This research has investigated the effect that altered supercoiling has on in vivo expression from bacterial promoters. In vivo expression was measured using the pKO system developed by McKenney et al. (1982). Supercoil levels were decreased through the use of the antibiotic nalidixic acid and were increased using a topoisomerase I deletion mutant. Galactokinase expression from a series of related hybrid promoters, the trp promoter flanked by oligonucleotide blocks, the trp promoter flanked by linker sections and the trp promoter flanked by a gyrase binding site was measured after incubation with nalidixic acid. Expression from the pBR322 tet promoter was inhibited by 50% under these conditions. Expression from the lacUV5, trp, and tettrp promoters was essentially unaffected. A 2-fold stimulation of expression was observed from the trplac and the trptet promoters. Expression from the trp promoter when flanked by upstream or downstream oligonucleotide blocks was similar to the values observed from control plasmids. The presence of a gyrase binding site (Kirkegaard and Wang, 1981) also exerted little influence on overall expression after a decrease in supercoiling. To further study control of expression from the trp promoter, a series of plasmids was constructed which contained a synthetic lac operator at defined positions either upstream or downstream of the promoter. Downstream lac repressor binding diminished the levels of expression, while upstream binding had little effect on expression. Placement of the lac operator farther downstream decreased the level of repression observed. Plasmids were also constructed containing two lac operators, one upstream at -39 or -52 and an additional operator downstream. Galactokinase expression from these plasmids was decreased more than in those plasmids with only one lac operator. In addition a system was developed that combined the M13 bacteriophage system of Messing (1981) and the pKO system of McKenney et al. (1982) in order to measure short term changes in transcription. Gene fragments from the (beta)-lactamase gene of pBR322 and the galactokinase gene of pKO-1 were cloned into M13mp8. Single strand DNA from these constructions will hybridize with mRNA transcribed from pKO derivatives, allowing measurement of transient changes in transcription.

The available cloned examples of the intron-28S coding sequence junctions from the rDNA of Calliphora erythrocephala have been sequenced. These introns interrupt the rDNA at the same position as the type 1 intron family first detected in D. melanogaster and D. virilis. A duplication of 14 bp of the 28S rRNA coding sequence surrounds a shortened 2.8 kb version of the major genomic length class of introns. This duplication is characteristic of the type 1 rDNA intron family in D. melanogaster and D virilis. Comparison of the intron sequences from the three Dipteran flies has revealed considerable homology between the 3' intron sequences of C. erythrocephala and D. virilis. The 28S coding sequences are highly conserved between these flies. However, a region of diverence has been located and found to correspond to the eukaryotic analogue of the L1 ribosomal protein binding site of the prokaryotic 23S rRNA. A set of overlapping phage clones representing 20.5 kb of the region surrounding the calmodulin gene of D. melanogaster has been isolated from a genomic library. Mapping of the protein coding regions suggests that the calmodulin gene is divided into four exons and three introns. Intron A has separated the initiation codon from the coding regions. Introns B and C, beween amino acids 58 and 59 and amino acids 139 and 140, respectively, split two of the four calcium binding domains. The amino acid sequence of residues 1 through 139 of calmodulin has been deduced from the DNA sequence. In comparison to the mammalian calmodulin sequences, there is a single amino acid substitution at position 99, where phenylalanine replaces tyrosine. A comparison of the D. melanogaster calmodulin gene sequences to other known calmodulin genes shows from 79% to 85% homology.

