The use of phosphorothioate-modified DNA in restriction enzyme reactions to prepare nicked DNA

Site-directed mutagenesis of beta-lactamase TEM-1. Investigating the potential role of specific residues on the activity of Pseudomonas-specific enzymes

From sequence alignments, two groups can be defined for the carbenicillin-hydrolysing beta-lactamases (CARB enzymes). One group includes the Pseudomonas-specific enzymes PSE-1, PSE-4, CARB-3, CARB-4 and also the Proteus mirabilis GN79, for which the well-conserved residue Lys 234 in all class-A beta-lactamases is changed to an arginine residue. The second group includes the enzymes PSE-3 and AER-1 which have an arginine or a lysine residue at position 165. All these enzymes also have leucine at position 68, threonine at position 104 and glycine at position 240. We engineered these mutations into the TEM-1 beta-lactamase to study their potential role in defining the substrate profile of the CARB enzymes. The mutations K234R and E240G in TEM-1 noticeably increased the hydrolysis of carboxypenicillins relative to other penicillins by approximately sixfold and twofold, respectively. The variant E240G also demonstrated an improved rate of second-generation cephalosporin and cefotaxime hydrolysis. In contrast, the substitution of Trp165 by arginine does not extend the substrate profile to alpha-carboxypenicillins nor does it noticeably modify the kinetic behavior of the enzyme. The mutations M68L and E104T do not have a large effect on the hydrolysis rate but the mutation E104T enhances the affinity of the enzyme for third-generation cephalosporins. As the mutation K234R resulted in a severe decrease in the affinity for carboxypenicillins, the double mutant E240G/K234R was constructed in an attempt to enhance the CARB character of the enzyme. Contrary to what could be expected, the additional mutation E240G for the TEM-1 K234R enzyme increases neither the catalytic constant for the carboxypenicillins nor the affinity towards these substrates. Consequently, this study strongly suggests that the three-dimensional structures of the active site of the TEM-1 enzyme and PSE-3, PSE-4 or other related enzymes are significantly different. This probably explains the discrepancy of the substrate profile between the CARB enzymes and the TEM-1 protein variants.

A survey of a functional amino acid of class C beta-lactamase corresponding to Glu166 of class A beta-lactamases

The class C beta-lactamase of Citrobacter freundii GN346 is a typical cephalosporinase comprising 361 amino acids. The aspartic acid at position 217 and glutamic acid at position 219 in this beta-lactamase were, respectively, previously shown not to be the counterpart of Glu166 (ABL166) in class A beta-lactamases, even though sequence alignment of class A and C enzymes strongly suggested this possibility [(1990) FEBS Lett. 264, 211-214; (1990) J. Bacteriol. 172, 4348-4351]. We tried again to assign candidates for the counterpart of Glu166 through sequence alignment based on other criteria, the glutamic acids at positions 195 and 205 in the class C beta-lactamase being selected. To investigate this possibility, these two glutamic acids were changed to glutamine, lysine or alanine, respectively. All the mutant enzymes showed more than 50% of the activity of the wild-type enzyme, indicating that the possibility was ruled out. These results strongly suggested the possibility that the class C beta-lactamase lacks a functional acidic residue corresponding to Glu166 in class A enzymes.

Effects of the residue adjacent to the reactive serine on the substrate interactions of Drosophila esterase 6

Esterase 6 from Drosophila melanogaster is a carboxylesterase that belongs to the serine esterase multigene family. It has a basic histidine (His) at residue 187, adjacent to the reactive serine (Ser) at residue 188, whereas most other characterized members of the family have an acidic glutamate (Glu) in the equivalent position. We have used site-directed in vitro mutagenesis to replace the His codon of the esterase 6 gene with either Gln or Glu codons. The enzymes encoded by these active-site mutants and a wild-type control have been expressed, purified, and characterized. Substitution of Gln for His at position 187 has little effect on the biochemical properties of esterase 6, but the presence of Glu at this position is associated with three major differences. First, the pH optimum is increased from 7 to 9. Second, the mutant enzyme shows decreased activity for beta-naphthyl esters and p-nitrophenyl acetate but has gained the ability to hydrolyze acetylthiocholine. Finally, the Gibb's free energy of activation for the enzyme is increased. These results suggest that residue 187 interacts directly with the substrate alkyl group and that this interaction is fully realized in the transition state. We further propose that the presence of His rather than Glu at position 187 in esterase 6 contributes significantly to its functional divergence from the cholinesterases and that this divergence is due to different interactions between residue 187 and the substrate alkyl group.

Dominant negative mutations of the scavenger receptor. Native receptor inactivation by expression of truncated variants

The bovine scavenger receptor was truncated at amino acid 266 or 310 to delete either all or part, respectively, of the collagen-like domain. The truncated receptors were inactive in the binding and internalization of acetyl (Ac) low density lipoprotein (LDL). Coexpression of truncated receptor with the native receptor dramatically reduced the percentage of cells internalizing fluorescently labeled Ac LDL, compared with cells expressing the native receptor alone. The mutant truncated at amino acid 266 was most effective in receptor inactivation, resulting in a 42% or 80% decrease in the percentage of cells expressing active receptor when transfected in a 1:1 or 1:2 molar ratio (native:mutant), respectively, with native receptor. Degradation of 125I-Ac LDL was reduced up to 90% when the native and truncated mutant receptors were coexpressed. Scavenger receptor inhibition was specific because the activity of the LDL receptor was not altered. Transient transfection of the mouse macrophage cell line P388D1 with truncated scavenger receptor resulted in a 65% decrease in the uptake and degradation of Ac LDL but did not decrease the degradation of beta-migrating very low density lipoprotein, which is LDL receptor-mediated. These results demonstrate that expression of truncated bovine scavenger receptor inactivates both the native bovine and murine scavenger receptors, producing a dominant negative phenotype in vitro.

