The yeast Saccharomyces cerevisiae is widely employed as a host for the production of recombinant proteins. In the work presented here, expression of recombinant reporter proteins has been analysed in yeast, with the aim of determining factors that contribute to the expression performance of recombinant genes. Initially, expression of bacterial aminoglycoside phosphotransferase was analysed in yeast. This heterologous gene showed very low expression levels, resulting from a small decrease in mRNA abundance, plus a large decrease in translation efficiency. This appeared to be due to an unfavourable codon usage confirmed with an alcohol dehydrogenase reporter gene having highly favourable codon usage. Addition of a PGKl, ADHl, or CYCl promoter to a promoterless aminoglycoside phosphotransferase gene on a 2mum-based plasmid resulted in a reduced transformation efficiency and copy number, despite the product being non-toxic to yeast. This was not manifest in ARS plasmids. The decrease in copy number, represented a loss of potential product-encoding genes. Further investigation implicated a number of factors; regulatory functions, transcription and codon usage/translation. The promoter-inhibition did not correlate with promoter strength, nor did it result from transcriptional bypass of terminators. It was increased by a PGK promoter without a coupled coding sequence, but was reduced by promoter deletions in heterologous gene constructions, although not homologous PGK. The copy number reduction was also accompanied by reduced background transcription which was not apparent in deleted promoters that had increased copy number. A high codon bias gene lacking a promoter, showed both lower plasmid copy number and lower background transcription than one with low codon bias. Plasmid-borne PGK promoters were found to cause transactivation of chromosomal ADH expression, which did not occur with the ADHl promoter. This was not influenced by expression of ADH protein, nor other promoters that were tested, but could be mediated by single additional PGK promoters. The effect was influenced by some promoter deletions.

Gram-negative bacterial conjugation is a specialised replicative event that increases the population size of a plasmid during its horizontal transfer between organisms. Work reported in this thesis has focused on the genetic structure of the leading regions of enterobacterial plasmids. A leading region is the first segment of a plasmid to enter the recipient cell during conjugation. Some leading regions carry conserved loci. Examples are ssb encoding a single-stranded DNA binding protein; psiB which functions to inhibit induction of the bacterial SOS response; ardA which encodes antirestriction activity, and hok, a gene that contributes to plasmid maintenance. The objective of this work was to determine the distribution of these genes on enterobacterial plasmids and their arrangement relative to the origin of conjugative transfer (oriT). Sequences homologous to ardA were found on plasmid representatives of five (FV, B, I, K and N) out of 23 incompatibility groups tested. Mapping data shows that the ardA homologues are leading region genes, and subcloning experiments coupled with phenotypic assays provide evidence that the IncB and IncFV ardA homologues are functional antirestriction genes. Sequence data reveal that the ardA gene family has diverged by at least 60% at the nucleotide sequence level when compared to the prototype ardA gene of ColIb-F9. A non-homologous antirestriction gene, ardB, was only detected on IncN plasmids. Conserved ssb and psiB genes are known to be present on a number of enterobacterial plasmids representing nine different incompatibility groups (B, com9, I1, FI, FII, FIV, K and Y). It is reported that the leading region of an IncFV plasmid, F0lac, also carries conserved ssb-psiB genes. In common with previous findings, the genes are separated by a spacer of ~2.5kb, indicating their presence in a conserved module. Genes of the hok family of killer genes are found on the leading region of FI and FII plasmids. Southern hybridisation experiments revealed that IncI1 plasmids lack a member of the hok subfamily but contain a member of the pnd subfamily. Functional and hybridisation tests localised pnd to the trailing region, defined as the last segment of a plasmid to enter the recipient during conjugation. Hybridisation data also localised the pnd genes of B and K plasmids to be outside of the leading region. The finding that the killer genes are scattered is consistent with the concept that the ecological role of the hok gene family is in plasmid maintenance rather than in conjugation. The results show that while enterobacterial plasmid leading regions collectively contain some conserved genes, there is considerable heterogeneity in their combination and arrangement on different plasmids with respect to oriT. It is postulated that the difference reflects independent acquisition of modules during the evolution of the different plasmid types. An important finding is that leading region genes are each orientated such that the transcribed strand is the same as the transferred strand. This arrangement holds for plasmids ColIb (IncI1), F (IncFI), pKM101 (IncN) and F0lac (IncFV) and could be important for the expression of these genes. One possibility is that the genes are expressed from the incoming plasmid strand before it is converted into duplex DNA.