The genomic organization and sequence of the 5S RNA genes of the Dipteran fly Calliphora erythrocephala have been determined. The 5S RNA genes consist of a tandem array of 480 base pair (bp) repeating units which are clustered at a single locus 4b on chromosome 2. The repeating units show only minimal sequences and length heterogeneity, and a consensus sequence for a cloned series of 5S repeat units was determined. The coding sequence for the mature 5S RNA contains a single residue change from the known gene sequences from three Drosophila species (19,30). A precursor RNA containing additional 3' sequences with some homology to the equivalent sequences of the Drosophila species (19,30) is indicated. Partial pseudogenes homologous to the extreme 3' end of the transcribed region and the adjacent termination sequence are found at two positions in the spacer. Comparison of the 5S RNA genes of C. erythrocephala to those of three Drosophila species (19,39) identified a striking series of perfectly conserved homologies identically positioned ((+OR-)1 nucleotide) within the 5' flanking DNA of all four Dipteran 5S RNA coding regions. One of the Dipteran homology blocks is highly conserved in sequence and position in all but one of the eukaryotic 5S RNA gene sequences examined (17/18 genes). A cloned library of genomic DNA sequences of C. erythrocephala was prepared and screened by differential hybridization to ('32)P-cDNA made against poly(A)('+) RNA from embryos and two stages of oogenesis to detect sequences specifically transcribed during oogenesis. Four phage clones showed transcriptional activity in specific stages of oogenesis but not during embryogenesis. Each clone hybridized to a single poly(A)('+) RNA transcript which differed in size and stage-specificity for the four clones. All four clones contain DNA sequences which are highly repeated within the C. erythrocephala genome, but the transcriptional activity, if any, of the repetitive elements in each clone is undetermined. Transcribed fragments from each clone failed to hybridize to D. melanogaster genomic DNA at hybridization conditions allowing 10% sequence divergence. Whole follicles were used in the identification of these oogenesis-specific phage clones, and the cell specificity (follicle cell or nurse cell) is, as yet, undetermined.

Low energy collisions between Rydberg atoms and neutral molecules have been investigated over a wide range of principal quantum numbers n, and for several different neutral targets. The results have been used to validate the free-electron, independent particle model of Rydberg atom collisions. Comparison between theory and experiment show that at large values of n, ionization of Rb(nS,nD) Rydberg atoms in the reaction: (UNFORMATTED TABLE FOLLOWS) Rb(nS,nD) + SF(,6) (--->) Rb('+) = SF(,6)('-) (1)(TABLE ENDS) proceeds by electron transfer from the Rydberg atom to the SF(,6) molecule. The rate constants measured for this reaction are much the same as for the attachment of free, low-energy electrons to SF(,6). Thus, Rydberg collision studies can provide information about low-energy free electron interactions. Studies of the rate constants for free ion production in the reaction: (UNFORMATTED TABLE FOLLOWS) K(nD) + SF(,6) (--->) K('+) + SF(,6)('-) (2)(TABLE ENDS) showed these to decrease sharply at smaller n, falling far below the value expected on the basis of Rydberg electron attachment to SF(,6). This behavior is attributed not to breakdown of the free-electron model, but to post-attachment electrostatic interactions between the product ions, which are formed closer to each other at lower n. Model calculations that take this electrostatic interaction into account confirm this prediction. Other Rydberg atom collision processes, such as: (UNFORMATTED TABLE FOLLOWS) K(nD) + O(,2) (--->) K('+) + O(,2)('-) (3) K(nD) + H(,2)O (--->) KH(,2)O('+) + e('-) (4)(TABLE ENDS) have been studied, as they require both the Rydberg ion core and electron to participate in the collision. Since O(,2)('-) ions formed by free electron attachment have short lifetimes against autodetachment, the observation of long-lived O(,2)('-) reaction product suggests that the K('+) core ion plays a role in stabilizing the excited O(,2)('-) ions formed by Rydberg electron attachment. Stable KH(,2)O('+) ions cannot be formed in two-body K('+) - H(,2)O interactions. The detection of long-lived KH(,2)O('+) ions thus demonstrates that the Rydberg electron can play an important role in collision processes involving the core ion by serving as a third body to carry off excess energy.

The effects of productive adenovirus infection on host gene expression were studied using a line of methotrexate-resistant HeLa cells containing amplified amounts of the gene coding for dihydrofolate reductase (DHFR). These cells contain several hundred copies of the DHFR gene and overproduce DHFR RNA and protein. Synthesis of DHFR is induced early in adenovirus infection, before the onset of viral replication, and is suppressed late in infection along with general host protein synthesis. These changes in DHFR protein synthesis are accompanied by changes in both the steady state cytoplasmic levels of DHFR mRNA and the rate of appearance of DHFR mRNA into the cytoplasm. These data indicate that nuclear events are important in the control of DHFR gene expression during productive infection. Transcription of DHFR-specific sequences continues at a constant rate throughout infection. In contrast, steady state nuclear levels of DHFR-specific sequences change in correspondence to the changing rate of appearance of DHFR mRNA into the cytoplasm through the course of infection. Thus, constitutive transcription of DHFR sequences is followed by a differential accumultion of DHFR nuclear RNA at different times during infection, indicating that post-transcriptional events are important in the regulation of DHFR gene expression by adenovirus. Sepcifically, the regulation of DHFR gene expression during infection is accomplished by a modulation of the relative stability of DHFR-specific sequences through the course of infection. During induction, DHFR sequences are more stable than in either uninfected cells or infected cells late in infection. The induction of DHFR is temporally linked to viral replication during both wild type infection and infection with defined early adenoviral mutants which delay replication. Studies with mutant defective in the production of E4 or E1b gene products indicate that E4 gene products are involved in the switch from early to late viral gene expression during the time of DHFR induction, and that both E4 and E1b gene products are necessary for the efficient shut off of DHFR and other host genes late in infection.