Homophilic adhesion of the myelin P0 protein requires glycosylation of both molecules in the homophilic pair

The myelin P0 protein is glycosylated at a single site, asparagine 93, within its only immunoglobulin (Ig)-like domain. We have previously shown that P0 behaves like a homophilic adhesion molecule (Filbin, M. T., F. S. Walsh, B. D. Trapp, J. A. Pizzey, and G. I. Tennekoon. 1990. Nature (Lond.). 344:871-872). To determine if the sugar residues of this molecule contribute to its adhesiveness, the glycosylation site was eliminated by replacing asparagine 93 with an alanine, through site-directed mutagenesis of the P0 cDNA. The mutated P0 cDNA was transfected into CHO cells and surface expression of the mutated P0 was assessed by immunofluorescence, limited trypsinization and an ELISA. A cell line was chosen which expressed approximately equivalent amounts of the unglcosylated P0 (UNGP0) at the cell surface as did a cell line expressing the fully glycosylated P0 (GPo); the adhesive properties of these two cell lines were compared. It was found that when a single cell suspension of the UNGPo cells were incubated, by 60 min, unlike the GP0 cells, they had not formed large aggregates; they were indistinguishable from the control transfected cells. This suggests that the UNGP0 protein does not behave like an adhesion molecule. To establish if only one molecule in the P0:P0 homophilic pair must be glycosylated for adhesion to occur, the ability of UNGP0 cells to adhere to GP0 cells was assessed both qualitatively and quantitatively. The results of both types of assay imply that, indeed, both P0 molecules in the homophilic pair must be glycosylated for adhesion to take place.

Association of p53 binding and immortalization of primary C57BL/6 mouse embryo fibroblasts by using simian virus 40 T-antigen mutants bearing internal overlapping deletion mutations

To more precisely map the immortalization and p53 binding domains of T antigen, a large series of overlapping deletion mutations were created between codons 251 to 651 by utilizing a combination of Bal 31 deletion and oligonucleotide-directed mutagenesis. Immortalization assay results indicated that amino acids (aa) 252 to 350, 400, and 451 to 532 could be removed without seriously compromising immortalization, although the appearance of immortal colonies was delayed in some cases. Western immunoblotting experiments indicated that the p53 binding capacities of T antigen produced by mutants missing aa 252 to 300, 301 to 350, 400, or 451 to 532 were only slightly reduced relative to that of wild-type T antigen. Within the limits of this deletion analysis, the immortalization and p53 binding domains appear to be colinear and, in fact, may represent two aspects of the same domain. This deletion analysis eliminates the entire zinc finger domain (aa 302 to 320), a small portion of the leucine-rich region (aa 345 to 350), and a large portion of the ATP binding domain (aa 451 to 528) as participants in p53 binding or in the immortalization process. The results also show that removal of T antigen amino acids within the region 451 to 532 appears to alter the capacity of newly synthesized but not older T antigen and p53 molecules to form complexes.

Significance of hydrophobic S4-P4 interactions in subtilisin 309 from Bacillus lentus

The subtilisins have an extended substrate binding cleft comprising at least 8 subsites. Two pockets at the S1 and S4 sites are particularly conspicuous, and the interactions between substrate and these two pockets are very important for the substrate specificity. Phe residues have mutationally been introduced at one of positions 102, 128, 130, and 132 of the subtilisin Savinase from Bacillus lentus to investigate the effects of introducing bulky groups along the rim of the S4 binding pocket. It is shown that the marked P4 preference of wild-type Savinase for aromatic groups is eliminated by the Gly102-->Phe and Ser128-->Phe mutations, indicating that bulky groups at positions 102 and 128 block the S4 binding site. In contrast, the activity toward hydrophilic P4 residues is not nearly as affected by these mutations, suggesting that the binding mode of the P4 side chain is dependent on its properties. Introduction of a bulky -CH2-S-CH2-CH2-pyridyl group at position 128, by mutational incorporation of Cys followed by chemical modification with 2-vinylpyridine, has essentially the same effect. The Ser130-->Phe mutation hardly affects the activity of the enzyme while the Ser-->Phe mutation at position 132 renders the preference for hydrophobic groups in P4 even more pronounced. This mutation furthermore affects the size of the S4 pocket. An analysis of double mutants at positions 132 and 104 suggests that the S4 region is flexible and is adjusted upon binding of substrates.

Interaction of Escherichia coli ribosomal protein S7 with 16S rRNA

The interaction between Escherichia coli ribosomal protein S7 and 16S rRNA was investigated using in vitro synthesized RNA transcripts. It was shown by nitrocellulose membrane filtration that RNA transcripts corresponding to the 3' major domain (nucleotides 926-1393) and to the lower half of this domain (nucleotides 926-986/1219-1393) bound S7 with the same affinity as 16S rRNA. A series of deletion mutants of the DNA coding for the lower half of the 3' major domain were constructed and the corresponding RNA fragments were assayed for their capacity to bind S7. A minimal domain of 108 nucleotides which can still efficiently bind S7 was thus obtained. In this domain, the 1304-1308/1329-1333 irregular helix and the 1351-1371 irregular hairpin were found to contain important determinants for S7 binding.

