Capped mRNAs with reduced secondary structure can function in extracts from poliovirus-infected cells

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.

mRNAs containing the unstructured 5' leader sequence of alfalfa mosaic virus RNA 4 translate inefficiently in lysates from poliovirus-infected HeLa cells

Poliovirus infection is accompanied by translational control that precludes translation of 5'-capped mRNAs and facilitates translation of the uncapped poliovirus RNA by an internal initiation mechanism. Previous reports have suggested that the capped alfalfa mosaic virus coat protein mRNA (AIMV CP RNA), which contains an unstructured 5' leader sequence, is unusual in being functionally active in extracts prepared from poliovirus-infected HeLa cells (PI-extracts). To identify the cis-acting nucleotide elements permitting selective AIMV CP expression, we tested capped mRNAs containing structured or unstructured 5' leader sequences in addition to an mRNA containing the poliovirus internal ribosome entry site (IRES). Translations were performed with PI-extracts and extracts prepared from mock-infected HeLa cells (MI-extracts). A number of control criteria demonstrated that the HeLa cells were infected by poliovirus and that the extracts were translationally active. The data strongly indicate that translation of RNAs lacking an internal ribosome entry site, including AIMV CP RNA, was severely compromised in PI-extracts, and we find no evidence that the unstructured AIMV CP RNA 5' leader sequence acts in cis to bypass the poliovirus translational control. Nevertheless, cotranslation assays in the MI-extracts demonstrate that mRNAs containing the unstructured AIMV CP RNA 5' untranslated region have a competitive advantage over those containing the rabbit alpha-globin 5' leader. Previous reports of AIMV CP RNA translation in PI-extracts likely describe inefficient expression that can be explained by residual cap-dependent initiation events, where AIMV CP RNA translation is competitive because of a diminished quantitative requirement for initiation factors.

A pollen-specific DEAD-box protein related to translation initiation factor eIF-4A from tobacco

A pollen-specific sequence, NeIF-4A8, has been isolated from a cDNA library from mature pollen of Nicotiana tabacum cv. Samsun. NeIF-4A8 is a full-length cDNA whose deduced amino acid sequence exhibits high homology to the eucaryotic translation initiation factor eIF-4A from mouse, Drosophila and tobacco. eIF-4A is an RNA helicase which belongs to the supergene family of DEAD-box proteins. Northern blot analysis with a gene-specific probe showed strict anther-specific expression of NeIF-4A8 starting at microspore mitosis. With antibodies raised against tobacco eIF-4A the presence of abundant eIF-4A-related proteins in developing anthers and pollen grains was demonstrated. The genomic analysis shows that the coding region is split by three introns whereas a large, fourth intron is situated in the 5'-untranslated region. A promoter construct with 2137 bp of upstream sequence fused to the GUS reporter gene was used to confirm that the expression is confined to the haploid cells within the anther. NeIF-4A8 is a prime candidate formediating translational control in the developing male gametophyte.

The role of the 5' untranslated region of eukaryotic messenger RNAs in translation and its investigation using antisense technologies

This chapter discusses the recent advances in the field of translational control and the possibility of applying the powerful antisense technology to investigate some of the unanswered questions, especially those pertaining to the role of the 5’untranslated region ( UTR) on translation initiation. Translational regulation is predominantly exerted during the initiation phase that is considered to be the rate-limiting step. Two types of translational regulation can be distinguished: global, in which the initiation rate of (nearly) all cellular messenger RNA (mRNA) is controlled and selective, in which the translation rate of specific mRNAs varies in response to the biological stimuli. In most cases of global regulation, control is exerted via the phosphorylation state of certain initiation factors, whereas only a few examples of selective regulation have been characterized well enough to define the underlying molecular events. Interestingly, cis-acting regulatory sequences, affecting translation initiation, have been found not only in the 5’UTRs of selectively regulated mRNAs, but also in the 3’UTRs. Thus, in addition to the protein encoding open reading frames, both the 5’ and 3’UTRs of mRNAs must be considered for their effect on translation.

Regulation of translation and cell growth by eIF-4E

This review discusses the regulation of a key controlling step in the initiation of protein synthesis, the binding of mRNA to ribosomes. Particular focus is given to the phosphorylation of the cap-binding factor, eIF-4E, and the role of this factor in the regulation of cell growth.

