Altered interactions between the A1 and A2 subunits of factor VIIIa following cleavage of A1 subunit by factor Xa

Factor VIIIa consists of subunits designated A1, A2, and A3-C1-C2. The limited cofactor activity observed with the isolated A2 subunit is markedly enhanced by the A1 subunit. A truncated A1 (A1(336)) was previously shown to possess similar affinity for A2 and retain approximately 60% of its A2 stimulatory activity. We now identify a second site in A1 at Lys(36) that is cleaved by factor Xa. A1 truncated at both cleavage sites (A1(37-336)) showed little if any affinity for A2 (K(d)>2 microm), whereas factor VIIIa reconstituted with A2 plus A1(37-336)/A3-C1-C2 dimer demonstrated significant cofactor activity ( approximately 30% that of factor VIIIa reconstituted with native A1) in a factor Xa generation assay. These affinity values were consistent with values obtained by fluorescence energy transfer using acrylodan-labeled A2 and fluorescein-labeled A1. In contrast, factor VIIIa reconstituted with A1(37-336) showed little activity in a one-stage clotting assay. This resulted in part from a 5-fold increase in K(m) for factor X when A1 was cleaved at Arg(336). These findings suggest that both A1 termini are necessary for functional interaction of A1 with A2. Furthermore, the C terminus of A1 contributes to the K(m) for factor X binding to factor Xase, and this parameter is critical for activity assessed in plasma-based assays.     

Different binding sites for the neuropeptide Y Y1 antagonists 1229U91 and J-104870 on human Y1 receptors

The peptidic Y1 antagonist 1229U91 and the non-peptidic antagonist J-104870 have high binding affinities for the human Y1 receptor. These Y1 antagonists show anorexigenic effects on NPY-induced feeding in rats, although they have completely different structures and molecular sizes. To identify the binding sites of these ligands, we substituted amino acid residues of the human Y1 receptor with alanine and examined the abilities of the mutant receptors to bind the radio-labeled ligands. Alanine substitutions, F98A, D104A, T125A, D200A, D205A, L215A, Q219A, L279A, F282A, F286A, W288A and H298A, in the human Y1 receptor lost their affinity for the peptide agonist PYY, but not for 1229U91 and J-104870, while L303A and F173A lost affinity for 1229U91 and J-104870, respectively. N283A retained its affinity for 1229U91, but not for PYY and J-104870. Y47A and N299A retained their affinity for J-104870, but not for PYY and 1229U91. W163A and D287A showed no affinity for any of the three ligands. Taken together, these data indicate that the binding sites of 1229U91 are widely located in the shallow region of the transmembrane (TM) domain of the receptor, especially TM1, TM6 and TM7. In contrast, J-104870 recognized the pocket formed by TM4, TM5 and TM6, based on the molecular modeling of the Y1 receptor and J-104870 complex. In conclusion, 1229U91 and J-104870 have high affinities for Y1 receptors using basically different binding sites. D287 of the common binding site in the TM6 domain could be crucial for the binding of Y1 antagonists.     

Identification of coagulation factor VIII A2 domain residues forming the binding epitope for low-density lipoprotein receptor-related protein

Regulation of the coagulation factor VIII (fVIII) level in circulation involves a hepatic receptor low-density lipoprotein receptor-related protein (LRP). One of two major LRP binding sites in fVIII is located within the A2 domain (A2), likely exposed within the fVIII complex with von Willebrand factor and contributing to regulation of fVIII via LRP. This work aimed to identify A2 residues forming its LRP-binding site, previously shown to involve residues 484-509. Isolated A2 was subjected to alanine-scanning mutagenesis followed by expression of a set of mutants in a baculovirus system. In competition and surface plasmon resonance assays, affinities of A2 mutants K466A, R471A, R484A, S488A, R489A, R490A, H497A, and K499A for LRP were found to be decreased by 2-4-fold. This correlated with 1.3-1.5-fold decreases in the degree of LRP-mediated internalization of the mutants in cell culture. Combining these mutations into pairs led to cumulative effects, i.e., 7-13-fold decrease in affinity for LRP and 1.6-2.2-fold decrease in the degree of LRP-mediated internalization in cell culture. We conclude that the residues mentioned above play a key role in formation of the A2 binding epitope for LRP. Experiments in mice revealed an approximately 4.5 times shorter half-life for A2 in the circulation in comparison with that of fVIII. The half-lives of A2 mutant R471A/R484A or A2 co-injected with receptor-associated protein, a classical ligand of LRP, were prolonged by approximately 1.9 and approximately 3.5 times, respectively, compared to that of A2. This further confirms the importance of the mutated residues for interaction of A2 with LRP and suggests the existence of an LRP-dependent mechanism for removing A2 as a product of dissociation of activated fVIII from the circulation.     