Identification of zinc-binding ligands in the class II fructose-1,6-bisphosphate aldolase of Escherichia coli

An expression and mutagenesis system for the E. coli Class II fructose-1,6-bisphosphate aldolase has been created by modification of the vector pKfda (Biochem. J. 257 (1989) 529-534). Large amounts of Class II aldolase (about 1 g/l in crude extracts), with properties consistent with those previously reported for the naturally occurring enzyme (Biochem. J. 169 (1978) 633-641) are obtained. The enzyme contains 2 zinc ions per enzyme dimer. We have investigated the nature of the zinc-binding site of the enzyme by site-directed mutagenesis. His-108, His-111, Cys-112 and His-142 were identified as possible zinc-binding ligands by sequence alignments and comparisons with other known zinc-containing enzymes. Mutation of these residues identified His-108 and His-111 as two of the ligands directly responsible for the tight binding of zinc. Mutation of the other two residues results in only a small effect on the amount of zinc bound per monomer and a corresponding change in specific activity. These residues are, therefore, unlikely to be directly involved in zinc binding, but may be indirectly involved in some manner in the zinc-binding environment.

Site-directed mutagenesis and chemical modification of histidine residues on an alpha-class chick liver glutathione S-transferase CL 3-3. Histidines are not needed for the activity of the enzyme and diethylpyrocarbonate modifies both histidine and lysine residues

Each chick liver glutathione S-transferase CL 3 subunit contains three histidine residues: His142, His158 and His228. CL 3-3 can be inactivated by treating with diethylpyrocarbonate. The inactivation process is pH dependent and the pKa of the modified residue is 6.4. The second-order inhibition rate constant is 741 M-1min-1 at pH 7.0. Based on difference-spectrum and kinetic analysis, inactivation coincides with the modification of one histidine residue. However, hydroxylamine treatment of the diethylpyrocarbonate-modified enzyme only partially restored the activity (30-50%) of CL 3-3. By tryptic mapping and amino acid sequence analysis, His228 and Lys14 have been identified as the modified residues. Mutants with histidine to serine replacement (H142S and H158S) or C-terminal histidine deletion (des-H228) were constructed and over-expressed in Spodoptera frugiperda cells using a baculovirus system. The mutants are enzymically active. Furthermore, the des-H228 mutant can be inactivated by diethylpyrocarbonate. These results support the conclusion that histidines are not involved in the enzymic mechanism of CL 3-3.

The effect of loop size in antisense and target RNAs on the efficiency of antisense RNA control

Most natural antisense RNAs display a high degree of secondary structure with stem-loops as their most prominent feature. Mutations affecting the inhibitory activity of these RNAs most often map in or close to loop regions in both the antisense and target RNAs. The primary recognition loops often contain 5-7 unpaired nucleotides. Nucleotide changes in the loops affect the binding rate and, hence, the inhibitory effect on the activity of the target RNA. Here we address the question whether loop sizes affect binding rates between antisense and target RNAs, using the replication control system of plasmid R1 as a model system. By creating a series of loop size mutants we show that loop size alterations have strong effects on the binding rates between the two reactant RNAs in vitro, and that most of the mutations analyzed display corresponding effects on antisense RNA control in vivo. Our data suggest that the three-dimensional structures of antisense and target RNA stem-loops are crucial for determining binding rates. The implications of these results for the design of efficient artificial antisense RNA control systems are discussed.

Mutational analysis of human T-cell leukemia virus type I Tax: regions necessary for function determined with 47 mutant proteins

We have made 47 mutations that span the length of the human T-cell leukemia virus type I (HTLV-I) Tax open reading frame. Of the 47 mutations, 38 were substitutions of single amino acids, 5 were missense changes in two or more amino acids, and 4 were deletions. A subset of these mutations includes individual changes of all 26 naturally occurring serines to alanines. By assaying each mutant protein separately on the HTLV-I long terminal repeat (LTR) and the human immunodeficiency virus type 1 (HIV-1) LTR in parallel, we were able to identify regions of Tax selectively necessary for each promoter. A small region in the carboxyl terminus, amino acids 315 to 325, was found to be selectively important for activation of the HTLV-I LTR. Three changes at serine 113, serine 160, and serine 258 were found to specifically affect function on the HIV-1 LTR. Surprisingly, we found that the great preponderance of missense changes (32 of 42) in Tax did not affect function.

Mutational replacements in subtilisin 309. Val104 has a modulating effect on the P4 substrate preference

The previous notion that the amino acid side chain at position 104 of subtilisins is involved in the binding of the side chain at position P4 of the substrate has been investigated. The amino acid residue Val104 in subtilisin 309 has been replaced by Ala, Arg, Asp, Phe, Ser, Trp and Tyr by site-directed mutagenesis. It is shown that the P4 specificity of this enzyme is not determined solely by the amino acid residue occupying position 104, as the enzyme exhibits a marked preference for aromatic groups in P4, regardless of the nature of the position-104 residue. With hydrophilic amino acid residues at this position, no involvement is seen in binding of either hydrophobic or hydrophilic amino acid residues at position P4 of the substrates. The substrate with Asp in P4 is an exception, as the preference for this substrate is increased dramatically by introduction of an arginine residue at position 104 in the enzyme, presumably due to a substrate-induced conformational change. However, when position 104 is occupied by hydrophobic residues, it is highly involved in binding of hydrophobic amino acid residues, either by increasing the hydrophobicity of S4 or by determining the size of the pocket. The results suggest that the amino acid residue at position 104 is mobile such that it is positioned in the S4 binding site only when it can interact favourably with the substrate's side chain at position P4.