Remarks on the mechanism of ribosome binding to eukaryotic mRNAs

It is evident from the data discussed here that the mechanism and the rules for mRNA binding in eukaryotes are complex and not well defined. The major points of this review are (1) ribosome binding could be preceded by the unwinding of mRNA secondary structure; (2) there is no obligatory ribosome entry through the 5' end of the mRNA; (3) there is no obligatory linear "scanning" of the 5'UTR; and (4) there are some interesting similarities between prokaryotes and eukaryotes in the mode of ribosome binding to mRNA, particularly in the ability of the small ribosomal subunit to diffuse or "scan" on the mRNA, and in the requirement for a minimally structured RNA for efficient ribosome binding.

Relationship of eukaryotic initiation factor 3 to poliovirus-induced p220 cleavage activity

The cleavage of the p220 subunit of eukaryotic initiation factor 4F (eIF-4F) that is induced by the poliovirus protease 2A has been shown previously to require another translation initiation factor, eIF-3. The role of eIF-3 in this cleavage reaction, however, is not known. An antiserum was raised against human eIF-3 and used to analyze the eIF-3 subunit composition in poliovirus-infected and uninfected HeLa cells and after incubation of eIF-3 in vitro with viral 2A protease. No evidence for 2Apro-dependent cleavage of any eIF-3 subunit was detected. Infected cells contain an activity that catalyzes the cleavage of p220 to a specific set of cleavage products. This activity is thought to be an activated form of a latent cellular protease. The p220-specific cleavage activity was partially purified. It was resolved from eIF-3 by both gel filtration and anion-exchange chromatography. Neither intact eIF-3 nor any detectable subunits of eIF-3 were found to copurify with the p220-specific cleavage activity. The latter activity behaves as a protein of 55,000 to 60,000 molecular weight and is inhibited by alkylating agents and metals, which indicates the presence of essential thiol groups. When this activity was incubated with partially purified p220, cleavage occurred only in the presence of eIF-3. Thus, eIF-3 appears to play a role in the p220 cleavage cascade which is subsequent to the 2Apro-induced activation of the p220-specific protease.

Signal transduction and regulation of translation initiation

Regulation of the rate of protein synthesis is important in the control of cellular proliferation. Changes in the rate of protein translation are brought about primarily at the level of initiation, which is usually rate limiting. This regulation involves the reversible phosphorylation of key initiation factors. Translation initiation factors eIF-4F, eIF-4B, and ribosomal protein S6 are phosphorylated in response to a wide variety of mitogens, growth factors, and tyrosine kinase oncogenes. Thus, translation initiation factors are important components of signal transduction pathways activated by extracellular factors and oncogenes. Of particular interest is the messenger RNA 5' cap-binding protein, eIF-4E. Overexpression of eIF-4E in fibroblasts results in malignant transformation, suggesting that it is an important transducer of growth signals, and that aberrant expression of a translation factor can cause malignancy. Elucidation of the components of the signalling pathways which regulate initiation factor activity should increase our understanding of how extracellular factors and oncogenes effect cellular proliferation, and the role that translation plays in this process.

The ATP requirement for initiation of eukaryotic translation varies according to the mRNA species

The requirement for ATP for initiation of eukaryotic mRNA translation was tested using gel-filtered rabbit reticulocyte lysates incubated with labelled Met-tRNAfMet and exogenous RNA templates, and assaying the formation of labelled 80S initiation complexes in the presence of GTP, or labelled 40S initiation complexes in the presence of a non-hydrolysable analogue of GTP. Initiation complex formation on globin mRNA, or on capped viral RNAs such as papaya mosaic virus RNA and tobacco mosaic virus RNA, was strongly stimulated by ATP. In contrast, initiation complex formation on (uncapped) encephalomyocarditis virus RNA was uninfluenced by the presence or absence of ATP, which may be correlated with the recent evidence for scanning-independent internal initiation on this viral RNA. In addition, initiation complex formation on uncapped cowpea mosaic virus RNA and on poly(A,U,G) was only slightly stimulated by ATP, much less than in the case of the capped RNAs. These results suggest that most of the ATP hydrolysed during translation initiation is consumed in cap-dependent processes, probably in unwinding the mRNA, and relatively little in the actual migration or scanning of 40S subunits along the mRNA.

