Cloning and mapping of the dihydrofolate reductase gene of Bacillus subtilis

The structural gene for dihydrofolate reductase (dfrA) from the Bacillus subtilis 168 chromosome has been cloned, along with the thyB gene, on a 4.5-kb insert contained on chimeric plasmid pER1. The presence of the dfrA gene on pER1 was demonstrated by showing that: (i) transformation of Escherichia coli strains RUE10(Thy-) and RUE11(Thy+) with pER1 resulted in a 60 to 130-fold increase in dihydrofolate reductase (DFRase) activity with a turnover number characteristic of that of B. subtilis and (ii) pER1-mediated transformation of trimethoprim-resistant E. coli strain D05, which overproduced a DFRase with a decreased affinity for trimethoprim, resulted in a 41-fold increase in DFRase activity with an affinity for trimethoprim similar to that of the B. subtilis enzyme. The dfrA gene was mapped to the 200 degrees region of the B. subtilis chromosome, and the gene order was established as thyB dfrA ilvA. Furthermore, the dfrA gene was shown to be linked closely (95-99% cotransformation) to the thyB gene.     

Pleiotropic effect of a rifampin-resistant mutation in Bacillus subtilis.

Rifampin-resistant (Rifr) mutants were isolated spontaneously from Bacillus subtilis strain 168. A fraction of the mutants did not grow on a minimal medium. A high concentration of one of the L-amino acids (glutamic acid, glutamine, arginine, proline, aspartic acid, or asparagine) was required to restore their growth on the medium. Further analysis of one of the mutants (strain RF 161) suggested that the mutant is unable to use ammonia as a nitrogen source and requires amino acids instead. Activity of glutamate synthase was not detected in the crude extract of the mutant. The Rifr mutation was closely located to cysA and the drug resistance was cotransformed with the property of amino acid requirement at 100% frequency. All revertants to prototrophy tested showed the rifampin-sensitive (Rifs) property. The activity of the DNA-dependent RNA polymerase of the mutant was resistant to rifampin. It is concluded that some alteration of RNA polymerase may cause absence of the activity of an enzyme involved in the nitrogen metabolism.     

Integration and expression of the human dihydrofolate reductase gene in the Bacillus subtilis chromosome

Integration of functionally active human dihydrofolate reductase (hDHFR) gene into the Bacillus subtilis chromosome was performed. The clones obtained contained 1 to 7 copies of hDHFR gene per chromosome equivalent and were resistant to trimethoprim. In cell lysates of such clones a protein with the molecular mass of hDHFR was detected. The hDHFR gene was stably maintained in all clones having this gene integrated into the bacterial chromosome, when grown under non-selective conditions.     

Glutamine-requiring mutants of Bacillus subtilis.

Two glutamine-requiring (Gln^?) mutants of $\text{Bacillus subtilis}$ SMY were deficient in glutamine synthetase activity $\text{in vitro}$. The Gln^? mutants sporulated poorly unless glutamine was provided at high concentrations. The differential rate of histidase synthesis following induction was 4- to 6-fold higher in the Gln^? mutants than in wild-type cells. In addition, glucose repression of utilization of alternative carbohydrates appeared to be partially relieved in the Gln^? mutants.     

Deoxyribonucleoside-requiring mutants of Bacillus subtilis.

A number of deoxyribonucleoside-requiring mutants (dns) of Bacillus subtilis were isolated and their growth characteristics and ribonucleotide reductase activities were compared with those of the wild type and of a dna mutant (tsA13). Both tsA13 and dns mutants required the presence of a mixture of deoxyribonucleosides for growth at 45 degrees C but not at 25 degrees C. All the mutant strains tested contained ribonucleotide reductase activity which showed heat sensitivity similar to that of the enzyme from a wild-type strain. The reductase in B. subtilis seemed to reduce ribonucleoside triphosphates in a similar manner to the enzyme in Lactobacillus leichmannii.     

Recombination-deficient mutants of Bacillus subtilis.