Modified DNA fragments activate NaeI cleavage of refractory DNA sites

Endonuclease NaeI cleaves DNA using a two-site mechanism. The DNA-binding sites are nonidentical: they recognize different families of flanking sequences. A unique NaeI site that is resistant to cleavage resides in M13 double-stranded DNA. NaeI can be activated to cleave this site by small DNA fragments containing one or more NaeI sites. These activators are not practical for genetic engineering because unphosphorylated activators that are consumed during the cleavage of substrate give ends that may interfere with subsequent ligations. We show that a DNA fragment containing phosphorothioate linkages at the NaeI scissile bonds (S-activator) is not cleaved by NaeI, even though this S-activator binds to the substrate site. The S-activator activates NaeI to cleave M13 DNA under conditions that completely exhaust unsubstituted activator. These results demonstrate that activation is not coupled to cleavage of activator, that NaeI reverts to its inactive state soon after dissociation of the EA complex, and that S-activator makes for a nondepletable activator during prolonged incubations.

Chimeric allosteric citrate synthases: construction and properties of citrate synthases containing domains from two different enzymes

The citrate synthases of the gram-negative bacteria, Escherichia coli and Acinetobacter anitratum, are allosterically inhibited by NADH. The kinetic properties, however, suggest that the equilibrium between active (R) and inactive (T) conformational states is shifted toward the T state in the E. coli enzyme. We have now manipulated the cloned genes for the two bacterial enzymes to produce two chimeric proteins, in which one folding domain of each subunit is derived from each enzyme. One chimera (the large domain from A. anitratum and the small domain from the E. coli enzyme) is designated CS ACI::eco; the other is called CS ECO::aci. Both chimeras are roughly as active as the wild type parents, but their Km values for both substrates are lower than those for the E. coli enzyme, and NADH inhibition is markedly sigmoid, while that for E. coli citrate synthases is hyperbolic. Curve-fitting to the allosteric equation suggests that these differences are the result of the destabilization of the T state in the chimeras. The ACI::eco chimera exists almost entirely as a hexamer, like the A. anitratum enzyme, while the ECO::aci chimera, like the E. coli synthase, forms three major bands on nondenaturing polyacrylamide gels, two of them hexamers of different net charge, and one a dimer. These findings indicate that subunit interactions leading to hexamer formation in allosteric citrate synthases of gram-negative bacteria involve mainly the large domains. The chimeras are also used to show that the NADH binding site of E. coli citrate synthase is located entirely in the large domain. Sensitivity of the chimeras to denaturation by urea, to which the A. anitratum enzyme is much more resistant than the E. coli enzyme, is determined by the large domains. Sensitivity to inactivation by subtilisin is intermediate between those shown by the E. coli (very sensitive) and A. anitratum (quite resistant) synthases. This result suggests that digestibility by subtilisin is determined by conformational factors as well as the amino acid sequences of the target regions.

Site-directed mutagenesis and chemical modification of cysteine residues of rat glutathione S-transferase 3-3

Rat liver glutathione S-transferase (GST) 3-3 is composed of two identical subunits, each containing three cysteine residues, Cys-86, Cys-114 and Cys-173. We have shown previously that Cys-86 is not involved in the enzymic activity of GST 3-3 [Hsieh, Huang, Chen, Lai & Tam (1991) Biochem, J. 278, 293-297]. At 50 degrees C, iodoacetamide can inactivate the enzyme by modifying Cys-86 and Cys-114. Cys-114 can be protected against iodoacetamide inhibition by S-(dinitrophenyl)glutathione. Site-directed mutagenesis was used to construct mutants in which serine replaced one (C114S and C173S) or all three (CallS) cysteine residues. These mutants were over-expressed in Spodoptera frugiperda cells in a baculovirus system and were found to be fully active. Replacing Cys-86 or Cys-114 with alanine (C86A and C114A) does not diminish the activity of the protein. The results suggest that cysteines are not involved in the enzymic mechanism, and Cys-114 is possibly located at the active site of GST 3-3.