The 5'-untranslated region of picornaviral genomes

Picornaviruses are small naked icosahedral viruses with a single-stranded RNA genome of positive polarity. According to current taxonomy, the family includes four genera: Enterouirus (polioviruses, coxsackieviruses, echoviruses, and other enteroviruses), Rhinovirus, Curdiouirus [encephalomyocarditis virus (EMCV), mengovirus, Theiler's murine encephalomyelitis virus (TMEV)], and Aphthouirus [foot-and-mouth disease viruses (FMDV)]. There are also some, as yet, unclassified picornaviruses [e.g., hepatitis A virus (HAW] that should certainly be assessed as a separate genus. Studies on the molecular biology of picornaviruses might be divided into two periods: those before and after the first sequencing of the poliovirus genome. The 5'-untranslated region (5-UTR) of the viral genome was one of the unexpected problems. This segment proved to be immensely long: about 750 nucleotides or ∼10% of the genome length. There were also other unusual features (e.g., multiple AUG triplets preceding the single open reading frame (ORF) that encodes the viral polyprotein). This chapter shows that the picornaviral 5-UTRs are not only involved in such essential events as the synthesis of viral proteins and RNAs that could be expected to some extent, although some of the underlying mechanisms appeared to be quite a surprise, but also may determine diverse biological phenotypes from the plaque size or thermosensitivity of reproduction to attenuation of neurovirulence. Furthermore, a close inspection of the 5-UTR structure unravels certain hidden facets of the evolution of the picornaviral genome. Finally, the conclusions drawn from the experiments with the picornaviral5-UTRs provide important clues for understanding the functional capabilities of the eukaryotic ribosomes.

Polyadenylated RNA sequences from vaccinia virus-infected cells selectively inhibit translation in a cell-free system: structural properties and mechanism of inhibition

The mechanism of vaccinia virus-induced selective inhibition of host cell protein synthesis was studied in a nonpermissive (Chinese hamster ovary, CHO) and in a permissive mouse cell line ( L cells). Small polyadenylated RNAs obtained from uninfected and infected cells were fractionated into six size classes by polyacrylamide gel electrophoresis. The RNAs from the first two largest fractions (greater than 500 nucleotide, nt) were translated into some low-molecular-weight polypeptides, whereas, the RNAs from the remaining fractions (400-500, 300-400, 200-300, and 100-200 nt) had no translational activity in reticulocyte lysates. When these nontranslating polyadenylated short sequences (POLADS) were added to the cell-free system together with HeLa cell mRNAs, translation was inhibited from 70%, by the 400- to 500-nt fraction, to about 20%, by the 100- to 200-nt fraction. The degree of inhibition of protein synthesis was clearly dependent on the size of POLADS. The translation of vaccinia virus mRNAs in the cell-free system was inhibited by about 25% with the 400- to 500-nt fraction, by 5% with the 300- to 400-nt fraction, while the smaller size POLADS had no inhibitory effect. The inhibition of HeLa cell and vaccinia virus mRNA translation by POLADS was reversed by the simultaneous addition of oligo(dT) to the cell-free system. POLADS were also obtained from uninfected cells, but they inhibited the translation of HeLa cell and vaccinia virus mRNAs to a much lesser extent. The removal of the poly(A) moiety from POLADS by treatment with ribonuclease H and oligo(dT) abolished their inhibitory effect on HeLa cell mRNA translation. The average length of the poly(A) tails of POLADS obtained from infected cells was longer than that of POLADS from normal cells. Inhibition of HeLa cell mRNA translation mediated by POLADS in the cell-free system was reversed (approximately 70%) by addition of crude initiation factors (ribosomal salt wash, RSW). Significantly, inhibition of translation of POLADS was reversed (greater than 90%) by addition of purified poly(A) binding protein (PAB). Purified initiation factor 4A (eIF-4A) also reversed this inhibition, but to a lesser extent than RSW and PAB. Our results show that the translation of vaccinia virus mRNAs is resistant to POLADS, suggesting that POLADS, by virtue of their long poly(A) tails, may sequester PAB and thus, play a role in selective inhibition.