This thesis describes an attempt to isolate polymorphic minisatellite and microsatellite loci from the male-specific region of the human Y chromosome. Y chromosome polymorphisms are useful in the study of human evolution, since the majority of the chromosome is exempt from meiotic recombination and is passed down intact in paternal lineages. Y-specific polymorphisms can be used to identify males that are related by descent. Base substitutions and sequence insertions are ideal for such analysis, because they represent very rare events and are unlikely to recur. However, these mutations are often highly population-specific, making them of little use in the investigation of inter-population relationships, or of intra-population studies in many cases. Minisatellite and microsatellite loci, on the other hand, show variability in all populations and complement the 'unique event' markers. A systematic search of a human Y-specific cosmid library for G+C-rich minisatellites identified many repetitive sequences, including 14 minisatellites from the Yp pseudoautosomal region. However no Y-specific minisatellites were found. Such sequences must be very rare, and may not be present at all in the Y-specific euchromatin, perhaps because inter-allelic exchange events are precluded in this region. Six novel Y-specific microsatellites were cloned using hybridisation selection from a degenerate PCR library. One multi-locus and two single-locus trinucleotide microsatellites were shown to be polymorphic and were used with two other microsatellites in a survey of diversity in 340 males, and to investigate the relationships between males in different haplotypic groups. No evidence for increased African diversity of Y chromosomes was found in an intercontinental comparison. These markers were also used to classify deletion and duplication events at the 50f2/C locus in 66 males from various populations. This analysis allowed the identification of novel deletion events that had not been distinguished by previous haplotyping analysis with base substitution and other polymorphisms.

D. arizonae and D. mojavensis are sibling species with overlapping distributions. There is character displacement for sexual isolation. Speciation may have been completed through reinforcement. Part of the sexual isolation can be explained by female preference for male lovesongs. Within the D. mojavensis cluster of the repleta species group lovesongs always contain pulse song with long inter-pulse intervals (L-IPIs) and generally feature pulse song with short inter-pulse intervals (S-IPIs) giving a bimodal IPI distribution. 'Normal' song is produced by a single wing. D. arizonae and D. mojavensis race B differ significantly in mean S-IPIs but not in mean L-IPIs. Divergent character displacement for mean IPIs is absent. The species do not differ for mean burst duration (MBD). Many pairwise comparisons between stocks show significant differences in mean IPIs and MBDs. Females from a sympatric D. arizonae stock prefer simulated songs with IPIs and bursts characteristic of conspecific males and discriminate against simulated songs with IPIs or bursts resembling a sympatric D. mojavensis stock. Females from the sympatric D. mojavensis stock do not discriminate between simulated songs of the D. arizonae and D. mojavensis stocks. In both species, mean IPIs depend on temperature which accounts for between 50-93% of the variation. MBD depends on temperature in 7/8 stocks. Mean IPIs and MBDs of a sympatric D. arizonae and a sympatric D. mojavensis stock remain distinct over 20°C. For sexually mature sympatric males, mean IPIs do not vary with age for a D. arizonae stock but they shorten with age for a D. mojavensis stock. A reduction in MBD with age was observed for a stock from each species. Song characters of sexually mature males, from a single stock of each species, remain distinct in sympatry. IPIs are probably controlled autosomally between species and influenced by the X-chromosome between D. mojavensis races. The Y chromosome may influence burst length. An D. arizonae stock has a shorter mean copulation duration than a D. mojavensis stock. Acoustical communication plays a more prominent role in the evolution of these species than previously recognised.