The effect of amino acid substitution at position 219 of Citrobacter freundii cephalosporinase on extension of its substrate spectrum

The cephalosporinase of Citrobacter freundii GN346 is a class-C beta-lactamase comprising 361 amino acids. The substitution of the glutamic acid at position 219 in the enzyme by lysine was previously shown to broaden its substrate specificity to unfavorable substrates such as oxyimino cephalosporins [Tsukamoto, K., Ohno, R. & Sawai, T. (1990) J. Bacteriol. 172, 4348-4351]. To investigate the cause of this phenomenon, Glu219 was changed to glutamine, cysteine or tryptophan. All the resultant enzymes showed higher cefuroxime-hydrolytic activities than the wild type, the order of increasing cefuroxime-hydrolytic activity being as follows: Trp greater than Lys greater than Cys greater than Gln greater than Glu. The rate of hydrolysis of cefuroxime by the Trp219 enzyme was approximately 3 x 10(4) times that of the wild-type enzyme. The order of increasing cefuroxime hydrolysis was approximately proportional to the molecular volume of the amino acid substituted and independent of the ionic character of the amino acids. The cysteine residue at position 219 in the Cys219 enzyme allowed its complete reaction with an SH-blocking reagent, 4-chloromercuriphenylsulfonic acid. The modified enzyme with the bulkier residue showed a 45% higher cefuroxime-hydrolytic activity than the untreated enzyme. These results suggested that extension of the substrate spectrum may be attributed to alteration in the configuration of the enzyme around position 219.

In vitro mutagenesis of the mitochondrial leucyl-tRNA synthetase of S. cerevisiae reveals residues critical for its in vivo activities

The mitochondrial leucyl-tRNA synthetase (mLRS) of Saccharomyces cerevisiae is involved in both mitochondrial protein synthesis and pre-mRNA splicing. We have created mutations in the regions HIGH, GWD and KMSKS, which are involved in ATP-, amino acid- and tRNA-binding respectively, and which have been conserved in the evolution of group I tRNA synthetases. The mutants GRD and NMSKS have no discernible phenotype. The mutants AWD and ARD act as null alleles and lead to the production of 100% cytoplasmic petites. The mutants HIGN, NIGH and KMSNS are unable to grow on glycerol even in the presence of an intronless mitochondrial genome and accumulate petites to a greater extent than the wild-type but less than 40%. Experiments with an imported bI4 maturase indicate that the lesion in these mutations primarily affects the synthetase and not the splicing functions.

Structural and functional features of cis-acting sequences in the basic replicon of plasmid ColIb-P9

We have structurally and functionally analyzed the cis-elements essential for ColIb-P9 plasmid DNA replication. The putative oriV region encompassed a region of 172 base pairs (bp) located 152 bp downstream of the repZ gene. A typical dnaA box found in this region proved nonessential for the DNA replication of ColIb-P9. The ssi signal of ColIb-P9 is a homologue of the G-sites of R1 and R100 plasmids. Deletion of the G-site led to 1.5-fold reduction of the copy number, suggesting that although this G-site is not essential, it is important for efficient ColIb-P9 DNA replication. In addition, the ColIb-P9 replicon is highly and extensively homologous with the P307 (RepFIC) replicon, and highly homologous with the R100 (RepFIIA) replicon around the G-site region. These facts imply a common ancestry from which the plasmids have evolved.

HTLV-I Tax is a zinc-binding protein: role of zinc in Tax structure and function

We have examined the functional significance of the cysteine- and histidine-rich region (amino acids 22-53) of HTLV-I Tax. A modeling of this region suggests two possible overlapping zinc-finger-like motifs. Using a zinc blotting technique, we show that Tax binds zinc. An N-terminal deletion in Tax that removed this zinc-finger region abolished the ability to bind zinc. Site-directed mutagenesis was used to generate 10 separate mutations so as to discriminate between the two alternative zinc-binding structures. Each Tax mutant was studied for its ability to trans-activate the HTLV-I LTR. Five of the ten mutations inactivated trans-activation. Our results support that one of the two putative zinc fingers is an integral element of Tax structure.

Direct expression of mature bovine adrenodoxin in Escherichia coli

Site-directed mutagenesis was utilized to enable direct expression of the mature form of bovine adrenodoxin cDNA using the pKK223-3 expression vector in Escherichia coli. Expression was under control of the "tac" promoter and resulted in a direct expression of soluble mature bovine adrenodoxin (greater than 15 mg per liter). Chromatographic behavior of recombinant adrenodoxin did not differ from that reported for mature native adrenodoxin. The purified recombinant protein was identical to native mitochondrial adrenodoxin on the basis of molecular weight, NH2 terminal sequencing and immunoreactivity. E. coli lysates were brown in color, and the purified protein possessed a visible absorbance spectra identical to native bovine adrenodoxin consistent with incorporation of a [2Fe-2S] cluster in vivo. Recombinant bovine adrenodoxin was active in cholesterol side-chain cleavage when reconstituted with adrenodoxin reductase and cytochrome P450scc and exhibited kinetics reported for native bovine adrenodoxin. The presence of the adrenodoxin amino terminal presequence does not appear to be essential for correct folding of mature recombinant adrenodoxin in E. coli. This expression system should prove useful for overexpression of adrenodoxin mutants in future structure/function studies. The approach described herein can potentially be used to directly express the mature form of any protein in bacteria.

Expression in Escherichia coli of a sub-gene encoding the lipoyl and peripheral subunit-binding domains of the dihydrolipoamide acetyltransferase component of the pyruvate dehydrogenase complex of Bacillus stearothermophilus

A sub-gene encoding the N-terminal 170 residues of the dihydrolipoamide acetyltransferase chain of the pyruvate dehydrogenase multienzyme complex of Bacillus stearothermophilus was over-expressed in Escherichia coli. The expressed polypeptide consists of the lipoyl domain, inter-domain linker and peripheral subunit-binding domain; these were found to have folded into their native functional conformations as judged by reductive acetylation of the lipoyl domain, limited proteolysis of the linker region and ability to bind the dihydrolipoamide dehydrogenase dimer. The di-domain was largely (80%) unlipoylated; a small proportion (4%) was correctly modified with lipoic acid and the remainder (16%) was aberrantly modified with octanoic acid. A polyclonal antiserum was raised that recognized both the di-domain and the individual component domains. The 400 MHz 1H-n.m.r. spectrum of the di-domain showed resonances corresponding to those seen in spectra of the lipoyl domain, plus others characteristic of amino acid residues in the flexible linker region. Further, as yet unidentified, resonances are likely to be derived from the peripheral subunit-binding domain. The existence and independent folding of the peripheral subunit-binding domain is thus confirmed and its purification in large-scale amounts for detailed structural analysis is now possible.