Two mutant strains of Bacillus subtilis Marburg, NIG43 and NIG45, were isolated. They showed high sensitivities to gamma rays, ultraviolet light (UV), and chemicals. Deficiencies in genetic recombination of these two mutants were shown by the experiments on their capacity in transformation. SPO2 transfection, and PBS1 phage transduction, as well as on their radiation and drug sensitivities and their Hcr+ capacity for UV-exposed phage M2. Some of these characteristics were compared with those of the known strains possessing the recA1 or recB2 alleles. Mapping studies revealed that the mutation rec-43 of strain NIG43 lies in the region of chromosome replication origin. The order was purA dna-8132 rec-43. Another mutation, rec-45, of strain NIG45 was found to be tightly linked to recA1. The mutation rec-43 reduced mainly the frequency of PBS1 transduction. On the other hand, the mutation rec-45 reduced the frequency of recombination involved both in transformation and PBS1 transduction. The mutation rec-43 of strain NIG43 is conditional, but rec-45 of strain NIG45 is not. The UV impairment in cellular survival of strain NIG43 was gradually reverted at higher salt or sucrose concentrations, suggesting cellular possession of a mutated gene produce whose function is conditional. In contrast to several other recombination-deficient strains, SPO2 lysogens of strain NIG43 and NIG45 were not inducible, indicating involvement of rec-43+ or rec-45+ gene product in the development of SPO2 prophage to a vegetative form. The UV-induced deoxyribonucleic acid degradation in vegetative cells was higher in rec-43 and rec-45 strains.     

Kasugamycin-resistant mutants of Bacillus subtilis.

Kasugamycin-resistant mutants of Bacillus subtilis were isolated and classified into two groups, one of which had resistance to kasugamycin in in vitro protein synthesis and mapped in the ribosomal region. The other group had no resistance to kasugamycin in in vitro protein synthesis and had weak cross-resistance to gentamicin and kanamycin. Neither group could sporulate in the presence of kasugamycin.     

Characterization of a rifampin-resistant, conditional asporogenous mutant of Bacillus subtilis.

A rifampin-resistant, conditionally asporgoenous mutant of Bacillus subtilis was isolated that sporulates poorly in Sterlini-Mandelstam sporulation medium, but that sporulates normally in modified Difco sporulation medium. Rifampin-resistant (Rif-r) and conditional asporogenous (Spo-c) phenotypes co-transformed at 100% frequency. Preliminary genetic studies indicated the Rif-r trait to lie between cysA14 and ery, a locus (rnp) common to Rif-r mutants. Ribonucleic acid polymerase from strains bearing this mutation was found to be rifampin resistant in vitro. The loss of ability to sporulate in Sterlini-Mandelstam medium was found to be corrected, to a large extent, by addition to the medium of arginine, methionine, valine, and isoleucine. Several other amino acids had small effects, whereas others had no effect at all. The restorative effect is approximately additive. Growth studies indicated that Rif-r strains grew more rapidly than the corresponding parent in minimal medium at temperatures higher than 37 C. Addition of certain amino acids to the medium resulted in identical growth rates at these temperatures. Extracellular protease and esterase activities of the Rif-r Spo-c mutant were normal. A slight difference was found in the heat sensitivity of partially purified ribonucleic acid polymerase preparations of this mutant compared to the wild type.     

Decadent sporulation mutants of Bacillus subtilis.

In decadent sporulation mutants, sporulating populations are heterogeneous: the cells reach successive chemical and physical resistances with progressively decreasing frequencies. Each decadent mutant can be characterized by the shape and slope of the curve describing the frequency of cells resistant to various agents ('the resistance spectrum'). In some mutants the resistance spectrum decreases progressively from xylene resistance to heat resistance; in other mutants it decreases rapidly between octanol resistance and chloroform resistance. Electron microscopy showed that in two mutants the majority of the cells are blocked at stages III and IV; the number of cells that develop further to reach successive morphological stages falls off progressively. In two other mutants most cells reach stage V. Cortexless spores are also frequent. One of the decadent mutations, SpoL1, was localized between aroD and acf. The phenotype of decadent mutants is discussed in terms of sequential gene activation.     

Catabolite repression-resistant mutants of Bacillus subtilis.

Mutants of Bacillus subtilis that are able to sporulate under the condition of catabolite repression were isolated by a simple selection technique. The mutants used in the present study were able to grow normally on minimal medium with ammonium sulphate as the nitrogen source and glucose as the carbon source. Studies carried out with these mutants show that there is no close relation between catabolite repression of an inducible enzyme, acetoin dehydrogenase, and that of sporulation. Certain mutants are able to sporulate in the presence of all the carbon sources tested but some mutants are resistant only to the carbon source used in isolation. It is suggested that several metabolic steps may be affected in catabolite repression of sporulation.     

Alcohol-resistant sporulation mutants of Bacillus subtilis.