In an attempt to elucidate the patterns of synthesis of some envelope components during the bacterial division cycle, the rate of synthesis of phospholipid and envelope proteins has been determined throughout the cell cycle of E.coli B/r. To facilitate this study, a convenient method of preparation of envelopes was devised which leads to the recovery of approximately half the membranous material of the bacterial envelope. Further characterisation of this material indicates that it is derived randomly from, and is therefore representative of the whole bacterial envelope. In the first instance the membrane elution technique devised by Helmstetter (1967) was employed, and the following results were obtained. The rate of phospholipid synthesis was found to increase continuously and exponentially throughout the division cycle, whilst the rate of envelope protein synthesis was found to be constant throughout part of the cycle which doubled in rate at a discrete time. Further fractionation of envelope proteins by SDS-PAGE showed that the rate of synthesis of approximately half the envelope proteins increased two-fold, whilst that of the other half increased three-fold. Similar results were obtained using synchronous cultures except that all envelope proteins showed a two-fold increase in rate of synthesis. Pulse-chase experiments performed with synchronous cultures showed that the class of proteins that underwent an apparent three-fold increase in rate of synthesis were also subsequently lost from the envelope at a greatly increased rate. Combining the results of all three sets of experiments it is concluded that, whilst the rate of synthesis of phospholipid increased continuously throughout the division cycle, the rate of envelope protein synthesis remained constant with discrete doublings in rate of synthesis. Some envelope proteins are lost from the envelope, probably also at a discrete time in the division cycle. The experiments performed with synchronous cultures demonstrated the appearance of a protein synthesised at about the time of termination of rounds of replication, which was only transiently associated with the bacterial envelope. Furthermore, the synthesis of this protein was greatly stimulated by inhibition of DNA synthesis caused by removal of thymine from the growth medium. The role this protein plays in the bacterial cell remains unknown.

The question of whether or not replication of the Hfr chromosome is an obligatory preliminary to transfer has remained controversial since 1963 Latterly, questions have been raised on the number of male DNA strainds which are transferred. It has also been suggested that DNA synthesis in the recipient may be necessary. The effect of a specific inhibitor of DNA synthesis on mating between sensitive parents was studied. It was found that nalidixic acid inhibited formation of recombinants and that this inhibition was apparently reversible. While NAL was found to have a slight inhibitory effect on pair formation, the major cause of recombinsuit loss appeared to be a reversible inhibition of chromosome transfer. Inhibition of DNA synthesis in the recipient by amino-acid starvation or NAL was found to lead to loss of recombinants. With the use of a method developed for quantitating DNA transfer, it was found that there was a corresponding drop in male DNA detectable in these mated females. Attempts to use density labels to indicate the state of replication of male DNA transferred into females were foiled by rapid integration of male DNA into the female genome. Male DNA thus appeared not in the heavy or hybrid positions on the CsCl gradient, which it was expected that unreplicated or once-replicated male DNA would occupy, but in the light band together with the female DNA. However, under conditions of thymine starvation this integration was not complete and some of the male DNA transferred was detected at a buoyant density consistent with single-stranded DNA. This has still to be confirmed. Use of a thymidine-i>hosphorylase negative recipient was made to label DNA in the male cell only with C thymine in a mating mixutre. It was found that mating caused a stimulus in the rate of DNA synthesis in the male population.

Conditions affecting the formation of beta-galactosidase activity in A.nidulans have been investigated. It is shown that mycelium grown on lactose or galactose as carbon sourse contains beta-galactosidase with a specific activity more than thirty times that in mycelium grown on many other carbon sources. The enzyme occurs in two forms with molecular weights of 120,000 and 450,000. A histochemical staining technique has been developed to detect beta-galactosidase activity in colonies grown on solid medium and used to isolate mutants unable to form beta-galactosidase activity. These mutants are defective at single loci, and the two forms of the enzyme probably represent different molecular states of a single polypeptide. Mutants unable to form beta-galactosidase grow on lactose on solid medium: however in liquid culture conidia of these strains do not germinate, indicating that the physiological role of the enzyme is connected with conidial germination when lactose is carbon source. The mutants have been tentatively assigned to three loci designated bgaA, bgaB and bgaC. As beta-galactosidase is not essential for the growth of mycelium on lactose, another pathway for lactose utilisation must exist. The nature of this pathway is not known. Evidence has accumulated that galactose is metabolised in two ways in the organism: via the Leloir pathway and by another unknown system. This alternate pathway is implicated in induction of beta-galactosidase, and mutants have been isolated which are unable to form the activity in response to galactose, though able to do so when grown on lactose. Mutants/contd. Mutants have been obtained which show an induction pattern in liquid culture which is different from that in solid culture. These mutants appear to synthesise beta-galactosidase constitutively on solid medium, but not in liquid culture. The difference is thought to be due to induction of the enzyme by metabolites derived from agar.

