Foot-and-mouth disease virus Lb proteinase can stimulate rhinovirus and enterovirus IRES-driven translation and cleave several proteins of cellular and viral origin.

Rhinovirus and enterovirus 2A proteinases stimulate translation initiation driven from the cognate internal ribosome entry segment (IRES) (S. J. Hambidge and P. Sarnow, Proc. Natl. Acad. Sci. USA 89:10272-10276, 1992; H.-D. Liebig, E. Ziegler, R. Yan, K. Hartmuth, H. Klump, H. Kowalski, D. Blaas, W. Sommergruber, L. Frasel, B. Lamphear, R. Rhoads, E. Kuechler, and T. Skern, Biochemistry 32:7581-7588, 1993). Given the functional similarities between the foot-and-mouth disease virus (FMDV) L proteinase and these 2A proteinases (autocatalytic excision from the nascent viral polyprotein and cleavage of eIF-4 gamma), we investigated whether the FMDV L proteinase would also be able to stimulate translation initiation. We found that purified recombinant FMDV Lb proteinase could stimulate in vitro translation driven from a rhinovirus or enterovirus IRES by 5- to 10-fold. In contrast, stimulation of translation initiation on a cardiovirus IRES by this proteinase was minimal, and stimulation of translation driven from the cognate FMDV IRES could not be evidenced. Studies using an inhibitor or a mutant Lb proteinase indicated that stimulation of IRES-driven translation is mediated via proteolysis of some cellular component(s). Our studies also demonstrated that the Lb proteinase is capable of stimulating initiation of translation on an uncapped cellular message. Unexpectedly, and in contrast to the 2A proteinases, the Lb proteinase specifically cleaved the products of the two reporter genes used in this study: Xenopus laevis cyclin B2 and influenza virus NS. Therefore, we also set out to investigate the requirements for substrate recognition by the Lb proteinase. Purified recombinant Lb proteinase recognized at least one mengovirus polypeptide and specifically cleaved human cyclin A and poliovirus replicase-related polypeptides. In the latter case, the site(s) of cleavage was located within the N-terminal part of polypeptide 3D. Sequence comparisons revealed no significant primary sequence similarities between the target proteins and the two sites already known to be recognized by the FMDV L proteinase.     

Evidence that the 5'-untranslated leader of mRNA affects the requirement for wheat germ initiation factors 4A, 4F, and 4G

The efficiency of translation of alfalfa mosaic virus (AMV) RNA 4, barley alpha-amylase (B alpha A) mRNA, and two chimeric mRNAs, AMV 4-B alpha A and B alpha A-AMV 4 (in which the 5' leader sequences of the two mRNAs were interchanged), was measured in an S30 extract from wheat germ and a fractionated system from wheat germ in which translation could be made dependent upon initiation factor (eIF) 3, 4A, 4F, or 4G. In the S30 system, AMV RNA 4 and the chimeric mRNA AMV 4-B alpha A are translated much more efficiently than B alpha A mRNA and the chimeric mRNA B alpha A-AMV 4. When the S30 system was supplemented with high amounts of purified eIF-3, eIF-4A, eIF-4F, and eIF-4G, B alpha A and B alpha A-AMV 4 mRNAs were translated as efficiently as AMV RNA 4 and AMV 4-B alpha A mRNA. These findings indicated that the mRNAs containing the B alpha A leader sequence required higher amounts of one or more of the initiation factors (eIF-3, eIF-4A, eIF-4F, and eIF-4G) for efficient translation. Determination of the amounts of the initiation factors required for translation in the fractionated system showed that AMV RNA 4 required 2-4-fold lower amounts of eIF-3, eIF-4A, eIF-4F, and eIF-4G than did B alpha A mRNA. Replacement of the B alpha A leader sequence with that of AMV RNA 4 decreased the amounts of eIF-4A, eIF-4G, and eIF-3 required, but did not affect the amount of eIF-4F required. Replacement of the AMV RNA 4 leader sequence with that of B alpha A mRNA increased the amounts of eIF-4F, eIF-4G, and eIF-3 required, but did not affect the amount of eIF-4A required. These data strongly suggest that the amounts of the factors required are affected not only by the 5' leader itself but also by interactions between the 5' leader and a region(s) of the mRNA 3' to the initiation codon.     