Asp-51 and Asp-120 are important for the transport function of the Escherichia coli melibiose carrier

Asp-51 and Asp-120 of the Escherichia coli melibiose carrier on plasmid pKKMB were separately replaced by amber codons and transformed into eight amber suppressor strains, producing eight amino acid substitutions for each site. Glu-51 and Glu-120 were the only replacements in the carrier that allowed the cells to ferment melibiose and that showed transport of melibiose against a concentration gradient. Revertants to Glu-51 and Glu-120 show less activity than the wild type. The Asp-51 position is more crucial for Na(+)-stimulated melibiose accumulation than is the Asp-120 site.

Reconstitution in vivo of penicillin G acylase activity from separately expressed subunits

Penicillin G acylase from Escherichia coli ATCC11105 is synthesized as a precursor polypeptide with a signal sequence for secretion into the periplasm and an endopeptide separating two subunit domains. Proteolytic processing leads to mature, heterodimeric penicillin G acylase. We have shown that the alpha- and beta-subunits of the enzyme, which have no detectable enzymatic activity on their own, can reconstitute enzyme activity when their genes are put into an E. coli host on separate plasmids. Activity is reconstituted in the cytoplasm whereas normally processing and formation of the active heterodimer occurs in the periplasm. Enzyme activity can reach levels close to wild type in the strain used. The activity recovered from a combination of alpha-subunit linked to a 54-amino-acid endopeptide and beta-subunit was lower than with the subunits alone.

Deletion of internal sequence on the HDV-ribozyme: elucidation of functionally important single-stranded loop regions

In elucidating functionally important single-stranded loop regions derived mainly from three models in genomic hepatitis delta virus (HDV) ribozyme possessing self-cleavage activity, we have constructed several internal deletion variants of the HDV133 molecule (654-786 nt on genomic RNA) by oligonucleotide-directed mutagenesis. When self-cleavage activities were compared among variants, the HDV133DI-1 (deletion of 701-718 nt) and HDV133DI-3 (deletion of 740-752 nt) ribozyme could maintain their self-cleavage activity, despite at reduced level. However, the activity could be regained in both mutants by some extent under partially denaturing conditions. These results suggest that the above two single-stranded RNA loop regions in HDV ribozyme are not part of the catalytic core but might be involved in the stability of the molecule. In contrast, deletion mutants such as HDV133DI-2 (deletion of 696-722 nt), HDV88DI-1 (deletion of 701-718 nt), HDV88DI-2 (deletion of 696-722 nt), and HDV88DI-4 (deletion of 733-760 nt) abolished catalytic activity. These results suggest that the remaining single-stranded regions of bases between 726-731 and 762-766 in the HDV88 ribozyme may be the potential regions to interact with Mg2+ ions.

Engineering surface charge. 1. A method for detecting subunit exchange in Escherichia coli glutathione reductase

The gene gor encoding Escherichia coli glutathione reductase was mutated to create a positively charged N-terminal extension consisting of five arginine residues followed by a factor Xa cleavage site to the enzyme polypeptide chain. The modified protein assembled in vivo to yield a dimeric enzyme with kinetic parameters indistinguishable from those of wild-type glutathione reductase. The N-terminal extension could not be released by treatment with factor Xa but could be removed by exposure to trypsin, again without effect on the enzyme activity. The modified enzyme was readily separated from the wild-type enzyme by means of ion-exchange chromatography or nondenaturing polyacrylamide gel electrophoresis. Incubation of the modified and wild-type enzymes, separately or as a mixture, with NADH led to their partial inactivation, and activity was restored by exposure to 1 mM reduced glutathione. No hybrid dimer was formed in the mixture of modified and wild-type enzymes, as judged by polyacrylamide gel electrophoresis, strongly suggesting that the inactivation induced by NADH was not due to dissociation of the parental dimers. The addition of otherwise benign positively or negatively charged extensions to the N- or C-terminal regions of the constituent polypeptide chains of oligomeric enzymes offers a simple route to detecting hybrid formation and the causative subunit dissociation and exchange.

