The glutamine ligand in the ferrous iron active site of isopenicillin N synthase of Streptomyces jumonjinensis is not essential for catalysis.

Isopenicillin N synthase (IPNS) is a non-heme ferrous iron dependent dioxygenase that catalyses the ring closure of delta-(L-alpha-aminoadipoyl)-L-cysteinyl-D-valine (ACV) to isopenicillin N. We previously used site-directed mutagenesis to identify in the IPNS of Streptomyces jumonjinensis two histidines and one aspartic acid that are essential for activity. The recent crystal structure of the IPNS of Aspergillus nidulans establishes that these amino acids are iron ligands and reveals that the fourth ligand is the penultimate glutamine. The two histidines and one aspartic acid are conserved in several classes of non-heme ferrous iron dioxygenases, whereas the glutamine is present only in IPNSs. In this paper we show that the penultimate glutamine in S. jumonjinensis IPNS Gln-328 is not essential for catalysis. In contrast, Gln-230 which is highly conserved among the above dioxygenases and is proximal to the active site is crucial for activity.

Ferrous active site of isopenicillin N synthase: genetic and sequence analysis of the endogenous ligands.

Isopenicillin N synthase (IPNS) from Streptomyces jumonjinensis (M(r) 37,902) is a non-heme ferrous iron-containing enzyme that catalyzes the oxidative cyclization of the tripeptide delta-(L-alpha-aminoadipoyl)-L-cysteinyl-D-valine (ACV) to form isopenicillin N. Spectroscopic studies [reviewed in Cooper, R. D. (1993) Biomed. Chem. 1, 1-17] have led to a model for the coordination environment of the iron atom possessing three histidine and one aspartic acid endogenous ligands and a solvent molecule. A refinement of that model proposes that formation of the Fe(II) IPNS-ACV complex occurs with displacement of the H2O from the metal center and that one of the histidines is subsequently replaced by a solvent molecule on binding of dioxygen. Here we report genetic studies to determine the nature and location of the endogenous ligands in the S. jumonjinensis IPNS primary amino acid sequence that constitute the ferrous active site. Site-directed mutagenesis was used to exchange each of the seven histidines and the five aspartic acids that are conserved in bacterial and fungal IPNS proteins. Biochemical analysis of the alanine-substituted mutant proteins shows that two histidines, His212 and His268, and one aspartic acid, Asp214, are essential for enzyme activity. The other mutant enzymes have specific activities 5-68% that of wild type. Sequence analysis of 10 IPNS and 42 other non-heme ferrous iron-dependent dioxygenases reveal the presence of a common motif, HisXAsp(53-57)XHis, which in IPNS contains the identical two histidines and one aspartic acid essential for function. Accordingly, we have assigned residues His212, His268, and Asp214 as three of the four endogenous ligands postulated to form the IPNS ferrous active site. Compelling support for these conclusions comes from the recent crystal structure determination of the manganese form of a fungal IPNS [Roach et al. (1995) Nature 375, 700-704].

High level expression in Escherichia coli of isopenicillin N synthase genes from Flavobacterium and Streptomyces, and recovery of active enzyme from inclusion bodies.

A T7 promoter-based vector was used to express the isopenicillin N synthase (IPNS) genes of Flavobacterium sp. 12,154 and Streptomyces jumonjinensis in Escherichia coli. Most of the IPNS synthesized at 37 degrees C, and representing some 22% and 51% of the total cell protein respectively, occurred in an insoluble, enzymatically inactive form. Active IPNS was recovered in a rapid and simple two-step procedure in which the insoluble material was first denatured in 5 M urea and then refolded by passing the solubilized IPNS through a G-25 Sephadex sizing column. Further chromatography on DEAE-Sepharose resulted in highly active IPNS preparations. This procedure was found to be well suited for scaling up to produce large amounts of IPNS.

Base analogue mutagenesis in yeast and its modulation by pyrimidine deoxynucleotide pool imbalances: incorporation of bromodeoxyuridylate and iododeoxyuridylate.