GRP94 (endoplasmin) co-purifies with and is phosphorylated by Golgi apparatus casein kinase

Certain picornaviruses encode proteinases which cleave the translation initiation factor eIF4G, a member of the eIF4F complex which recruits mRNA to the 40S ribosomal subunit during initiation of protein synthesis in eukaryotes. We have compared the efficiency of eIF4G cleavage in rabbit reticulocyte lysates during translation of mRNAs encoding the foot-and-mouth disease virus leader proteinase (L^pro) or the human rhinovirus 2A^pro. Under standard translation conditions, L^pro cleaved 50% of eIF4G within 4 min after initiation of protein synthesis, whereas 2A^pro required 15 min. At these times, the molar ratios of proteinase to eIF4G were 1:130 for L^pro and 1:12 for 2A^pro, indicating a much more efficient in vitro cleavage than previously observed. The molar ratios are similar to those observed during viral infection in vivo.     

Recognition of eukaryotic initiation factor 4G isoforms by picornaviral proteinases

The leader proteinase (L(pro)) of foot and mouth disease virus is a papain-like cysteine proteinase. After processing itself from the polyprotein, L(pro) then cleaves the host protein eukaryotic initiation factor (eIf) 4GI, thus preventing protein synthesis from capped mRNA in the infected cell. We have investigated L(pro) interaction with eIF4GI and its isoform, eIF4GII. L(pro), expressed as a catalytically inactive fusion protein with glutathione S-transferase, binds specifically to eIF4G isomers in rabbit reticulocyte lysates. Deletion and specific mutagenesis were used to map the binding domain on L(pro) to residues 183-195 of the C-terminal extension and to residue Cys(133). These residues of the C-terminal extension and Cys(133) are adjacent in the crystal structure but lie about 25 A from the active site. The region on eIF4GI recognized by the L(pro) C-terminal extension was mapped to residues 640-669 using eIF4GI fragments generated by proteolysis or by in vitro translation. The L(pro) cleavage site at Gly(674) downward arrow Arg(675) was not necessary for binding. Similar experiments with human rhinovirus 2A proteinase (2A(pro)), a chymotrypsin-like cysteine proteinase that also cleaves eIF4G isoforms, revealed that 2A(pro) can also bind to eIF4GI fragments lacking its cleavage site. These experiments strongly suggest a novel interaction between picornaviral proteinases and eIF4G isoforms.     

elF4G and its proteolytic cleavage products: effect on initiation of protein synthesis from capped, uncapped, and IRES-containing mRNAs.

Rhinovirus 2A and foot-and-mouth disease virus Lb proteinases stimulate the translation of uncapped messages and those carrying the rhinovirus and enterovirus Internal Ribosome Entry Segments (IRESes) by a mechanism involving the cleavage of host cell proteins. Here, we investigate this mechanism using an artificial dicistronic RNA containing the human rhinovirus IRES as intercistronic spacer. Because both proteinases cleave eukaryotic initiation factor 4G (eIF4G), we examined whether the cleavage products of eIF4G could stimulate uncapped or IRES-driven translation. Addition of intact eIF4F to translation extracts inhibited IRES-driven translation and reduced the translation stimulation observed in reactions pre-treated with Lb proteinase. Prolonged incubation of translation extracts with Lb proteinase removed all endogenous eIF4G and a substantial amount of the primary C- and N-terminal cleavage products. The translation of all mRNAs was reduced in such extracts. Capped mRNA translation was rescued by the addition of intact eIF4F. In contrast, addition of pre-cleaved eIF4F stimulated translation of uncapped or IRES-bearing messages to the levels seen upon proteinase addition. Furthermore, fractions containing the C-terminal, but not N-terminal, cleavage product of eIF4G stimulated translation moderately. These results demonstrate that the Lb and 2A proteinases stimulate translation of uncapped RNAs and those carrying IRESes by the production of cleavage products of eIF4G that enhance translation and by the removal of intact eIF4G that interferes with this stimulation.     