HLA antigens and clinical subgroups of schizophrenia

Frequencies of HLA A, B, C, and DR antigens were studied in 100 schizophrenic patients and 919 controls from South Sweden. The patients were diagnosed according to the DSM III criteria and divided into four clinical subgroups (hebephrenic, paranoid, residual, and undifferentiated). In the schizophrenic patients as a whole significant increases were found for A2, A3, B17, B27, and Cw2 and decreases for A1, A11, and B8. A previous positive association with A9 from the same population was not confirmed. A significant heterogeneity between the four clinical subgroups was found for A3 and Bw35. Most of the associations between HLA antigens and schizophrenia reported in the literature appear to be fortuitous and dependent on the large number of trials made. However, confirmed increases have been found for A9 and B17, and confirmed decreases have been observed for A1 and B7. Some evidence for a heterogeneity between clinical subgroups was found in the present as well as in previous investigations.     

Cytotoxicity of polyspermine-ribonuclease A and polyspermine-dimeric ribonuclease A

Polyspermine-ribonuclease A (PS-RNase A) and polyspermine-dimeric ribonuclease A (PS-dimeric RNase A) were prepared by cross-linking ribonuclease A or its covalently linked dimer to polyspermine (PS) using dimethyl suberimidate. The two RNase A derivatives were tested for a possible antitumor action. The in vitro and in vivo cytotoxic activity of PS-RNase A, although strong, is not higher than that known for free polyspermine. PS-dimeric RNase A, which was characterized by mass spectroscopy, titration of free amine groups, and enzymatic assays, proved instead to be a definitely more efficient antitumor agent, both in vitro and in vivo. This result could tentatively be explained in view of the importance of positive charges for ribonuclease activity, considering the higher basicity of PS-dimeric RNase A compared to that of PS-(monomeric)RNase A. It must be also taken into account that the dimeric RNase A moiety of PS-dimeric RNase A could evade the cytoplasmic ribonuclease inhibitor, which instead could trap the monomeric RNase A moiety of the other derivative. The two RNase A derivatives degrade poly(A).poly(U) under conditions where native RNase A is inactive. The results of this work demonstrate once again the importance of positive charges for the functions of mammalian pancreatic type ribonucleases in general, in particular for RNase A derivatives, and the potential therapeutic use of the ribonuclease A derivatives.     

Ebola virus glycoprotein 1: identification of residues important for binding and postbinding events

The filoviruses Ebola virus (EBOV) and Marburg virus (MARV) are responsible for devastating hemorrhagic fever outbreaks. No therapies are available against these viruses. An understanding of filoviral glycoprotein 1 (GP1) residues involved in entry events would facilitate the development of antivirals. Towards this end, we performed alanine scanning mutagenesis on selected residues in the amino terminus of GP1. Mutant GPs were evaluated for their incorporation onto feline immunodeficiency virus (FIV) particles, transduction efficiency, receptor binding, and ability to be cleaved by cathepsins L and B. FIV virions bearing 39 out of 63 mutant glycoproteins transduced cells efficiently, whereas virions bearing the other 24 had reduced levels of transduction. Virions pseudotyped with 23 of the poorly transducing GPs were characterized for their block in entry. Ten mutant GPs were very poorly incorporated onto viral particles. Nine additional mutant GPs (G87A/F88A, K114A/K115A, K140A, G143A, P146A/C147A, F153A/H154A, F159A, F160A, and Y162A) competed poorly with wild-type GP for binding to permissive cells. Four of these nine mutants (P146A/C147A, F153A/H154A, F159A, and F160A) were also inefficiently cleaved by cathepsins. An additional four mutant GPs (K84A, R134A, D150A, and E305/E306A) that were partially defective in transduction were found to compete effectively for receptor binding and were readily cleaved by cathepsins. This finding suggested that this latter group of mutants might be defective at a postbinding, cathepsin cleavage-independent step. In total, our study confirms the role of some GP1 residues in EBOV entry that had previously been recognized and identifies for the first time other residues that are important for productive entry.     