Cells of the yeast, Saccharomyces cerevisiae, which are auxotrophic for thymidylate (tmpl) can also incorporate analogues of thymidylate. When the base analogue, 5-bromodeoxyuridylate, is incorporated into tmpl yeast cells it is lethal and mutagenic. Both lethality and mutation induction can be drastically altered by perturbation of the pyrimidine nucleotide pools. Analysis of mutation induction, bromodeoxyuridylate incorporation into DNA, and cell viability under various conditions revealed: (1) lethality and mutagenesis can be uncoupled, (2) thymidylate enhances mutagenesis and deoxycytidylate suppresses it, (3) mutation induction is not correlated with the magnitude of bromodeoxyuridylate incorporation into DNA. Therefore, in yeast, the pyrimidine nucleotide pools have a powerful effect on bromodeoxyuridylate mutagenesis. Both bromodeoxyuridylate and iododeoxyuridylate are extensively incorporated into the DNA of tmpl yeast cells; however, iododeoxyuridylate is non-mutagenic. Replication proceeds at the same rate in the presence of the natural substrate or either analogue. When cells are supplied with thymidylate and bromodeoxyuridylate together, there is no discrimination against bromodeoxyuridylate as a DNA precursor. However, in the presence of thymidylate and iododeoxyuridylate, there is a 3 to 1 discrimination against iododeoxyuridylate as compared to thymidylate.

Competence related proteins in the supernatant of competent cells of Bacillus subtilis.

We report that centrifugation at relatively high g-forces reduces the ability of competent cells of Bacillus subtilis to bind and take up DNA, and to be transformed. The centrifugation supernatant from competent cells restores this reduction of competence; the supernatant from non-competent cells is inactive. Phosphocellulose chromatography of centrifugation supernatants from radioactive competent cultures gave rise to six sharp peaks, together, these were shown by subsequent SDS polyacrylamide gel electrophoresis to contain over 60 different polypeptide bands. Peak II, which showed competence restoring activity, produced three polypeptides. When these bands were further examined, one of these exhibited DNA binding activity and the other two each contained a different endonuclease. Competence restoring activity was not recovered from the SDS polyacrylamide gel of peak II. The three peaks from non-competent cultures produced altogether five faint bands in gel electrophoresis. None of these bands were similar to those found in peak II.

Adsorption of bacteriophages phi 29 and 22a to protoplasts of Bacillus subtilis 168.

Adsorption of bacteriophages phi 29 and 22a to protoplasts of Bacillus subtilis 168 is described. The number of binding sites on bacilli and protoplasts is determined for each phage. Bacilli and protoplasts possess roughly the same number of sites per unit area for phi 29, i.e., approximately 700 sites per bacillus. There are also approximately 700 sites per bacillus for 22a, but only about one-third as many sites per unit area on the protoplast surface. A model for phi 29 adsorption is proposed.

Inhibitory protein controls the reversion of protoplasts and L forms of Bacillus subtilis to the walled state.

When the cell wall of Bacillus subtilis is removed by lysozyme and the resultant protoplasts are plated on hypertonic soft agar medium, each protoplast forms an L colony. L bodies from such L colonies again plate as L-colony-forming units (CFU). However, if protoplasts or L bodies are "conditioned" by 1 h of incubation in 0.4% casein hydrolysate medium and then incubated in 25% gelatin medium for 1 h, 60 to 100% of the formerly naked cells give rist to bacillary colonies. The present experiments largely explain the mechanism responsible for the "heritable" persistence of the wall-less state in B. subtilis. It is shown that protoplasts produce a reversion inhibitory factor (RIF) which blocks reversion when the cell concentration exceeds 5 x 105 CFU/ml. This inhibitor is nondialyzable and sensitive to trypsin, heat, and detergent. Efficient reversion at 2 x 107 CFU/ml is obtained if the protoplasts are treated with trypsin after conditioning and chloramphenicol is incorporated into the gelatin reversion medium. In the presence of 500 mug of trypsin per ml, the requirement for gelatin is sharply reduced, and reversion occurs rapidly in liquid medium containing only 10% gelatin. Trypsin also stimulates reversion in L colonies growing on soft agar. Latent RIF is activated by beta-mercaptoethanol. This reagent blocks reversion of protoplast suspensions at densities of 5 x 105 CFU/ml. Comparison of the autolytic behavior of B. subtilis and of the RIF revealed that several or the properties of the two activities coincide: both are inhibited by high concentrations of gelatin, both are activated by beta-mercaptoethanol, and both have high affinity for cell wall. Going on the assumption that RIF is autolysin, models for protoplast reversion is suggested by the finding that mutants with altered teichoic acid show altered reversion behavior.

Interaction of protoplasts, L forms, and bacilli of Bacillus subtilis with 12 strains of bacteriophage.