Eukaryotic initiation factor 4GII (eIF4GII), but not eIF4GI, cleavage correlates with inhibition of host cell protein synthesis after human rhinovirus infection

For many members of the Picornaviridae family, infection of cells results in a shutoff of host protein synthesis. For rhinoviruses and enteroviruses, the shutoff has been explained in part by the cleavage of eukaryotic initiation factor 4GI (eIF4GI), a component of the cap-binding protein complex eIF4F. The cleavage of eIF4GI is mediated by the virus-specific proteinase 2Apro and results in inhibition of cap-dependent, but not cap-independent, translation. The inhibition of host protein synthesis after infection with human rhinovirus 14 (HRV-14) lags behind the cleavage of eIF4GI. Recently, we discovered a functional homolog of eIF4GI, termed eIF4GII, and showed that cleavage of eIF4GII coincides with the shutoff of host cell protein synthesis after poliovirus infection (Gradi et al., Proc. Natl. Acad. Sci. USA 95:11089-11094, 1998). We wished to determine whether eIF4GII cleavage kinetics could also explain the lack of correlation between the kinetics of eIF4GI cleavage and the shutoff of host protein synthesis after rhinovirus infection. In this study, we examined the correlation between human rhinovirus-induced shutoff of host protein synthesis and cleavage of eIF4GI and eIF4GII. In HRV-14-infected HeLa cells, almost no intact eIF4GI could be detected by 4 h postinfection, while only 4% of eIF4GII was cleaved at this time. By 6 h, however, 67% of eIF4GII was cleaved, and this cleavage coincided with a significant (60%) decline of host translation. These results suggest that cleavage of both eIF4GI and eIF4GII is required for HRV-mediated inhibition of host cell protein synthesis and that the cleavage of eIF4GII is the rate-limiting step in the shutoff of host cell protein synthesis after rhinovirus infection.     

The eIF4G-eIF4E complex is the target for direct cleavage by the rhinovirus 2A proteinase.

The 2A proteinases (2Apro) of certain picornaviruses induce the cleavage of the eIF4G subunit of the cap-binding protein complex, eIF4F. Several reports have demonstrated that 2Apro of rhinovirus and coxsackievirus B4 cleave eIF4G directly. However, it was suggested that in poliovirus infection, the 2Apro induces the activation of a cellular proteinase which in turn cleaves eIF4G. Furthermore, it is not clear whether eIF4G is cleaved as part of the eIF4F complex or as an individual polypeptide. To address these issues, recombinant eIF4G was purified from Sf9 insect cells and tested for cleavage by purified rhinovirus 2Apro. Here we report that eIF4G alone is a relatively poor substrate for cleavage by the rhinovirus 2Apro. However, an eIF4G-eIF4E complex is cleaved efficiently by the 2Apro, suggesting that eIF4F is a preferred substrate for cleavage by rhinovirus 2Apro. Furthermore, 2Apr drastically reduced the translation of a capped mRNA. An eIF4G-eIF4E complex, but not eIF4G alone, was required to restore translation.     

2A proteinase of human rhinovirus cleaves cytokeratin 8 in infected HeLa cells

Rhino- and enteroviruses encode two proteinases, 2A and 3C, which are responsible for the processing of the viral polyprotein and for cleavage of several cellular proteins. To identify further targets of the 2A proteinase of human rhinovirus serotype 2 (HRV2), an in vitro cleavage assay followed by two-dimensional electrophoresis was employed. Cytokeratin 8, a member of the intermediate filament group of proteins, was found to be proteolytically cleaved in vitro by the 2A proteinase of HRV2 and of coxsackievirus B4 and in vivo during HRV2 infection of HeLa cells. The cleavage results in removal of 14 amino acids from the N-terminal head domain of cytokeratin 8. However, other intermediate filament proteins (cytokeratins 7 and 18 and vimentin) were not cleaved in the course of the HRV2 infection. Compared with the processing of the eucaryotic translation initiation factors 4GI and 4GII, cleavage of cytokeratin 8 occurs late in the infection cycle at the time of the onset of the cytopathic effect.     

Primary structure of bovine endothelin ETB receptor and identification of signal peptidase and metal proteinase cleavage sites.