Description of cellobiohydrolases Ce16A and Ce17A fromTrichoderma reesei using Langmuir-type models

The binding of cellobiohydrolases to cellulose is a crucial initial step in cellulose hydrolysis. In the search for a detailed understanding of the function of cellobiohydrolases, much information concerning how the enzymes and their constituent catalytic and cellulose-binding changes during hydrolysis is still needed. The adsorption of purffied two cellobiohydrolases (Ce17A and Ce16A) fromTrichoderma reesei cellulase to microcrystalline cellulose has been studied. Cellobiohydrolase II (Ce16A) does not affect the adsorption of cellobiohydrolase I (Ce17A) significantly, and there are specific binding sites for both Ce17A and Ce16A. The adsorption affinity and tightness of the cellulase binding domain (CBD) for Ce17A are larger that those of the CBD for Ce16A. The CBD for Ce17A binds more rapidly and tightly to Avicel than the CBD for Ce16A. The decrease in adsorption observed when the two cellobihydrolases are studied together would appear to be the result of competition for binding sites on the cellulose. Ce17A competes more efficiently for binding sites than Ce16A. Competition for binding sites is the dominating factor when the two enzymes are acting together, furthermore adsorption to sites specific for Ce17A and Ce16A, also contributes to the total adsorption.     

Mapping of sites on the surface membrane of mammalian cells

Summary Cells transformed by Simian Virus 40 have sites on the surface membrane for Concanavalin A (Con A) and a copolymer of ornithine, leucine (POL). The cells can be rapidly agglutinated by Con A, more slowly aggregated by POL, and they can be killed by both compounds. Treatment with Con A or POL has been used to select resistant cell variants from the transformed cells. Variants selected for resistance to Con A were also resistant to POL, but variants selected for resistance to POL were not resistant to Con A. The POL-selected variants showed less aggregation by POL but no decrease in agglutinability by Con A, whereas Con A-selected variants showed a decrease both in POL aggregation and Con A agglutination. The selection for both sites by Con A and only for POL sites by POL, can be explained in that the sites for POL are part of the sites for Con A and/or are included in clusters of Con A sites.Paper I in this series is Inbar, Ben-Bassat and Sachs (1971a).     

Differential requirement for conserved tryptophans in human immunodeficiency virus type 1 Vif for the selective suppression of APOBEC3G and APOBEC3F

APOBEC3G (A3G) and related cytidine deaminases, such as APOBEC3F (A3F), are potent inhibitors of retroviruses. Formation of infectious human immunodeficiency virus type 1 (HIV-1) requires suppression of multiple cytidine deaminases by Vif. Whether HIV-1 Vif recognizes various APOBEC3 proteins through a common mechanism is unclear. The domains in Vif that mediate APOBEC3 recognitions are also poorly defined. The N-terminal region of HIV-1 Vif is unusually rich in Trp residues, which are highly conserved. In the present study, we examined the role of these Trp residues in the suppression of APOBEC3 proteins by HIV-1 Vif. We found that most of the highly conserved Trp residues were required for efficient suppression of both A3G and A3F, but some of these residues were selectively required for the suppression of A3F but not A3G. Mutant Vif molecules in which Ala was substituted for Trp79 and, to a lesser extent, for Trp11 remained competent for A3G interaction and its suppression; however, they were defective for A3F interaction and therefore could not efficiently suppress the antiviral activity of A3F. Interestingly, while the HIV-1 Vif-mediated degradation of A3G was not affected by the different C-terminal tag peptides, that of A3F was significantly influenced by its C-terminal tags. These data indicate that the mechanisms by which HIV-1 Vif recognizes its target molecules, A3G and A3F, are not identical. The fact that several highly conserved residues in Vif are required for the suppression of A3F but not that of A3G suggests a critical role for A3F in the restriction of HIV-1 in vivo.     

Analysis of the substrate specificity of human sulfotransferases SULT1A1 and SULT1A3: site-directed mutagenesis and kinetic studies.