The interaction of 12 phage strains with bacilli, protoplasts, and L forms of Bacillus subtilis 168 and with eight of its mutants and two of its lysogens is described qualitatively and quantitatively. After removal of the cell wall from B. subtilis 168, 11 of the 12 phage strains can still adsorb to the protoplasts, nine kill their wall-less host cells, and five multiply in the naked bacteria, forming plaques on L form lawns. Individual gene mutations can have similarly pleiotropic effects, strongly dependent upon the plating medium. Thus, the gta A mutation, which causes loss of glucosylation of the wall teichoic acid, results in loss of wall adsorption sites for phi (but not membrane sites) and for phi105. Phages phi25, SP82G and phie can still adsorb to gta A bacilli and plaque in unstabilized and sorbitol-stabilized lawns of this mutant, but they can not plaque in sucrose-stabilized lawns. The lysogenized wild type, B. subtilis 168 (SPO2), also exhibits a pleiotropic pattern, showing different levels of resistance to phages SPO2, phi1, phie, and phi25. Its resistance pattern is very similar to that of wild-type protoplasts. On the basis of such patterns, the bacterial mutants and strain B. subtilis 168 (SPO2) could be ordered into four classes and the phage strains classified into four to six groups. Together, they form four to six interaction complexes, based partly on adsorption sites and perhaps partly on metabolic blocks in phage development.

Temperature-sensitive divisionless mutant of Bacillus subtilis defective in the initiation of septation.

A temperature-sensitive divisionless mutant of Bacillus subtilis 168, tms-12, is shown to be defective in an early step in septum formation at the restrictive temperature. The nature of this defect has been studied by comparing the growth and composition of mutant and wild-type (tms-12(+)) cells at the restrictive (48 C) and permissive (34 C) temperatures. At 48 C, tms-12 cells grow as nonseptate, multinucleate filaments. Filamentation does not appear to be a result of alterations in properties of the cell wall, since the ratio of mucopeptide to teichoic acid, the autolytic activity, and the ability of the walls to protect cells against osmotic shock are comparable in tms-12 filaments and tms-12(+) bacilli grown at 48 C. Synthesis of deoxyribonucleic acid and the segregation of nucleoids also proceed normally during filamentation. The synthesis of membrane, however, is delayed during filamentation of tms-12. No gross alterations were observed in the protein or lipid composition of membranes isolated from mutant filaments. Septum formation resumes when filaments are returned to 34 C and appears to be associated with an increased synthesis of membrane. The occurrence of septa was monitored both by microscopic observation of cross walls and by assays of the number of viable protoplasts released from bacillary filaments upon removal of the cell wall. Septation recovery can be blocked by inhibitors of ribonucleic acid and protein synthesis added during, but not after, the first 7 min of recovery at 34 C. By contrast, inhibition of deoxyribonucleic synthesis does not block recovery.

Penetration of a bacteriophage into Bacillus subtilis: blockage of infection by deoxyribonuclease.

Plaquing of a newly isolated phage of Bacillus subtilis, phage 41c, is only 2% efficient in agar containing 200 mug of deoxyribonuclease per ml. Timed deoxyribonuclease addition experiments showed that phage development is blocked in 90% of the cells if deoxyribonuclease is present during adsorption (zero-time samples), whereas 10 min after adsorption the enzyme has little effect (10-min samples). The fate of (32)P-deoxyribonucleic acid label of phage 41c in zero-time samples was compared to that in 10-min samples. In both, about 80% of the label remained with the phage-bacterium complex on initial centrifugation. However, four successive washings removed 90% of the (32)P from the zero-time samples but only 25% from the 10-min samples. In both samples, most of the washed-out label was of low molecular weight. When the time course of interruption of infection by blending was compared with interruption by deoxyribonuclease treatment, the two processes exhibited similar kinetics. It is postulated that both processes block injection at the same site, namely, the point of contact between phage tail and cell wall surface. Partitioning of (32)P label during protoplasting of zero-time and 10-min samples was similar to that observed during washing. For the protoplasting experiments, a quantitative method for plaquing protoplasts was developed. A single bacillus made of several cells can give rise to several protoplast plaque-forming units. Strain 41c was the only phage of seven tested to be inhibited by deoxyribonuclease. No other deoxyribonuclease-sensitive phages have been described.

Comparison of deoxyribonucleic acid uptake and marker integration in bacilli and protoplasts of Bacillus subtilis.

trp(+)his(-) donor deoxyribonucleic acid (DNA) was added to highly competent trp(-)his(+) recipient bacilli and to protoplasts prepared from these bacilli, and the cell-DNA complexes were incubated for 30 min. The complexes were then washed and lysed, and their DNA was analyzed on a trp(-)his(-) strain for the donor marker trp(+), the resident marker his(+), and for the recombinant trp(+)his(+) combination. The extracts of the bacillary complexes contained a normal percentage of donor markers (0.1-0.2%), and the number of trp(+)his(+) doubles (20% of all trp(+) transformants) indicated that the donor DNA had become integrated into the resident genomes. The protoplast complexes contained 10 to 1,000 times fewer donor markers and almost no recombinants. This indicated that, in protoplasts, marker uptake was minimal and recombination was absent. Uptake was also measured with (3)H-labeled DNA. On the average, protoplasts took up one-fiftieth as much DNA as bacilli. It was concluded that, probably, protoplasts took up no DNA at all, that there were no DNA affinity sites on the surface of the protoplasts, and that the residual marker and radioactivity uptake was due to imperfections in the experimental system. The data and conclusions differed sharply from earlier ones of Hirokawa and Ikeda despite the fact that the techniques of these authors were followed in repeat experiments.