A cDNA clone corresponding to the entire coding region of the bovine ETB endothelin receptor mRNA was isolated from a lung cDNA library and sequenced. The cDNA encodes 441 amino acids: 26 constituting an NH2-terminal signal peptide and 415 constituting the mature receptor. The signal peptidase cleavage site was determined by direct amino acid sequencing of purified receptor. A comparison of the predicted amino acid sequence with the available bovine ETA and rat ETB endothelin receptor sequences revealed 63 and 85% homology, respectively. Endothelin receptors of various species are known to be very sensitive to a certain metal proteinase(s) and have been shown to be converted to a lower Mr form in the absence of EDTA. The metal proteinase cleavage site was also determined by direct protein sequencing of the proteolysis product. The amino acid sequence (Ala-Gly-X-Pro-Pro-Arg) surrounding the cleavage site (between Ala-79 and Gly-80) is conserved among the ETB endothelin receptors, explaining the above mentioned proteolytic conversion from the higher to lower Mr forms observed in various species.     

Cap-independent translation initiation in Xenopus oocytes.

Eukaryotic cellular mRNAs contain a cap at their 5'-ends, but some viral and cellular mRNAs bypass the cap-dependent mechanism of translation initiation in favor of internal entry of ribosomes at specific RNA sequences. Cap-dependent initiation requires intact initiation factor eIF4G (formerly eIF-4gamma, eIF-4Fgamma or p220), whereas internal initiation can proceed with eIF4G cleaved by picornaviral 2A or L proteases. Injection of recombinant coxsackievirus B4 protease 2A into Xenopus oocytes led to complete cleavage of endogenous eIF4G, but protein synthesis decreased by only 35%. Co-injection of edeine reduced synthesis by >90%, indicating that eIF4G-independent synthesis involved ongoing initiation. The spectrum of endogenous proteins synthesized was very similar in the presence or absence of intact eIF4G. Translation of exogenous rabbit globin mRNA, by contrast, was drastically inhibited by eIF4G cleavage. The N-terminal cleavage product of eIF4G (cpN), which binds eIF4E, was completely degraded within 6-12 h, while the C-terminal cleavage product (cpC), which binds to eIF3 and eIF4A, was more stable over the same period. Thus, translation initiation of most endogenous mRNAs inXenopusoocytes requires no eIF4G, or perhaps only cpC, suggesting a cap-independent mechanism.     

Eukaryotic initiation factor 4GI is a poor substrate for HIV-1 proteinase

Eukaryotic initiation factor (eIF) 4GI is efficiently cleaved during picornaviral replication. eIF4GI processing has also recently been observed during HIV-1 replication. We have compared the efficiency of eIF4GI proteolysis in rabbit reticulocyte lysates during translation of mRNAs encoding the foot-and-mouth disease virus leader proteinase (L(pro)) or the HIV-1 proteinase (HIV-1(pro)). L(pro) cleaved 50% eIF4GI within 12 min whereas HIV-1(pro) required 4 h; the concentrations were 2 pg/microl (0.1 nM) for L(pro) and 60 pg/microl (2.66 nM) for HIV-1(pro). HIV-1(pro) processing of eIF4GI is therefore not quantitatively analogous to that of L(pro), suggesting that the primary function of eIF4GI cleavage in HIV-1 replication may not be protein synthesis inhibition.     

Translation by the adenovirus tripartite leader: elements which determine independence from cap-binding protein complex.

The adenovirus tripartite leader is a 200-nucleotide-long 5' noncoding region which facilitates translation of viral mRNAs at late times after infection. The tripartite leader also confers the ability to initiate translation independent of the requirement for cap-binding protein complex or eIF-4F without any requirement for adenovirus gene products. To elucidate the manner by which the tripartite leader functions, the primary determinants of leader activity were investigated in vivo by testing a series of mutations expressed from transfected plasmids. The results of these experiments indicate that the tripartite leader does not promote internal ribosome binding, at least in a manner recently described for picornavirus mRNAs. In addition, despite an unusual arrangement of sequences complementary to the 3' end of 18S rRNA in the tripartite leader, we could find no evidence for involvement in its translation activity. Instead, our results are consistent with a model in which much of the first leader is maintained in an unstructured conformation which determines the ability of the tripartite leader to facilitate translation and bypass a normal requirement for eIF-4F activity. Several possible translation models are discussed, as well as the implications for translation of late viral mRNAs.     