We have used quasi-elastic neutron scattering techniques to study the dynamics of solvent water protons in H$\sb 2$O solutions of deuterated poly(ethylene oxide) (Polyox) at room temperature. Using a slightly modified technique, we also have obtained values for the mean-square vibrational amplitude of polymer protons in powder and D$\sb 2$O solutions of Polyox and trypsin at temperatures between 300K and 75K. To study the dynamics of solvent water in deuterated Polyox (dPolyox) solutions, we used a high energy-resolution spectrometer ($\Gamma\sb{\rm res}\approx$ 100$\mu$eV) so that we could measure the width of the quasi-elastic line for various values of the momentum transfer, Q. The Q-dependence of the quasi-elastic linewidth shows that the diffusive properties of water in these Polyox/water solutions is slightly affected at low (5 polymer wt%) and intermediate (10-20 wt%) polymer concentrations. At high polymer concentration (38 wt%), the water clearly has a reduced diffusion coefficient and an increased residence time compared to bulk water and this may be attributed to the nature of the interaction of the water molecules with the polymer. We were motivated to study the mean-square vibrational amplitude of protons in powder and D$\sb 2$O solutions of Polyox and trypsin because of recent computer simulations on biological macromolecules. We measured the vibrational mean-square amplitude by using a low energy-resolution spectrometer ($\Gamma\sb{\rm res}\approx$ 700$\mu$eV) to measure the intensity of the quasi-elastic peak in a relatively short time compared to standard high energy-resolution techniques. Our results on these 20 wt% solutions at 300K are consistent with conventional quasi-elastic neutron scattering results on the Polyox and with computer simulations of trypsin in solution. The vibrational amplitude was proportional to temperature throughout the temperature range studied, and the proportionality constant is interpreted in terms of an effective spring constant for the motion of the hydrogens. These spring constants are smaller than those for some common chemical bonds indicating that there are significant contributions to the motion of the hydrogens besides vibrational motion.

To determine the mechanisms that the yeast S. cerevisiae uses to adapt to osmotic stress, mutants were isolated which were impaired in their ability to accumulate glycerol in response to a high osmolarity challenge. These mutants identified four genes, HOG1 through HOG4 (High Osmolarity Glycerol Response). HOG2 was cloned from a genomic library on a centromere vector by complementation of the high osmolarity sensitive growth (Osm$\sp{\rm s}$) phenotype of hog2-1. Integrative confirmation of the genetic linkage between the cloned DNA and the locus of the original point mutation was performed. The null mutant hog2-$\Delta$3, created by deletion of the gene, had phenotypes similar to those of hog2-1. In addition to being Osm$\sp{\rm s}$, hog2-$\Delta$3 mutants are heat shock-sensitive, fail to accumulate glycogen and are defective for sporulation. We sequenced HOG2 and determined that this gene encodes a 102 kDa protein, the N-terminus of which contains a region of 350 amino acids showing 38% identity to CIF1(GLC6), a gene required for metabolic responses to glucose. HOG2 encodes trehalose-6-phosphate phosphatase. HOG2 mRNA levels increase in response to heat shock and osmotic stress. Yeast cells containing a HOG2::lacZ fusion showed a four-fold increase in $\beta$-galactosidase one hour after exposure to 0.5 M NaCl, but no increase was seen after a heat shock at 42$\sp\circ$ for one hour. The pleiotropic phenotype of hog2$\Delta$ mutants suggests that HOG2 is a regulatory gene necessary for adaptive responses to a variety of environmental stresses. We propose that the promoter region of HOG2 contains two separate stress elements in the Upstream Activating Sequence (UAS). We suggest that the HOG2 promoter region located between nucleotides $-$1 and $-$400 of the HOG2 coding region contains a novel osmotic stress element which is induced in response to an increased extracellular osmolarity. The inducement of this osmotic stress element is not dependent on the presence of the TRS, the heat shock inducible element, found farther upstream.