Sulfonation is an important metabolic process involved in the excretion and in some cases activation of various endogenous compounds and xenobiotics. This reaction is catalyzed by a family of enzymes named sulfotransferases. The cytosolic human sulfotransferases SULT1A1 and SULT1A3 have overlapping yet distinct substrate specificities. SULT1A1 favors simple phenolic substrates such as p-nitrophenol, whereas SULT1A3 prefers monoamine substrates such as dopamine. In this study we have used a variety of phenolic substrates to functionally characterize the role of the amino acid at position 146 in SULT1A1 and SULT1A3. First, the mutation A146E in SULT1A1 yielded a SULT1A3-like protein with respect to the Michaelis constant for simple phenols. The mutation E146A in SULT1A3 resulted in a SULT1A1-like protein with respect to the Michaelis constant for both simple phenols and monoamine compounds. When comparing the specificity of SULT1A3 toward tyramine with that for p-ethylphenol (which differs from tyramine in having no amine group on the carbon side chain), we saw a 200-fold preference for tyramine. The kinetic data obtained with the E146A mutant of SULT1A3 for these two substrates clearly showed that this protein preferred substrates without an amine group attached. Second, changing the glutamic acid at position 146 of SULT1A3 to a glutamine, thereby neutralizing the negative charge at this position, resulted in a 360-fold decrease in the specificity constant for dopamine. The results provide strong evidence that residue 146 is crucial in determining the substrate specificity of both SULT1A1 and SULT1A3 and suggest that there is a direct interaction between glutamic acid 146 in SULT1A3 and monoamine substrates.     

A central region in the hepatitis C virus NS4A protein allows formation of an active NS3-NS4A serine proteinase complex in vivo and in vitro.

A virus-encoded serine proteinase mediates four site-specific cleavages in the hepatitis C virus polyprotein. In addition to the catalytic domain, which is located in the N-terminal one-third of nonstructural protein NS3, the 54-residue NS4A protein is required for cleavage at some but not all sites. Here, we provide evidence for a non-ionic detergent-stable interaction between NS4A and the NS3 serine proteinase domain and demonstrate that the central region of NS4A plays a key role in NS4A-dependent processing. Hydrophobic residues, in particular Ile-29, were shown to be important for NS4A activity, and a synthetic peptide, spanning NS4A residues 22 to 34, could substitute for intact NS4A in a cell-free trans cleavage assay. Furthermore, NS4A mutations, which abolished or inhibited processing, correlated with destabilization of the NS3-NS4A complex. These results suggest that a stable interaction exists between the central region of NS4A and the NS3 catalytic domain which is required for NS4A-dependent processing. Since NS4A is required for processing at certain serine proteinase-dependent cleavage sites, this interaction may represent a new target for development of antiviral compounds.     

Structural characterization of the homotropic cooperative binding of azamulin to human cytochrome P450 3A5

Cytochrome P450 3A4 and 3A5 catalyze the metabolic clearance of a large portion of therapeutic drugs. Azamulin is used as a selective inhibitor for 3A4 and 3A5 to define their roles in metabolism of new chemical entities during drug development. In contrast to 3A4, 3A5 exhibits homotropic cooperativity for the sequential binding of two azamulin molecules at concentrations used for inhibition. To define the underlying sites and mechanisms for cooperativity, an X-ray crystal structure of 3A5 was determined with two azamulin molecules in the active site that are stacked in an antiparallel orientation. One azamulin resides proximal to the heme in a pose similar to the 3A4–azamulin complex. Comparison to the 3A5 apo structure indicates that the distal azamulin in 3A5 ternary complex causes a significant induced fit that excludes water from the hydrophobic surfaces of binding cavity and the distal azamulin, which is augmented by the stacking interaction with the proximal azamulin. Homotropic cooperativity was not observed for the binding of related pleuromutilin antibiotics, tiamulin, retapamulin, and lefamulin, to 3A5, which are larger and unlikely to bind in the distal site in a stacked orientation. Formation of the 3A5 complex with two azamulin molecules may prevent time-dependent inhibition that is seen for 3A4 by restricting alternate product formation and/or access of reactive intermediates to vulnerable protein sites. These results also contribute to a better understanding of sites for cooperative binding and the differential structural plasticity of 3A5 and 3A4 that contribute to differential substrate and inhibitor binding.     

The A1 and A2 subunits of factor VIIIa synergistically stimulate factor IXa catalytic activity.