The foot-and-mouth disease virus leader proteinase gene is not required for viral replication.

The foot-and-mouth disease virus (FMDV) leader (L) proteinase has only two known functions: (i) autocatalytic removal from the N terminus of the viral polyprotein and (ii) cleavage of the p220 subunit of the eukaryotic initiation factor 4F complex, which helps to shut off host protein synthesis. Cleavage of p220 appears to be important for picornavirus replication, since rhinoviruses and enteroviruses utilize a different proteinase (2A) to cleave p220. To explore the role of L in FMDV replication, we generated synthetic FMDV genomes lacking the L gene and tested their viability in cells. Genomes were constructed with the N-terminal Gly codon of VP4 positioned directly following either the first (Lab) or second (Lb) Met codon of the L protein. Cells transfected with synthetic RNAs lacking L and initiating with the Lab Met codon failed to produce viable virus, but cells transfected with RNAs that utilized the second AUG to drive translation of the viral polyprotein produced viable viruses. These leader-deleted viruses produced plaques on BHK cells that were slightly smaller than those produced by wild-type (WT) virus, grew to slightly lower titers than WT virus in BHK cells, shut off host protein synthesis more slowly than WT virus, and were slightly attenuated in mice. These studies indicate that the L proteinase is not essential for FMDV replication and show that in the cells and animals tested the L gene has a limited effect on virus replication.     

The translation initiation factor eIF-4E binds to a common motif shared by the translation factor eIF-4 gamma and the translational repressors 4E-binding proteins.

Eukaryotic translation initiation factor 4E (eIF-4E), which possesses cap-binding activity, functions in the recruitment of mRNA to polysomes as part of a three-subunit complex, eIF-4F (cap-binding complex). eIF-4E is the least abundant of all translation initiation factors and a target of growth regulatory pathways. Recently, two human cDNAs encoding novel eIF-4E-binding proteins (4E-BPs) which function as repressors of cap-dependent translation have been cloned. Their interaction with eIF-4E is negatively regulated by phosphorylation in response to cell treatment with insulin or growth factors. The present study aimed to characterize the molecular interactions between eIF-4E and the other subunits of eIF-4F and to similarly characterize the molecular interactions between eIF-4E and the 4E-BPs. A 49-amino-acid region of eIF-4 gamma, located in the N-terminal side of the site of cleavage by Picornaviridae protease 2A, was found to be sufficient for interacting with eIF-4E. Analysis of deletion mutants in this region led to the identification of a 12-amino-acid sequence conserved between mammals and Saccharomyces cerevisiae that is critical for the interaction with eIF-4E. A similar motif is found in the amino acid sequence of the 4E-BPs, and point mutations in this motif abolish the interaction with eIF-4E. These results shed light on the mechanisms of eIF-4F assembly and on the translational regulation by insulin and growth factors.     

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.     

Cleavage of eukaryotic translation initiation factor 4GII within foot-and-mouth disease virus-infected cells: identification of the L-protease cleavage site in vitro

Foot-and-mouth disease virus (FMDV) induces a very rapid inhibition of host cell protein synthesis within infected cells. This is accompanied by the cleavage of the eukaryotic translation initiation factor 4GI (eIF4GI). The cleavage of the related protein eIF4GII has now been analyzed. Within FMDV-infected cells, cleavage of eIF4GI and eIF4GII occurs with similar kinetics. Cleavage of eIF4GII is induced in cells and in cell extracts by the FMDV leader protease (L(pro)) alone, generating cleavage products similar to those induced by enterovirus and rhinovirus 2A protease (2A(pro)). By the use of a fusion protein containing residues 445 to 744 of human eIF4GII, it was demonstrated that the FMDV L(pro) specifically cleaves this protein between residues G700 and S701, immediately adjacent to the site (V699/G700) cleaved by rhinovirus 2A(pro) in vitro. The G700/S701 cleavage site does not correspond, by amino acid sequence alignment, to that cleaved in eIF4GI by the FMDV L(pro) in vitro. Knowledge of the cleavage sites and the three-dimensional structures of the FMDV L(pro) and rhinovirus 2A(pro) enabled mutant forms of the eIF4GII sequence to be generated that are differentially resistant to either one of these proteases. These results confirmed the specificity of each protease and showed that the mutant forms of the fusion protein substrate retained their correct sensitivity to other proteases.     