A 5 cm. radius Rowland circle vacuum spectrometer has been used to neasure the mass absorption coefficients of Be, Mg, A1, Cu and Ag in the wavelength range 7-17A. A mica crystal was used as the monochromator. The construction and alignment procedure for the spectrometer are described together with the operation of a thin-window gas flow proportional counting system for measurement of the incident and transmitted X-ray flux. Nordfors method for measuring the scattered X-ray background in the spectrometer and calculating the optimum counting times for a given absorber thickness has been employed throughout. The preparation and mounting of the absorbing foils is described, including the use of Polypropylene as a substrate. The experimental values are compared with those available from the literature and also with the 'Universal' curve of Jonsson. It is concluded that new measurements of mass absorption coefficients are desirable in all wavelength ranges and that Jonssons curve is seriously in error in the long wavelength region. Numerical values for mass absorption coefficients have been obtained from calculations using a screened hydrogen-like atomic model and compared with those produced experimentally. The fullness of agreement varies widely with the atomic shells but is considered to be adequate for such a treatment.

'Minisatellites' are a class of tandemly repeated sequences ubiquitous to eukaryotic genomes. A subset of minisatellites have been found to be highly variable in tandem repeat copy number. This variability makes such loci highly informative markers for genetic mapping, establishing family relationships in humans and other animals, and individual identification from forensic samples. One highly unstable mouse minisatellite locus, Ms6- hm, has been previously identified in mouse DNA fingerprints by crosshybridization with the human minisatellite probe 33.6. Ms6-hm showed a high germline mutation rate to new length alleles (2.5% per gamete) causing multi-allelism and heterozygosity even within inbred strains. Mice mosaic for cells carrying a non-parental allele in somatic tissue are also seen (2.8% of mice). At reduced stringency Ms6-hm detects other loci, one of which, Hm-2, also appeared to be highly unstable This work describes the characterization of this locus. Hm-2 has been cloned, localized to chromosome 9 using recombinant inbred strains, and sequenced. The repeat sequence, (GGCA)n, is short and the largest Hm-2 alleles can have over 5000 tandem repeats. Like Ms6-hm, Hm-2 shows a high germline mutation rate (3.6% per gamete). The incidence of mosaicism at Hm-2 (20.4% of adult mice) is much greater than at Ms6-hm, however, making Hm-2 an ideal locus for further study of somatic mutation events. Analysis of Hm-2 in embryonic and extra-embryonic tissues has shown that many mutation events occur early in development to produce in some cases divergence in Hm-2 genotype between the embryo and trophoblast. In others, mosaicism was shared between the embryo and trophoblast suggesting that these mutations arose very early in development before the fifth cell division following fertilization. The degree of mosaicism in 60 mosaics has been calculated and comparison of this data with computer simulations suggests that the somatic mutation rate is not constant throughout development but rather that mutations are biased towards the first two cell divisions after fertilization.

Expanded Simple Tandem Repeat (ESTR) loci provide a useful system to assess the effect of exposure to ionising radiation on the germline of male mice; however, little is known about the mutation mechanism(s) at these loci. Information about mutation processes at these loci may provide important clues concerning the damaging effects of irradiation at the DNA level. A number of approaches have been used to investigate possible mutation mechanisms. No correlation was observed between the levels of meiotic recombination and ESTR mutation rate in the germline of exposed male mice, ruling out the possibility that radiation induced mutation at ESTR loci resulted from an increase in meiotic crossing-over. The analysis of the murine scid mutation on ESTR mutation rate demonstrated that the process of non-homologous end-joining (NHEJ) is important in the stability of ESTR loci in the germline of non-exposed mice, but was unable to ascertain whether the activation of NHEJ could provide a plausible explanation for radiation-induced increases in ESTR mutation rate. A transgenerational study of the descendants of directly exposed male mice provided evidence for a long-term effect of ionising radiation on ESTR stability in the mouse germline. ESTR instability was observed in the germline of the offspring and grandoffspring of the initially irradiated males, with no evidence for a decrease in mutation rate. This analysis also provided additional information about the inheritance of ESTR instability in the mouse germline, demonstrating the possibility that the transmission of instability was epigenetic in nature, and showing that the effect could be observed after exposure to both high- or low-LET sources of irradiation. The data also showed transgenerational effects in three different mouse strains, and that there were no differences in the inheritance of ESTR instability between the male and female germlines. The work presented here provides the basis for a number of new and exciting directions to further analyse radiation-induced mutation at ESTR loci.