Inhibition of avian retrovirus protein synthesis in the presence of host cellular mRNA

Two conditions of in vitro translation are described in which the RNA of Rous sarcoma virus is significantly less efficient at protein synthesis than host cellular chick mRNA. In direct competition, addition of increasing amounts of host cell mRNA inhibits synthesis of gag gene products from RSV RNA. When the denaturant dimethyl sulfoxide is added to a mixture of the two RNA's the inhibition is even greater. These results support the idea that protein synthesis from avian retrovirus RNA is negatively regulated in vivo in the presence of the host cell mRNA.

Dissociation of double-stranded polynucleotide helical structures by eukaryotic initiation factors, as revealed by a novel assay

A new technique has been applied to the study of the RNA secondary structure unwinding activity of the eukaryotic initiation factors (eIFs) 4F, 4A, and 4B. Secondary structures were generated at the 5' ends of reovirus and globin mRNA molecules by hybridization with 32P-labeled cDNA molecules 15 nucleotide residues long. The dissociation of the labeled cDNAs from the mRNAs was assayed by a gel filtration chromatography procedure which separates the free cDNAs from mRNAs and mRNA/cDNA hybrids. When the three factors were tested alone, only eIF-4F stimulated dissociation of hybrids. The combination of eIF-4A plus eIF-4B also exhibited a strong hybrid dissociating activity, which was markedly temperature dependent. Under optimum conditions, up to 90% of the hybrid structures are disrupted in 60 min. These results demonstrate for the first time that stable double-stranded regions can be melted and dissociated by eIFs. They also characterize more precisely the first step in the structure unwinding reaction.

Control of reovirus messenger RNA translation efficiency by the regions upstream of initiation codons

The 10 species of reovirus messenger RNA are translated in vivo with efficiencies/frequencies that differ by as much as 100-fold. The s1 mRNA, which is translated 10 times less efficiently than the s4 mRNA but 10 times more efficiently than the/1 and m1 mRNAs, has a unique BamH1 cleavage site located immediately downstream of its initiation codon. Because the reovirus mRNAs have been cloned, this provides the opportunity for placing modified and altered sequences upstream of its coding sequence. The translation efficiencies of the variant mRNAs, transcribed via the SP6 in vitro transcription system, can then be measured in the rabbit reticulocyte lysate in vitro translation system. Using this system it was found that replacing the 5'-upstream sequence of the s1 mRNA with that of the s4 mRNA increases its in vitro translation efficiency by 4-fold; that the trinucleotide immediately upstream of the s1 initiation codon renders it very weak, and that it is only slightly superior to the weakest Kozak consensus sequence; that the nature of the nucleotides further upstream than position -3 can profoundly affect translation efficiency; that the nature of this effect is in turn markedly modified by the nature of nucleotides in positions -1 to -3; and that there is a minimum optimal 5'-upstream sequence length of about 14 nucleotides. We also investigated the effect of secondary structure involvement on the ability of 5'-upstream sequences to promote translation. Two effects were noted. First, being part of moderately stable stem loops (delta G, -18 kcal/mol) decreased translation efficiency about 3-fold; second, mRNA in which only three 5'-terminal nucleotides were unpaired were translated five times less efficiently than mRNA in which six nucleotides were unpaired. Accessibility of the 5'-cap as well as secondary structure of the 5'-upstream sequences are therefore factors that affect translation efficiency. Finally, we showed that the m1 mRNA, which is transcribed very poorly in vivo, is translated very efficiently in vitro; and that its 5'-upstream sequence is as effective in increasing protein sigma 1 formation as that of s4 mRNA. Since both m1 mRNA and protein mu 2 are stable in infected cells, the reason why m1 mRNA is translated so inefficiently in vivo therefore remains unexplained.

Studies on the mechanism of translational enhancement by the 5'-leader sequence of tobacco mosaic virus RNA