About 80% of Bacillus subtilis cells form spores when grown in nutrient broth. In medium containing various short-chain aliphatic alcohols, the frequency of sporulation was reduced to 0.5%. Mutants sporulated in the presence of alcohols at a frequency of 30 to 40%. Sporulation in the wild-type cells was sensitive to alcohol at the beginning of sporulation (stage zero). Sensitivity to alcohol in the mutants was also at stage zero, even though the sensitivity was considerably reduced. This sensitivity of sporulation to alcohol is the phenotypic expression of a genetic locus designated ssa. Mutations at this locus lead to a decreased sensitivity of sporulation to alcohol without modifying the sensitivity of growth. Genetic analysis by transduction was bacteriophage PBS1 revealed that ssa mutations are near the previously described spo0A locus. ssa mutants also differ from wild-type cells in the composition of membrane phospholipids. The relative amount of phosphatidylglycerol increased, whereas the relative amount of phosphatidylethanolamine and lysylphosphatidylglycerol decreased relative to the proportions in the wild type. The distribution of fatty acids in membrane lipids is the same as in the wild type. No differential sensitivity of phospholipid metabolism to alcohol could be detected in the mutant. This work therefore reveals that the extensive, pleiotropic changes in the membranes of ssa mutants are the phenotypic reflection of alterations at a specific gene locus.     

Biofilm-defective mutants of Bacillus subtilis

Many bacteria can adopt organized, sessile, communal lifestyles. The gram-positive bacterium, Bacillus subtilis,forms biofilms on solid surfaces and at air-liquid interfaces, and biofilm development is dependent on environmental conditions. We demonstrate that biofilm formation by B. subtilis strain JH642 can be either activated or repressed by glucose, depending on the growth medium used, and that these glucose effects are at least in part mediated by the catabolite control protein, CcpA. Starting with a chromosomal Tn917-LTV3 insertional library, we isolated mutants that are defective for biofilm formation. The biofilm defects of these mutants were observable in both rich and minimal media, and both on polyvinylchloride abiotic surfaces and in borosilicate tubes. Two mutants were defective in flagellar synthesis. Chemotaxis was shown to be less important for biofilm formation than was flagellar-driven motility. Although motility is known to be required for biofilm formation in other bacteria, this had not previously been demonstrated for B. subtilis. In addition, our study suggests roles for glutamate synthase, GltAB, and an aminopeptidase, AmpS. The loss of these enzymes did not decrease growth or cellular motility but had dramatic effects on biofilm formation under all conditions assayed. The effect of the gltAB defect on biofilm formation could not be due to a decrease in poly-gamma-glutamate synthesis since this polymer proved to be nonessential for robust biofilm formation. High exogenous concentrations of glutamate, aspartate, glutamine or proline did not override the glutamate synthase requirement. This is the first report showing that glutamate synthase and a cytoplasmic aminopeptidase play roles in bacterial biofilm formation. Possible mechanistic implications and potential roles of biofilm formation in other developmental processes are discussed.     

Structure and function of alternative proton-relay mutants of dihydrofolate reductase.

Using site-specific mutagenesis, we have constructed two mutants of Escherichia coli dihydrofolate reductase (ecDHFR) to investigate further the function of a weakly acidic side chain at position 27 in substrate protonation: Asp27-->Glu (D27E) and Asp27-->Cys (D27C). The crystal structure of D27E ecDHFR in a binary complex with methotrexate shows that the side-chain oxygen atoms of Glu27 are in almost precisely the same location as those of Asp27 in the wild-type enzyme. Kinetic evidence indicates that Glu27 can indeed function efficiently in the proton relay to dihydrofolate. Even though vertebrate DHFRs all have a glutamic acid at the structurally equivalent position, the kinetic properties of Glu27 ecDHFR more closely resemble those of wild-type bacterial DHFRs than of vertebrate DHFRs. The D27C mutation produced an enzyme still capable of relaying a proton to dihydrofolate, but with the intrinsic pKa in its pH-activity profiles shifted upward to values characteristic of the more basic thiolate group. The crystal structure of the binary complex with methotrexate reveals two unexpected features: (1) the Cys27 sulfhydryl group does not point toward the pteridine-binding site, but the side chain of this residue is instead rotated 120 degrees to interact with a tyrosine side chain projecting from a neighboring beta-strand; (2) a bound ethanol molecule occupies a cavity adjacent to methotrexate. Ethanol is a component of the crystallization medium.     

Various means of integration of the expressible human dihydrofolate reductase gene into the Bacillus subtilis genome

Integration of expressible DNA corresponding to the human dihydrofolate reductase (hDHFR) gene into the Bacillus subtilis genome has been achieved in different ways. The clones obtained contained one to seven copies of this gene per genome equivalent and were resistant to trimethoprim. Clones produce a new protein coded by the integrated hDHFR gene. In all clones, the integrated DNA was stably maintained even under nonselective growth conditions.     

