which is the minor 135-kDa protein in the preparation, can immunodeplete Pan1p but not PAN activity. The protein sequence of the major 135-kDa protein, Pan2p, reveals a novel protein that was also found in the previously reported PAN purification (Sachs, A. B., and Deardorff, J. A. (1992) Cell 70, 961-973). Deletion of the non-essential PAN2 gene results in an increase of the average length of mRNA poly(A) tails in vivo, and a loss of Pab1p-stimulated PAN activity in crude extracts. These data confirm that Pan2p and not Pan1p is required for PAN activity, and they suggest that ribonucleases other than the Pab1p-stimulated PAN are capable of shortening poly(A) tails in vivo.

Xenopus poly (A) binding protein maternal RNA is localized during oogenesis and associated with large complexes in blastula

Maternal mRNAs are synthesized during oogenesis and often stored for use during early embryogenesis, before the onset of zygotic transcription. The temporal and spatial regulation of maternal RNAs is likely to be crucial mechanism for the establishment of the body pattern. In the course of a study that identified a Xenopus maternal mRNA that is translationally regulated along the dorsoventral axis, several RNAs were found to behave anomalously in polysomal analysis and are further characterized here. As controls for polysome analysis, elF4E RNA and D7.1 RNA were equally translated in both dorsal and ventral cells, whereas the cell-cell signaling factor noggin RNA was not translated in either cell type. Maternal RNAs encoding poly (A) binding protein (PABP), Vg1 and Xcat-2 were associated with large complexes that, in contrast to polysomes, were not dissociated in magnesium-free buffer. Vg1 and Xcat-2 maternal mRNAs have been shown to be localized during oogenesis to the vegetal hemisphere of the oocyte [Rebagliati et al., 1985; Mosquera et al., 1993]. In situ hybridization analysis indicated that PABP RNA was also localized during oogenesis, to the animal hemisphere in stage VI oocytes. This suggests that association of maternal mRNAs with large EDTA-insensitive mRNP complexes is correlated with intracellular localization, but the specific localization within the oocyte is dependent upon the RNA species.

Multiple regions of yeast ribosomal protein L1 are important for its interaction with 5 S rRNA and assembly into ribosomes

Yeast ribosomal protein L1 binds to 5 S rRNA and can be released from 60 S ribosomal subunits as an intact ribonucleoprotein particle. To identify residues important for binding of Saccharomyces cerevisiae rpL1 to 5 S rRNA and assembly into functional ribosomes, we have isolated mutant alleles of the yeast RPL1 gene by site-directed and random mutagenesis. The rpl1 mutants were assayed for association of rpL1 with 5 S rRNA in vivo and in vitro and assembly of rpL1 into functional 60 S ribosomal subunits. Consistent with previous data implicating the importance of the carboxyl-terminal 47 amino acids of rpL1 for binding to 5 S rRNA in vitro, we find that deletion of the carboxyl-terminal 8, 25, or 44 amino acids of rpL1 confers lethality in vivo. Missense mutations elsewhere in rpL1 also affect its function, indicating that multiple regions of rpL1 are important for its association with 5 S rRNA and assembly into ribosomes.

A common function for mRNA 5' and 3' ends in translation initiation in yeast

The mRNA poly(A) tail and its associated poly(A) binding protein (Pab1p) are ubiquitous in eukaryotes. The function of the poly(A) tail is to stabilize mRNA and to stimulate its translation. The development of a poly(A)- and cap-dependent yeast in vitro translation system has allowed us to understand how poly(A) stimulates translation. We find that Pab1p but not the cap binding protein eIF-4E is required for poly(A) tail-dependent translation, and that the Pab1p-poly(A) tail complex functions to recruit the 40S ribosomal subunit to the mRNA. These data introduce a new step into the pathway of translation initiation and merge the translational functions of the two ends of mRNA.

iPABP, an inducible poly(A)-binding protein detected in activated human T cells

The poly(A)-binding protein (PABP) binds to the poly(A) tail present at the 3' ends of most eukaryotic mRNAs. PABP is thought to play a role in both translation and mRNA stability. Here we describe the molecular cloning and characterization of an inducible PABP, iPABP, from a cDNA library prepared from activated T cells. iPABP shows 79% sequence identity to PABP at the amino acid level. The RNA binding domains of iPABP and PABP are nearly identical, while their C termini are more divergent. Like PABP, iPABP is primarily localized to the cytoplasm. iPABP is expressed at low levels in resting normal human T cells; following T-cell activation, however, iPABP mRNA levels are rapidly up-regulated. In contrast, PABP is constitutively expressed in both resting and activated T cells. iPABP mRNA was also expressed at much higher levels than PABP mRNA in heart and skeletal muscle tissue. These data suggest that the regulation of cytoplasmic poly(A)-binding activity is more complex than previously believed. In most tissues, poly(A)-binding activity is likely to be the result of the combined effects of constitutively expressed PABP and iPABP, whose expression is subject to more complex regulation.