Two experimental approaches to dissociative resonance Raman spectroscopy, Raman emission spectra and Raman excitation profiles, have been applied to investigate the photodissociation dynamics of ozone in the Hartley and Huggins absorption bands. Resonance Raman spectra of ozone at 266 and 270 nm excitation wavelengths have been measured up to 11000 cm$\sp{-1}$ with improved accuracy. Several bands beyond the dissociation limit have been observed. All prominent bands observed are well fit by an analytical Darling-Dennison model. A new analytical two-dimensional potential energy surface for the electronic ground state of ozone has been constructed based on the new experimental data. Resonance Raman spectra with excitation in the Huggins band were also measured. The symmetry issue of the electronic state responsible for Huggins band is addressed. A novel continuously-scanned Raman excitation profile technique has been applied to the red wing (276-297 nm) of the Hartley band. Several Raman excitation profiles for the lowest three vibrational bands have been determined. These spectra show characteristic oscillatory structures. We also report the discovery of the chemiluminescence of O$\sb2$ and OH following the photodissociation of ozone.

The biodegradative arginine decarboxylase and lysine decarboxylases, encoded by adi and cadA, respectively, are induced to maximal levels when E. coli is grown anaerobically in rich medium at acidic pH. They were used as model systems to study how E. coli responds to a specific environmental stress, moderately acidic pH. Two Mu dlac fusion strains, GNB7145K (adi::lac) and GNB8385K (cadA::lac), were used as genetic tools in our search for transregulating factors involved in acid-induction of these two gene systems. By characterization of random transposon mutagenesis-generated mutants that showed altered expression patterns in response to medium pH, and by studying a series of plasmids that could complement the mutant phenotypes, we identified six genes whose products might be involved in the acid-induction processes. Four of them are likely to be responsible for the basal level of expression at noninducing conditions, three of which, hns, hnsB, and hfq, code for known or putative bacterial histone-like proteins, and one of which, leuO, codes for a putative transcription regulator. The other two, rpoA encoding the $\alpha$ subunit of RNA polymerase, and cysB encoding the positive regulator for cysteine biosynthesis operons, are likely to be involved in the adi activation process at inducing conditions. These findings have allowed us to speculate upon the mechanisms of acid-regulation, particularly of the adi gene system where the regulation seems complex and little previous information was available.

The core domain (residues 62-323) of the regulatory protein lac repressor has been aligned to several sugar binding proteins of known structure. The overall homology based on two separate matrix scoring systems (minimum base change per codon and amino acid homology per residue) is significant. Similarly, the predicted secondary structure of the repressor exhibits excellent agreement with the known secondary structures of the sugar binding proteins. Using this primary sequence alignment, the tertiary structure of the core domain of lac repressor was modeled using the structures of the sugar binding proteins as templates. Further refinements of this model using the purine repressor provided improvement in regions not well defined based on the sugar binding proteins. Important residues involved in operator and sugar binding and in protein assembly have been identified using genetic methods, and placement of these residues in the model is consistent with their known function. The recent solution of the crystallographic structure confirms the elements of this homology model. This approach provides an effective means to visualize the core domain of the lac repressor and to interpret the mutational data for specific residues. The availability of models of this type provides a structural basis for rational design of experiments. Mutations that affect both monomer-monomer and dimer-dimer subunit interfaces have been combined to generate a new family of mutant proteins designed to explore the role of subunit interactions in this regulatory protein. Combination of apolar substitutions at residue 84 and C-terminal deletions to generate dimer results in mutant repressor proteins with increased stability, even in the presence of Tyr282Asp, a mutation that generates monomer in the wild-type background. Operator binding studies demonstrate the linkage between dimer formation and operator binding and confirm that dimers are the minimal unit required for repression. The nature of the side chain at position 84 not only influences subunit association and stability, but also appears to contribute to subunit communication and DNA binding domain orientation. Placement of the sites of these mutations in the homology model of the lac repressor is consistent with the characteristics found for these proteins.