Translation of foreign mRNAs is enhanced by a cis-acting derivative (omega') of the 5'-leader sequence (omega) of tobacco mosaic virus RNA (vulgare strain). To explain this effect we have conducted several experiments in vitro. 1. The presence of various 5'-terminal sequences, including omega', did not significantly increase the half-lives of chloramphenicol acetyltransferase (CAT) or neomycin phosphotransferase (NPTII) mRNAs in wheat-germ extract. Also, a long leader sequence, unrelated to omega', did not enhance expression of NPTII mRNA in vitro. 2. The ability of several leader sequences, including omega', to form multiple initiation complexes with 80S (wheat germ) ribosomes was examined using CAT or NPTII mRNAs incubated in the presence of sparsomycin. Formation of disome complexes was unrelated to the capacity of a 5'-leader sequence to enhance translation. 3. Expression of CAT mRNA in both wheat germ extract and messenger-dependent rabbit reticulocyte lysate was less susceptible to inhibition by increasing salt concentration when a 5'-proximal omega' sequence was present. This effect was less marked when the CAT mRNA was capped. Conversely at high salt concentrations, capping was less stimulatory for mRNA with a 5'-proximal omega' sequence. These data suggest that omega' and the cap enhance translation, at least in part, by a similar mechanism. We propose that both features reduce RNA secondary structure, thereby rendering the 5' terminus more accessible to scanning by 40S ribosomal subunits and/or interaction with associated initiation factors. This conclusion was supported by computer-based secondary-structure analyses of our SP6 RNA polymerase transcript sequences. The ability of 5' leader sequences from brome mosaic virus RNA 3, alfalfa mosaic virus RNA 4, and the genomic RNAs of turnip yellow mosaic virus, Rous sarcoma virus or tobacco mosaic virus (tomato strain) to enhance mRNA translation in eukaryotic systems may also be correlated with their respective secondary structures. A different mechanism probably accounts for the omega'-dependent enhancement of mRNA expression in Escherichia coli or in E. coli cell-free systems.

Identification of a protein kinase activity in rabbit reticulocytes that phosphorylates the mRNA cap binding protein

The 25 kDa mRNA cap binding protein can be purified in a partially phosphorylated state and the extent of its phosphorylation appears to be regulated during heat shock and mitosis in mammalian cells. We demonstrated that a nonabundant serine protein kinase activity exists in rabbit reticulocytes that phosphorylates the 25 kDa cap binding protein in both the free (eIF-4E) and complexed (eIF-4F) state. This kinase was not inhibited by the cAMP-dependent protein kinase inhibitory peptide IAAGRTGRRNAIHDILVAA, did not phosphorylate S6 ribosomal protein, did not phosphorylate p220 of eIF-4F as protein kinase C does and no other substrates for this kinase were apparent in reticulocyte ribosomal salt wash. The molecular identity of this kinase, the specific site(s) of eIF-4E that it phosphorylates and its in vivo regulatory role remain to be studied.

A fluorescence study of the binding of eucaryotic initiation factors to messenger RNA and messenger RNA analogues

The binding of the eucaryotic polypeptide chain initiation factors (eIFs) 4A, 4B, and 4F to poly(1,N6-ethenoadenylic acid) [poly(epsilon A)] was investigated by fluorescence spectroscopy. Competition experiments allowed us to determine the relative affinity of these proteins for mRNA cap analogues and the triplets AUG, GUG, UUU, UAA, and UGA. The salt dependence of eIF-4A binding to poly(epsilon A) and mRNA suggested that the binding was largely electrostatic and was enhanced in the presence of Mg2+ and ATP. The size of the binding site of eIF-4A, eIF-4B, and eIF-4F on poly(epsilon A) was approximately 13, 25, and 35 nucleotides, respectively. Fluorescence studies with the cap analogue 7-methylguanosine triphosphate as well as competition studies with poly(epsilon A) provide further evidence for a direct interaction of eIF-4F with the cap region. There was no evidence that either eIF-4B or eIF-4A bound the mRNA cap directly. In contrast to the other two factors, eIF-4B was found to bind preferentially to AUG, and of all the triplets tested, AUG was the most effective competitor for poly(epsilon A) binding.

Enhanced translation of chimaeric messenger RNAs containing a plant viral untranslated leader sequence

Eukaryotic messenger RNAs are translated with unequal efficiencies in vivo and in vitro and the molecular basis of this phenomenon is not understood. As an approach to understanding the role of the 5' untranslated leader sequence in regulating mRNA translational efficiency, chimaeric mRNAs have been generated by joining a heterologous leader to complementary DNA (cDNA) sequences, followed by in vitro transcription using SP6 RNA polymerase and in vitro protein synthesis. We used the untranslated leader from the coat protein mRNA of alfalfa mosaic virus (AMV RNA 4), a well-translated, highly competitive message, to replace the leader sequence of barley alpha-amylase (B alpha A) and human interleukin 1 beta (IL-1 beta) cDNAs. Deletion of transcribed vector sequences and replacement of the native untranslated region with the AMV RNA 4 leader can result in as much as a 35-fold increase in mRNA translational efficiency; moreover, the translational efficiency of the chimaeric mRNAs containing the AMV RNA 4 leader is at least as great as that of virion RNA 4. The results suggest that the chimaeric AMV-mRNAs have either a higher relative affinity or a diminished requirement for a limiting component(s) of the translational machinery; in addition, it may be feasible, through use of heterologous leader sequences, to increase expression of engineered genes or cDNAs without changing the antigenic or biological properties of the encoded protein.