Amino acid sequence of the human protein synthesis initiation factor eIF-4 gamma.

Eukaryotic protein synthesis initiation factor (eIF) 4 gamma, also known as p220, is a component of the protein complex eIF-4, which is involved in the recognition of the mRNA cap, ATP-dependent unwinding of 5'-terminal secondary structure and recruitment of mRNA to the ribosome. Peptide sequence data from rabbit reticulocyte eIF-4 gamma was used to synthesize oligonucleotide probes and polymerase chain reaction primers. These were used to screen lambda-cDNA libraries from rabbit and human brain, yielding a partial rabbit and a complete human cDNA sequence of 5.1 kilobases. Northern blot and primer extension analysis indicated that the cDNA sequence was complete. To confirm that the cDNA represented that of eIF-4 gamma, three peptides were synthesized based on cDNA sequences and used to produce anti-peptide antibodies. The antibodies specifically recognized intact eIF-4 gamma and its cleavage products following poliovirus infection. The eIF-4 gamma mRNA contains AUG codons at nucleotides 6, 67, 90, 165, and 369, but only the last is followed by a long open reading frame. The eIF-4 gamma polypeptide is 154 kDa (1396 amino acid residues) and contains sequence motifs of potential interest: a sequence (AGLGPR) that is similar to the substrate recognition sequence of protease 2A from rhinovirus serotype 14, five PEST regions with scores greater than 10, which are characteristic of rapidly degraded proteins, stretches of polyglutamic acid, and numerous potential phosphorylation sites.     

Hybrid proteins between Pseudomonas aeruginosa exotoxin A and poliovirus 2Apro cleave p220 in HeLa cells.

Cleavage of p220, a component of the initiation factor eIF-4F, has been correlated with the inhibition of host translation during poliovirus infection. To obtain p220 cleavage in the absence of any other poliovirus gene products, hybrid proteins containing Pseudomonas aeruginosa exotoxin A and poliovirus protease 2Apro have been constructed. The addition of the hybrid molecules to cultured cells did not lead to substantial p220 cleavage. However, the simultaneous presence of the hybrid toxin with replicationally inactive chicken adenovirus particles results in efficient cleavage of p220 in the intact cells. Under these conditions, cellular translation continues unabated for several hours, arguing against a direct requirement for intact p220 in each round of the initiation of translation of cellular mRNAs.     

Dominant negative mutants of mammalian translation initiation factor eIF-4A define a critical role for eIF-4F in cap-dependent and cap-independent initiation of translation.

Eukaryotic translation initiation factor-4A (eIF-4A) plays a critical role in binding of eukaryotic mRNAs to ribosomes. It has been biochemically characterized as an RNA-dependent ATPase and RNA helicase and is a prototype for a growing family of putative RNA helicases termed the DEAD box family. It is required for mRNA-ribosome binding both in its free form and as a subunit of the cap binding protein complex, eIF-4F. To gain further understanding into the mechanism of action of eIF-4A in mRNA-ribosome binding, defective eIF-4A mutants were tested for their abilities to function in a dominant negative manner in a rabbit reticulocyte translation system. Several mutants were demonstrated to be potent inhibitors of translation. Addition of mutant eIF-4A to a rabbit reticulocyte translation system strongly inhibited translation of all mRNAs studied including those translated by a cap-independent internal initiation mechanism. Addition of eIF-4A or eIF-4F relieved inhibition of translation, but eIF-4F was six times more effective than eIF-4A, whereas eIF-4B or other translation factors failed to relieve the inhibition. Kinetic experiments demonstrated that mutant eIF-4A is defective in recycling through eIF-4F, thus explaining the dramatic inhibition of translation. Mutant eIF-4A proteins also inhibited eIF-4F-dependent, but not eIF-4A-dependent RNA helicase activity. Taken together these results suggest that eIF-4A functions primarily as a subunit of eIF-4F, and that singular eIF-4A is required to recycle through the complex during translation. Surprisingly, eIF-4F, which binds to the cap structure, appears to be also required for the translation of naturally uncapped mRNAs.