Comparison of the biochemistry and rates of synthesis of mesosomal and peripheral membranes in Bacillus subtilis.

The membrane vesicle (beaded chain) portion of the mesosomes and peripheral (ghost) membrane of Bacillus subtilis were obtained by protoplast lysis and separated by differential and sucrose gradient centrifugation. Electron microscopy revealed that both fractions were satisfactorily homogeneous. Comparison of the two membrane preparations showed that they were similar with respect to total protein, total phosphorus, and lipid-soluble phosphorus content. Their protein patterns on acrylamide gel electrophreograms did not differ significantly. A possible point of distinction was revealed by a difference spectrum analysis of their cytochromes. The two preparations showed clear quantitative differences in all five of the enzyme activities assayed. Acrylamide gel electrophreograms of peripheral membrane stained for malate dehydrogenase showed four weak isozyme bands, whereas electrophreograms of mesosome membranes exhibited a single strong peak. (A survey of published data on enzymes in mesosome fractions shows a marked lack of correspondence between different species of bacteria.) Comparison of (3)H-acetate incorporation into the two membrane fractions showed that both were labeled at the same rate. Similarly, (35)SO(4) was taken up by both fractions at a comparable rate and was chased from both comparably. Lipid and protein labeling thus indicates that mesosome vesicle membrane is not a precursor or special growing point of peripheral membrane.

Gelatin-induced reversion of protoplasts of Bacillus subtilis to the bacillary form: biosynthesis of macromolecules and wall during successive steps.

Protoplasts of Bacillus subtilis plated on SDG medium formed L colonies in quantative yield and propagated in the L-form indefinitely. Protoplasts or L bodies placed in 25% gelatin medium formed bacillary colonies. Details of the reversion of these naked bodies to the walled form are reported here. Protoplasts prepared in minimal medium reverted fairly synchronously 3 to 4 hr after inoculation into gelatin, but protoplasts preincubated in casein hydrolysate (CH)-enriched minimal medium were primed to revert within 1 hr in the gelatin. Preincubation for 1.5 hr in 0.44% CH was required for good priming. Cells must be subjected to this preincubation (step 1) in the naked state; it is effective for L bodies as well as protoplasts. Priming was blocked by chloramphenicol, puromycin, and actinomycin D but was not affected by penicillin, lysozyme, or inhibition of deoxyribonucleic acid (DNA) synthesis. It is concluded that protein and ribonucleic acid (RNA) synthesis are required during step 1, that DNA synthesis is not required, and that wall mucopeptide is not made. The reversion of well-primed protoplasts in the gelatin (step 2) proceeded undisturbed in thymine-starved cells with chromosomes arrested at the terminus. It was scarcely slowed by chloramphenicol in the gelatin but was delayed about 3 hr by both puromycin and actinomycin D. Escape from inhibition occurred while the inhibitors were still actively blocking growth. Penicillin and cycloserine inhibited and lysozyme reversed reversion. Momentary melting of the gelatin delayed reversion. It is concluded that mucopeptide synthesis occurs in step 2, that concomitant RNA, DNA, or protein synthesis is not essential, but that physical immobilization of excreted cell products at the protoplast surface is necessary early in step 2. Newly reverted cells were misshapen and osmotically sensitive. Processes which confer osmotic stability after reversion (step 3) did not occur in the presence of chloramphenicol or actinomycin D.

Development of competence in thymine-starved Bacillus subtilis with chromosomes arrested at the terminus.

Tryptophan- and thymine-requiring cells of Bacillus subtilis, emerging from an amino acid starvation treatment which causes arrest of the chromosomes at the terminus, were not transformable. During subsequent incubation in a thymineless medium supplemented with amino acids, the cultures developed competence while retaining chromosome arrest. The competent subpopulation apparently shares the synchronous chromosome arrest of the bulk population. This was shown by different methods. The principal method was marker frequency analysis of the deoxyribonucleic acid extracted from a population enriched for competent cells by a column-chromatographic method. It is concluded that development of the competent state can occur in nondividing cells, and that the presence of a replication fork actively engaged in synthesis of deoxyribonucleic acid is not required for the development of this state.