Factor VIIIa, the protein cofactor for factor IXa, is comprised of A1, A2, and A3-C1-C2 subunits. Recently, we showed that isolated A2 subunit enhanced the kcat for factor IXa-catalyzed activation of factor X by approximately 100-fold ( approximately 1 min-1), whereas isolated A1 or A3-C1-C2 subunits showed no effect on this rate (Fay, P. J., and Koshibu, K. J. (1998) J. Biol. Chem. 273, 19049-19054). However, A1 subunit increased the A2-dependent stimulation by approximately 10-fold. The Km for factor X in the presence of A2 subunit was unaffected by A1 subunit, whereas the kcat observed in the presence of saturating A1 and A2 subunits ( approximately 15 min-1) represented 5-10% of the value observed for native factor VIIIa (approximately 200 min-1). An anti-A1 subunit antibody that blocks the association of A2 eliminated the A1-dependent contribution to factor IXa activity. Inclusion of both A1 and A2 subunits resulted in greater increases in the fluorescence anisotropy of fluorescein-Phe-Phe-Arg factor IXa than that observed for A2 subunit alone and approached values obtained with factor VIIIa. These results indicate that A1 subunit alters the A2 subunit-dependent modulation of the active site of factor IXa to synergistically increase cofactor activity, yielding an overall increase in kcat of over 1000-fold compared with factor IXa alone.     

Phenylethyl-substituted pyrimido[2,1-f]purinediones and related compounds: structure-activity relationships as adenosine A(1) and A(2A) receptor ligands

The synthesis of N-(un)substituted-phenylalkylpyrimido[2,1-f]purinediones was performed starting with 7-(3-chloropropyl)-8-bromotheophylline and 7-(3-chloropropyl)-8-bromo-1,3-dipropylxanthine. Compounds with unsubstituted or substituted ethylene spacer to an aromatic ring were synthesized. Additionally variations in the spacer-elongation of the linker containing more than two atoms, introduction of a double bond or heteroatoms were performed. Physicochemical properties of the synthesized compounds were described. The obtained compounds envisaged as sterically fixed and configurationally stable analogs of 8-styrylxanthines, were evaluated for their affinity to adenosine A(1) and A(2A) receptors, the receptor subtypes that are predominant in the brain. Selected compounds were also investigated for the affinity to the A(2B) and A(3) receptor subtypes. It was stated that phenylethyl pyrimido[2,1-f]purinediones and their analogs with variations of the ethylene spacer (substituted or extended) exhibit micromolar or submicromolar affinity for A(2A) ARs (adenosine receptors); for example compound 2Ac with p-hydroxy substituent displayed a K(i) value of 0.23 microM at the rat A(2A) receptor. In comparison to the previously obtained phenyl and benzyl pyrimido[2,1-f]purinediones compounds with a shorter spacer, phenethyl derivatives were optimal for A(2A) AR. The kind of substituent at the aromatic ring was important for the affinity. Oxygen and nitrogen atoms in the spacer resulted frequently in a slight decrease of the A(2A) AR affinity, introduction of more heteroatoms into the spacer-in carbamates-caused distinctly negative effect on the activity. In this series of compounds more frequently the adenosine A(1) activity was observed, also in submicromolar range as for dipropyl derivative 2Ba with K(i) value of 0.62 microM at the rat A(2A) AR. 3D-QSAR models were developed for the compounds presented in this paper as well as in the previous publications showing activity at adenosine A(1) and A(2A) ARs. It was concluded that for the activity at adenosine A(1) and A(2A) receptors lipophilicity, steric effects along with the molecule's electrostatic surface properties had greatest value. Chosen compounds were evaluated in vivo as anticonvulsants in MES, scMet tests and examined for neurotoxicity. Contrary to previously obtained phenyl and benzyl pyrimido[2,1-f]purinediones, all tested compounds were inactive as anticonvulsants.     

Interaction of isozymes of myosin subfragment 1 with actin: effect of ionic strength and nucleotide

Myosin subfragment 1 (S-1) can be fractionated into two isozymes, (A1)S-1 containing alkali light chain 1 and (A2)S-1 containing alkali light chain 2. The predominant difference in the behavior of the two isozymes of S-1 is that, at low ionic strength, the actin concentration required for half-maximal ATPase activity is considerably lower for (A1)S-1 than for (A2)S-1; that is, the apparent binding constant KATPase for (A1)S-1 is greater than KATPase for (A2)S-1 [Weeds, A.G., & Taylor, R.S. (1975) Nature (London) 257, 54-56]. This difference disappears at high ionic strength [Wagner, P. D., Slater, C. S., Pope, B., & Weeds, A.G. (1979) Eur. J. Biochem. 99, 385-394]. In the present study we investigated whether the difference in the KATPase values of (A1)S-1 and (A2)S-1 is due to a difference in the actual affinity of these S-1 isozymes for actin. Binding was measured in the presence of ATP and AMP-PNP and in the absence of nucleotide at varied ionic strengths. We found that at low ionic strength where KATPase is several times stronger for (A1)S-1 than for (A2)S-1, the binding of (A1)S-1 to actin is correspondingly stronger than that of (A2)S-1 irrespective of the nucleotide present. Furthermore, as the ionic strength is increased, just as the difference between the KATPase values for (A1)S-1 and (A2)S-1 disappears so too does the difference in the affinity of the two isozymes for actin.(ABSTRACT TRUNCATED AT 250 WORDS)     

Ligand binding to heme proteins. VI. Interconversion of taxonomic substates in carbonmonoxymyoglobin.