Engineering surface charge. 2. A method for purifying heterodimers of Escherichia coli glutathione reductase

Two gor genes encoding different mutants of Escherichia coli glutathione reductase have been expressed in the same E. coli cell, leading to the creation of a hybrid form of the enzyme dimer. One of the gor genes carried, in addition to various directed mutations, a 5' extension that encodes a benign penta-arginine "arm" added to the N-terminus of the glutathione reductase polypeptide chain [Deonarain, M.P., Scrutton, N.S., & Perham, R.N. (1992) Biochemistry (preceding paper in this issue)]. This made possible, by means of ion-exchange chromatography or nondenaturing polyacrylamide gel electrophoresis, the facile separation of the hybrid enzyme from the two parental forms. Moreover, the two subunits in the hybrid enzyme could be made to carry different mutations. In this way, glutathione reductases with only one active site per dimer were generated: the effects of replacing tyrosine-177 with glycine in the NADPH-binding site, which greatly diminishes the Km for glutathione and switches the kinetic mechanism from ping-pong to ordered sequential, and of replacing His-439 with glutamine in the glutathione-binding site, which greatly diminishes the Km for NADPH, were both found to be restricted to the one active site carrying the mutations. This system of generating separable enzyme hybrids is generally applicable and should make it possible now to undertake a more systematic study of catalytic mechanism and assembly for the many enzymes with quaternary structure.

Analysis of the integration function of the streptomycete bacteriophage phi C31

A 2.1 kb (1 kb = 10(3) base-pairs) segment of DNA from the streptomycete bacteriophage phi C31 was found to be sufficient to direct site-specific integration of plasmid vectors in Streptomyces ambofaciens and Streptomyces fradiae in the absence of any streptomycete origin of replication. Sequencing and analysis of phage, chromosomal and junction attachment sites of S. ambofaciens and S. fradiae revealed that recombination is conservative and that crossover takes place within three bases of homology between phage and host. Deletion analysis, sequencing and site-specific mutagenesis of the phi C31 DNA revealed a large open reading frame (ORF 613) whose expression was necessary for integration. This ORF begins near the point of crossover and reads away from the attachment site. A comparison of the predicted amino acid sequence of ORF 613 with known recombinases did not reveal any significant similarities. A genetic analysis of the amino-terminal region of ORF 613 suggested that translation could initiate at any one of three possible start codons. Primer extension experiments showed that transcriptional initiation occurred at a T and a C only four and five bases, respectively, from the site of crossover. This analysis suggested that ORF 613 would be separated from its promoter upon integration.

High-level secretion of correctly processed recombinant human interleukin-1 beta in Kluyveromyces lactis

The lactose-assimilating yeast, Kluyveromyces lactis, has been developed as a microbial host for the synthesis and secretion of human proteins. Here, we report the use of multi-copy vectors based on the 2 mu-like plasmid pKD1 from Kluyveromyces drosophilarum [Chen et al., Nucleic Acids Res. 14 (1986) 4471-4481] for the secretion of recombinant human interleukin-1 beta (reIL-1 beta). High levels of reIL-1 beta were secreted into the growth medium when the structural gene was fused in-frame to a synthetic secretion signal derived from the 'pre'-region of the K. lactis killer toxin. N-terminal sequencing of the excreted protein showed highly efficient (greater than 95%) maturation of the signal sequence. Synthesis as prepro-IL-1 beta, the 'pro'-sequence being derived from the human serum albumin-encoding gene, resulted in equally efficient secretion of mature IL-1 beta. Cytoplasmic production of Met-IL-1 beta, without a secretion signal, was found to be toxic to K. lactis. As in Saccharomyces cerevisiae [Baldari et al., EMBO J. 6 (1987) 229-234], but unlike native human IL-1 beta, K. lactis reIL-1 beta is glycosylated. This glycosylation led to a 95% loss of its biological activity. Removal of the carbohydrate chains by endo-beta-N-acetyl-glucosamidase H treatment fully restored the biological activity. A modified form of IL-1 beta (Asn7----Gln7), in which the unique site for Asn-linked glycosylation was deleted, exhibited the same biological activity as native IL-1 beta. The level of secretion of mature recombinant IL-1 beta ws glycosylation-independent.

Regulation of replication of plasmid R1: an analysis of the intergenic region between copA and repA

The synthesis of the rate-limiting RepA replication initiator protein of plasmid R1 is negatively controlled by an antisense RNA, CopA. The regulation is posttranscriptional and involves an inhibitory effect on RepA translation mediated by the binding of CopA to its target (CopT) in the leader region of the RepA mRNA. The evolutionary conservation of the intergenic region between the copA gene and the repA reading frame among plasmids related to R1 may be indicative of an important function in this regulation. One possibility is that sequences/structures in this region might be required for the presumed distal effect of CopAQCopT binding. We have performed a mutational analysis of this region, starting with a mutant repA-lacZ fusion plasmid that shows decreased RepA-LacZ synthesis compared to a wild-type construct, and have identified five compensatory mutations that increase repA-lacZ expression. Two of these were single base-pair substitutions in the copA promoter leading to a decrease in CopA transcription. The other three mutations increased RepA synthesis in the presence as well as in the absence of functional CopA. Reconstructed plasmids carrying these mutations--in conjunction with the original down-mutation or in an otherwise wild-type background--show the expected increase in copy number. The effect of two of these mutations is consistent with the destabilization of a putative secondary structure which may be responsible for the normally low translation rate of the RepA reading frame. The implications of the types of mutations found in this study, as well as the absence of other classes of mutations, are discussed in terms of alternative possible models of CopA-mediated inhibition of RepA synthesis.