Synthesis of epidermal growth factor (EGF) receptor in vitro using SP6 RNA polymerase-transcribed template mRNA

The epidermal growth factor (EGF) receptor plays a key role in the control cellular proliferation, and its homology to the avian erythroblastosis virus erb B oncogene implicates its involvement in cellular transformation. The establishment of a correlation between the various structural domains of the EGF receptor and their functional counterparts would greatly advance our understanding of these processes. To this end, we have constructed an expression vector containing the SP6 viral promoter and an adjacent cDNA fragment encoding the full-length EGF receptor. Upon addition of SP6 RNA polymerase, this DNA is capable of generating large amounts of EGF receptor mRNA; this RNA can then be translated in vitro into immunoprecipitable EGF receptor protein. The translational efficiency of this EGF receptor RNA was found to be relatively low: approx. 100-fold lower than globin RNA synthesized using SP6 RNA polymerase. Use of these tools should now permit the synthesis and analysis of mutated EGF receptor protein in an effort to clarify the role of this receptor in growth control.

Specific and stable intron-factor interactions are established early during in vitro pre-mRNA splicing

Biochemical components (splicing factors) interact with specific intron regions during pre-mRNA splicing in vitro. The pre-mRNA specifically associates with factors at both the branch point and the 5' splice site and these RNA-factor interactions are maintained in the intron-containing RNA processing products. The first detectable event, the ATP-dependent association of a factor (or factors) with the branch point, is mediated by at least one factor containing an essential nucleic acid component. Mutant RNA substrates that lack either the 5' splice site or the vast majority of exon sequences can still associate with the branch point binding factor(s). However, this branch point-factor interaction does not occur with a mutant RNA substrate that contains the branch point but that lacks the 3' splice site consensus sequence. These results suggest that selection of the 3' splice site accompanied by the association of a factor with the branch point may be the initial step in mammalian pre-mRNA splicing.

Cap accessibility correlates with the initiation efficiency of alfalfa mosaic virus RNAs

The rate of cap removal from the alfalfa mosaic virus (A1MV) RNAs with tobacco acid pyrophosphatase (TAP) depends on the RNA species. At 37 degrees C and in the absence of divalent cation, RNA 3 reacts more slowly than the other three, which are decapped at similar rates. In the presence of magnesium, at 25 degrees C, TAP also discriminates against RNA 1. Thus the order of reactivity with TAP largely mimics the hierarchy of initiation efficiencies of the A1MV RNAs (Godefroy-Colburn et al., preceding paper in this journal). Our interpretation of these findings is that cap accessibility is what limits the rate of reaction with initiation factors as well as with TAP. In this hypothesis, translational discrimination between naturally capped messages would be related to the rate of 'breathing' of their 5' ends.

Insertion mutagenesis to increase secondary structure within the 5' noncoding region of a eukaryotic mRNA reduces translational efficiency

The thymidine kinase gene of herpes simplex virus 1 was mutated by inserting oligodeoxynucleotide linkers into the region of the gene corresponding to the 5' untranslated portion of the mRNA. These linkers, when transcribed into mRNA, might be expected to form hairpin loops and hence to increase the secondary structure of the 5' end of the mRNA. Thymidine kinase insertion derivatives were examined in vivo and in vitro to determine translational efficiency. For the in vivo studies, thymidine kinase insertion derivatives were transfected into thymidine kinase deficient L cells alone and together with a selectable dominant marker, or were assayed in the COS-1 transient expression system. For in vitro studies, thymidine kinase insertion derivatives were subcloned into pSP64. Capped transcripts were analyzed for their ability to bind ribosomes and translate in rabbit reticulocyte lysates and wheat-germ extracts. The results demonstrate that translation efficiency is decreased as the number of linkers is increased and support the view that excessive secondary structure at the 5' end of eukaryotic mRNA impedes translation.