Degradation of the soybean ribulose-1,5-bisphosphate carboxylase small-subunit mRNA, SRS4, initiates with endonucleolytic cleavage

The degradation of the soybean SRS4 mRNA, which encodes the small subunit of ribulose-1,5-bisphosphate carboxylase, yields a set of proximal (5' intact) and distal (3' intact) products both in vivo and in vitro. These products are generated by endonucleolytic cleavages that occur essentially in a random order, although some products are produced more rapidly than others. Comparison of sizes of products on Northern (RNA) blots showed that the combined sizes of pairs of proximal and distal products form contiguous full-length SRS4 mRNAs. When the 3' ends of the proximal products and the 5' ends of the distal products were mapped by S1 nuclease and primer extension assays, respectively, both sets of ends mapped to the same sequences within the SRS4 mRNA. A small in vitro-synthesized RNA fragment containing one cleavage site inhibited cleavage of all major sites, equivalently consistent with one enzymatic activity generating the endonucleolytic cleavage products. These products were rich in GU nucleotides, but no obvious consensus sequence was found among several cleavage sites. Preliminary evidence suggested that secondary structure could play a role in site selection. The structures of the 5' ends of the proximal products and the 3' ends of the distal products were examined. Proximal products were found with approximately equal frequency in both m7G cap(+) and m7G cap(-) fractions, suggesting that the endonucleolytic cleavage events occurred independently of the removal of the 5' cap structure. Distal products were distributed among fractions with poly(A) tails ranging from undetectable to greater than 100 nucleotides in length, suggesting that the endonucleolytic cleavage events occurred independently of poly(A) tail shortening. Together, these data support a stochastic endonuclease model in which an endonucleolytic cleavage event is the initial step in SRS4 mRNA degradation.

Cordycepin blocks recovery of non-heat-shock mRNA translation following heat shock in Drosophila

Treatment of cells with cordycepin (3-deoxyadenosine), an inhibitor of cytoplasmic adenylation, blocks the restoration of normal translation following heat shock. Cordycepin also reduces heat-shock protein 70 (Hsp70) protein synthesis greater than 10-fold, while having little to no effect on mRNA accumulation. Parallel analysis of the poly(A)-binding protein detects no change in its abundance during heat shock or subsequent recovery. These results suggest that normal, non-heat-shock mRNA translational repression during heat shock may be caused by deadenylation, and that readenylation is required for restoration of activity. However, three independent analyses of the adenylation status of mRNAs during heat shock and recovery indicate that no significant changes in polyadenylation occur. (a) The total poly(A) content decreases by only about 10% during heat shock; (b) the size of the poly(A) tract decreases only marginally, from an average length of 75-90 nucleotides in non-heated cells to 45-60 nucleotides during heat shock; (c) virtually all mRNAs bind to oligo d(T)-cellulose, whether extracted from normal-temperature, heat-shock or recovered cells. Our results are most consistent with a model where the process of readenylation, rather than the specific poly(A) tail length, influences translational activation during recovery, paralleling a proposed model for the activation of translation during Xenopus oocyte maturation.

Identification of an additional gene required for eukaryotic nonsense mRNA turnover

Loss of function of any one of three UPF genes prevents the accelerated decay of nonsense mRNAs in Saccharomyces cerevisiae. We report the identification and DNA sequence of UPF3, which is present in one nonessential copy on chromosome VII. Upf3 contains three putative nuclear localization signal sequences, suggesting that it may be located in a different compartment than the cytoplasmic Upf1 protein. Epitope-tagged Upf3 (FLAG-Upf3) does not cofractionate with polyribosomes or 80S ribosomal particles. Double disruptions of UPF1 and UPF3 affect nonsense mRNA decay in a manner indistinguishable from single disruptions. These results suggest that the Upf proteins perform related functions in a common pathway.

Multiple functions for the poly(A)-binding protein in mRNA decapping and deadenylation in yeast

The first step in the decay of many eukaryotic mRNAs is shortening of the poly(A) tail. In yeast, deadenylation leads to mRNA decapping and subsequent 5' --> 3' exonucleolytic degradation of the transcript body. We have determined that the major poly(A)-binding protein Pab1p plays at least two critical roles in this pathway. First, mRNAs in pab1 delta strains were decapped prior to deadenylation. This observation defines a new function for Pab1p as an inhibitor of mRNA decapping. Moreover, mutations that inhibit mRNA turnover suppress the inviability of a pab1 delta mutation, suggesting that premature mRNA decapping in pab1 delta strains contributes to cell death. Second, we find that Pab1p is not required for deadenylation, although in its absence poly(A) tail shortening rates are significantly reduced. In addition, in the absence of Pab1p, newly synthesized mRNAs had poly(A) tails longer than those in wild-type strains and showed an unexpected temporal delay prior to the initiation of deadenylation and degradation. These results define new and critical functions for Pab1p in the regulation of mRNA decapping and deadenylation, two important control points in the specification of mRNA half-lives. Moreover, these results suggest that Pab1p functions in additional phases of mRNA metabolism such as mRNP maturation.

Translational regulation in response to changes in amino acid availability in Neurospora crassa

We examined the regulation of Neurospora crassa arg-2 and cpc-1 in response to amino acid availability.arg-2 encodes the small subunit of arginine-specific carbamoyl phosphate synthetase; it is subject to unique negative regulation by Arg and is positively regulated in response to limitation for many different amino acids through a mechanism known as cross-pathway control. cpc-1 specifies a transcriptional activator important for crosspathway control. Expression of these genes was compared with that of the cytochrome oxidase subunit V gene, cox-5. Analyses of mRNA levels, polypeptide pulse-labeling results, and the distribution of mRNA in polysomes indicated that Arg-specific negative regulation of arg-2 affected the levels of both arg-2 mRNA and arg-2 mRNA translation. Negative translational effects on arg-2 and positive translational effects on cpc-1 were apparent soon after cells were provided with exogenous Arg. In cells limited for His, increased expression of arg-2 and cpc-1, and decreased expression of cox-5, also had translational and transcriptional components. The arg-2 and cpc-1 transcripts contain upstream open reading frames (uORFs), as do their Saccharomyces cerevisiae homologs CPA1 and GCN4. We examined the regulation of arg-2-lacZ reporter genes containing or lacking the uORF start codon; the capacity for arg-2 uORF translation appeared critical for controlling gene expression.