Transport of donor deoxyribonucleic acid into the cell interior of thymine-starved Bacillus subtilis with chromosomes arrested at the terminus.

The chromosomes of a tryptophan(-), thymine(-) double auxotroph of Bacillus subtilis were uniformly aligned at the chromosome terminus by an amino acid starvation treatment. By subsequent incubations, the starved culture was rendered competent, while its state of synchronous chromosome arrest was maintained by thymine starvation. The competent, chromosome-arrested cells were transformed for three unlinked markers, located in two different chromosome regions. Shortly after addition of deoxyribonucleic acid, the cell walls were removed with lysozyme in a medium containing deoxyribonuclease and no thymine, and the protoplasted culture was assayed for single and double transformants. It was found that markers both near and distant from the terminus entered freely into the cell interior. There was no important difference in the relative frequency of entry of different markers between synchronously arrested cells and nonsynchronized control cultures. It is concluded that entry of a given marker into the cell interior can occur even if the replication site of the chromosome is stationary at a location distant from the locus of the resident homolog of the entering marker. A mechanism of donor deoxyribonucleic acid entry involving homology at the replication fork is excluded.

Transformation in quasi spheroplasts of Bacillus subtilis.

RECENTLY DEVELOPED DIFFERENTIAL PLATING MEDIA PERMIT THE DISTINCTION OF FOUR CELL TYPES IN INCOMPLETELY PROTOPLASTED POPULATIONS: intact, osmotically insensitive bacilli; osmotically sensitive rods; spheres with adherent wall residues, called quasi spheroplasts; and protoplasts. Such population mixtures were washed free of lysozyme, and then transforming deoxyribonucleic acid (DNA) was added. Transformation was nil in the protoplasts, very low in the residual osmotically insensitive bacilli, and markedly enhanced in both osmotically sensitive rods and quasi spheroplasts. Transformation in the latter two population fractions was reduced, respectively, by about 60% and about 80% by deoxyribonuclease treatment. DNA adhering to the quasi spheroplasts transforms these cells only if they are permitted to resume wall synthesis; when the same cells are plated on a medium where they shed the residual wall and form L colonies, no transformant L colonies are recovered. It is inferred that far-reaching or complete protoplasting blocks all entry of transforming DNA into the cell interior. This may be owing to eversion of mesosomes. Evidence that intact mesosomes may be required for DNA entry is provided by the finding that the recovery of transformants in the intact cell system is sharply reduced on plating media containing 25% gelatin. On such media, cells expel their mesosomes and 75% of them do not re-form any. Our own data and a survey of published results suggest the generalization that partial depolymerization of the cell wall by lysozyme may enhance competence, whereas its complete removal abolishes it.

Gelatin-induced reversion of protoplasts of Bacillus subtilis to the bacillary form: electron-microscopic and physical study.

Protoplasts of Bacillus subtilis plated on SD medium form L colonies in quantitative yield and propagate in the L form indefinitely. L bodies or protoplasts placed in 25% gelatin medium form bacillary colonies. Details of the reversion of naked bodies to the walled form are reported. In 25% gelatin medium, reversion begins earlier (about 50% reversion in 4 hr) than the multiplication of bacilli. Thus, virtually all the observed bacillary forms are themselves revertants and not the offspring of a few growing clones. The optimal temperature for reversion is 26 C in 25% gelatin. When cells reverting at 26 C are warmed to 40 C for 3 min, reversion is delayed markedly, whereas viability is unaffected. For electron microscopy, a dense protoplast inoculum was placed on a gelatin surface, incubated, and then fixed in situ. There was no multiplication, but crowding delayed reversion markedly. Successive events of reversion are as follows. The loose nucleoid of the protoplasts condenses in response to the gelatin medium and condenses further and further as reversion proceeds. A thin coat of wall develops around the bodies of various sizes and shapes and then increases uniformly in thickness until a wall of normal aspect is formed. Rod-shaped cells grow out from these bodies-sometimes in several directions at once. A few mesosomes begin to appear only after a thin coat of wall has been formed. These are dense, atypical structures compartmented by membranes. They are located at the cell periphery and do not seem to be in contact with the nucleoids. Quantitative estimates showed that only 20 to 25% of revertant cells or cells grown on gelatin contain even a single mesosome. The others have no mesosome at all. Mesosomes thus do not appear to play a significant role in reversion, and normal mesosome functions must presumably be performed elsewhere in the cell in gelatin-grown bacilli. The role of cell wall, its synthesis, and its chemical nature in successive steps in reversion are discussed.