The kinetic properties of the three taxonomic A substates of sperm whale carbonmonoxy myoglobin in 75% glycerol/buffer are studied by flash photolysis with monitoring in the infrared stretch bands of bound CO at nu(A0) approximately 1967 cm-1, nu(A1) approximately 1947 cm-1, and nu(A3) approximately 1929 cm-1 between 60 and 300 K. Below 160 K the photodissociated CO rebinds from the heme pocket, no interconversion among the A substates is observed, and rebinding in each A substate is nonexponential in time and described by a different temperature-independent distribution of enthalpy barriers with a different preexponential. Measurements in the electronic bands, e.g., the Soret, contain contributions of all three A substates and can, therefore, be only approximately modeled with a single enthalpy distribution and a single preexponential. The bond formation step at the heme is fastest for the A0 substate, intermediate for the A1 substate, and slowest for A3. Rebinding between 200 and 300 K displays several processes, including geminate rebinding, rebinding after ligand escape to the solvent, and interconversion among the A substates. Different kinetics are measured in each of the A bands for times shorter than the characteristic time of fluctuations among the A substates. At longer times, fluctuational averaging yields the same kinetics in all three A substates. The interconversion rates between A1 and A3 are determined from the time when the scaled kinetic traces of the two substates merge. Fluctuations between A1 and A3 are much faster than those between A0 and either A1 or A3, so A1 and A3 appear as one kinetic species in the exchange with A0. The maximum-entropy method is used to extract the distribution of rate coefficients for the interconversion process A0 <--> A1 + A3 from the flash photolysis data. The temperature dependencies of the A substate interconversion processes are fitted with a non-Arrhenius expression similar to that used to describe relaxation processes in glasses. At 300 K the interconversion time for A0 <--> A1 + A3 is 10 microseconds, and extrapolation yields approximately 1 ns for A1 <--> A3. The pronounced kinetic differences imply different structural rearrangements. Crystallographic data support this conclusion: They show that formation of the A0 substate involves a major change of the protein structure; the distal histidine rotates about the C(alpha)-C(beta) bond, and its imidazole sidechain swings out of the heme pocket into the solvent, whereas it remains in the heme pocket in the A1 <--> A3 interconversion. The fast A1 <--> A3 exchange is inconsistent with structural models that involve differences in the protonation between A1 and A3.     

Mutational analyses of human eIF5A-1--identification of amino acid residues critical for eIF5A activity and hypusine modification

The eukaryotic translation initiation factor 5A (eIF5A) is the only protein that contains hypusine [Nepsilon-(4-amino-2-hydroxybutyl)lysine], which is required for its activity. Hypusine is formed by post-translational modification of one specific lysine (Lys50 for human eIF5A) by deoxyhypusine synthase and deoxyhypusine hydroxylase. To investigate the features of eIF5A required for its activity, we generated 49 mutations in human eIF5A-1, with a single amino acid substitution at the highly conserved residues or with N-terminal or C-terminal truncations, and tested mutant proteins in complementing the growth of a Saccharomyces cerevisiae eIF5A null strain. Growth-supporting activity was abolished in only a few mutant eIF5As (K47D, G49A, K50A, K50D, K50I, K50R, G52A and K55A), with substitutions at or near the hypusine modification site or with truncation of 21 amino acids from either the N-terminus or C-terminus. The inactivity of the Lys50 substitution proteins is obviously due to lack of deoxyhypusine modification. In contrast, K47D and G49A were effective substrates for deoxyhypusine synthase, yet failed to support growth, suggesting critical roles of Lys47 and Gly49 in eIF5A activity, possibly in its interaction with effector(s). By use of a UBHY-R strain harboring genetically engineered unstable eIF5A, we present evidence for the primary function of eIF5A in protein synthesis. When selected eIF5A mutant proteins were tested for their activity in protein synthesis, a close correlation was observed between their ability to enhance protein synthesis and growth, lending further support for a central role of eIF5A in translation.