Porphyrin and corrinoid mutants of Bacillus subtilis.

Porphyrin auxotrophs of Bacillus subtilis can be divided into two groups. Strains belonging to the first group (hemA, hemB, or hemC) are not able to synthesize or metabolize porphobilinogen. These strains require cysteine, cystine, and methionine, respectively. Traces of aminolevulinic acid, in a hemin-containing medium, can replace the cysteine requirement in a mutant lacking aminolevulinic acid synthetase. In bacteria belonging to the second group (hemE, hemF, or hemG), porphyrin biosynthesis is blocked at later steps, and the amino acids mentioned above are not required. It is of interest that both the activity of ribonucleotide reductase and the amount of vitamin B12 were significantly lower in the first group. The addition of vitamin B12 to the medium did not promote the growth of strains examined. We assume that porphobilinogen deaminase is essential for the synthesis of corrinoids.     

Autolytic enzyme-deficient mutants of Bacillus subtilis 168.

Mutants of Bacillus subtilis strain 168 have been isolated that are at least 90 to 95% deficient in the autolytic enzymes N-acetylmuramyl-L-alanine amidase and endo-beta-N-acetylglucosaminidase. These mutants grow at normal rates as very long chains of unseparated cells. The length of the chains is directly related to the growth rates. They are nonmotile and have no flagella, but otherwise appear to have normal cell morphology. Their walls are fully sysceptible to enzymes formed by the wild type and have the same chemical composition as the latter. Cell wall preparations from the mutants lyse at about 10% of the rate of those from the isogenic wild type, with the correspondingly small liberation of both the amino groups of alanine at pH 8.0 and of reducing groups at pH 5.6. Likewise, Microcococcus luteus walls at pH 5.6 and B. subtilis walls at pH 8 are lysed only very slowly by LiCl extracts made from the mutants as compared with rates obtained with wild-type extracts. Thus, the activity of both autolytic enzymes in the mutants is depressed. The frequencies of transformation, the isolation of revertants, and observations with a temperature-sensitive mutant all point to the likelihood that the pleiotropic, phenotypic properties of the strains are due to a single mutation. The mutants did not produce more protease or amylase than did the wild type. They sporulate and the spores germinate normally. The addition of antibiotics to exponentially growing cultures prevents wall synthesis but leads to less lysis than is obtained with the wild type. The bacteriophage PBSX can be induced in the mutants by treatment with mitomycin C.     

Nucleotide sequence of the thymidylate synthase B and dihydrofolate reductase genes contained in one Bacillus subtilis operon

The nucleotide sequence of the thymidylate synthase B (thyB) and dihydrofolate reductase (dfrA) gene regions from wild-type and trimethoprim-resistant (TpR) mutant strains of Bacillus subtilis 168 was determined. The sequenced region contains two open reading frames, ORF1 and ORF2, which correspond to thyB and dfrA, respectively, and overlap by one nucleotide. The thyB-dfrA genes encode 267 and 168 amino acid polypeptides, respectively, and are present in the order of thyB - dfrA in 5'----3' orientation. This gene order differs from those which have been found in other organisms so far. S1 mapping analysis indicated that both genes were transcribed from a single promoter located upstream from the thyB gene. Thus, the genes belong to an operon. A nucleotide substitution from 'A' in the wild type to 'C' in the TpR mutant was located in the dfrA gene region, with predicted conversion of isoleucine-95 (wild type) to leucine-95 (mutant) in dihydrofolate reductase (DHFR). It is suggested that the affinity between DHFR and Tp is reduced by this alteration.     

Increased levels of dihydrofolate reductase mRNA can be measured in normal, growth-stimulated mouse fibroblasts

Levels of mRNA for the enzyme dihydrofolate reductase (EC 1.5.1.3) were determined in growth-stimulated 3T6 cells which contained wild-type dosage of the gene coding for this enzyme. As in the case of methotrexate-resistent cells having highly amplified levels of genes for dihydrofolate reductase, an increase in dihydrofolate reductase mRNA by a factor of 2–4 can be determined when cells enter the S phase. This increase is inhibited by sodium butyrate (which inhibits growth-stimulated 3T6 cells in mid G1 phase) but not by hydroxyurea (which inhibits in early S phase). We conclude that with the available methods it is possible to study the regulation of S phase-specific enzymes after growth stimulation at the level of the mRNA, even if gene amplification is not possible or desirable.