Engineering the substrate specificity of glutathione reductase toward that of trypanothione reduction

Glutathione reductase (EC 1.6.4.2; CAS registry number 9001-48-3) and trypanothione reductase (CAS registry number 102210-35-5), which are related flavoprotein disulfide oxidoreductases, have marked specificities for glutathione and trypanothione, respectively. A combination of primary sequence alignments and molecular modeling, together with the high-resolution crystal structure of human glutathione reductase, identified certain residues as potentially being responsible for substrate discrimination. Site-directed mutagenesis of Escherichia coli glutathione reductase was used to test these predictions. The mutation of Asn-21 to Arg demonstrated that this single change was insufficient to generate the greater discrimination against trypanothione shown by human glutathione reductase compared with the E. coli enzyme. However, the mutation of Ala-18, Asn-21, and Arg-22 to the amino acid residues (Glu, Trp, and Asn, respectively) in corresponding positions in Trypanosoma congolense trypanothione reductase confirmed that this region of polypeptide chain is intimately involved in substrate recognition. It led to a mutant form of E. coli glutathione reductase that possessed essentially no activity with glutathione but that was able to catalyze trypanothione reduction with a kcat/Km value that was 10% of that measured for natural trypanothione reductases. These results should be of considerable importance in the design of trypanocidal drugs targeted at the differences between glutathione and trypanothione metabolism in trypanosomatids and their hosts.

Gaucher disease: heterologous expression of two alleles associated with neuronopathic phenotypes

To investigate the molecular basis for the distinct neuronopathic phenotypes of Gaucher disease, acid beta-glucosidases expressed from mutant DNAs in Gaucher disease type 2 (acute) and type 3 (subacute) patients were characterized in fibroblasts and with the baculovirus expression system in insect cells. Expression of the mutant DNA encoding a proline-for-leucine substitution at amino acid 444 (L444P) resulted in a catalytically defective, unstable acid beta-glucosidase in either fibroblasts from L444P/L444P homozygotes or in insect cells. This mutation was found to be homoallelic in subacute neuronopathic (type 3) Gaucher disease. In comparison, expression of the mutant cDNA encoding an arginine-for-proline substitution at amino acid 415 (P415R) resulted in an inactive and unstable protein in insect cells. This allele was found only in a type 2 patient with the L444P/P415R genotype. The substantial variation in the type 3 phenotype (L444P homozygotes) suggests the complex nature of the molecular basis of phenotypic variation in Gaucher disease. Yet, the association of neuronopathic phenotypes with alleles producing severely compromised (L444P) or functionally null (P415R) enzymes indicates that the effective level of residual activity at the lysosome is likely to be a major determinant of the severity of Gaucher disease.

Utilization of DNA recombination for the two-step replacement of growth factor sequences in the vaccinia virus genome

An efficient procedure for the generation of sequence-specific alterations of the vaccinia virus genome was demonstrated. Homologous DNA recombination within cells infected with vaccinia virus was used for the deletion or replacement of promoter sequences of the viral growth factor gene by a procedure comparable to transplacement in Saccharomyces cerevisiae. This DNA replacement procedure can potentially be used to generate any sequence alteration within the vaccinia virus genome. Deletion of growth factor promoter sequences resulted in a dramatic reduction in growth factor gene transcription and protein synthesis. Replacement of growth factor promoter sequences with promoter sequences of the strong constitutive 40-kDa gene resulted in an increase in gene transcription and protein synthesis and an altered temporal pattern of expression. Virus containing mutations in the growth factor gene demonstrated different plaque morphologies on cell culture monolayers.

Cysteine-86 is not needed for the enzymic activity of glutathione S-transferase 3-3

Recombinant glutathione S-transferase 3-3 expressed in Spodoptera frugiperda (SF9) cells with the use of a baculovirus expression system was modified with 1 mM-iodoacetamide. Amino acid analysis indicated that 0.79 +/- 0.15 cysteine residue was modified per enzyme subunit. The S-carbaminomethylated protein retains the GSH-conjugating activity. Glutathione S-transferase 3-3 modified with iodo[14C]acetamide was digested with Achromobacter proteinase I and the resulting peptides were separated by h.p.l.c. The modified peptides were pooled and further digested with Staphylococcus aureus V8 proteinase. Isotope-labelled peptides were isolated and collected for N-terminal sequence analysis. By this procedure, cysteine-86 was identified as the major S-carbaminomethylated residue. Verification of this findings came from the use of site-directed mutagenesis in which this cysteine was replaced by serine (C86S mutant). The C86S mutant is enzymically active. Therefore cysteine-86 is not needed for the conjugation of GSH with electrophilic compounds on glutathione S-transferase 3-3.

A single-strand specific endonuclease activity copurifies with overexpressed T5 D15 exonuclease

The T5 D15 exonuclease purified from an overproducing strain of E. coli was shown to possess a low level of endonucleolytic activity specific for single-stranded DNA when assayed with 1-10 mM Mg2+ as co-factor. Endonuclease activity on double-stranded circular DNA could not be detected under these conditions. Nicked circular DNA was first gapped by the enzyme's exonucleolytic activity, creating a single-stranded region. This gapped substrate was then endonucleolytically cleaved and rapidly degraded. We show that a gapped and not a nicked substrate is required for this activity as previously suggested (Moyer, R. W. and Roth, C. T. 1977, J. Virol. 24, 177-193). The single-strand endonuclease activity could be selectively suppressed by using low concentrations of Mg2+ as co-factor (less than 1 mM), thus allowing nicked double-stranded circular DNA to be gapped to a single-stranded circular species. We also report on sequence similarities between the T5 exonuclease and several prokaryotic DNA polymerases.