RNA molecules play a major role in cellular processes such as replication, transcription, and translation. As a result, RNA-based engineering methods have emerged as important tools in biotechnology. However, the structure and function of RNA depends on global interactions, which often prevents the use of a modular design strategy, particularly with allosteric conformations. Using RNA secondary structure prediction tools, computational methods can successfully design RNA switches that work in E. coli. The overarching aim of my research is to develop synthetic RNA switches that could be used for regulation and sensing of molecules in living cells. We have engineered RNA based synthetic signal transduction cascade consisting of a single RNA molecule (regazyme, an RNA chimera of an aptazyme with a riboregulator) that upon sensing a ligand (theophylline) self cleaves and releases a riboregulating small RNA. This small RNA binds to a cis-repressed mRNA allowing translation of a reporter protein. This system can be adapted to be induced by other ligands and can be used as a biosensor. I have also integrated a riboregulated RNA switch into the E. coli genome to study its behaviour at single-cell level. This reduces the transcriptional and translational noise in data collection to inform more accurate computational design of RNA regulatory units. We used computational design to engineer higher-order RNA-triggered riboregulators organized as a hierarchical toehold activation cascade. This has been studied in a single cell as well as in a population of E. coli cells. These RNA riboregulators can be used for construction of new, complex and portable synthetic gene circuits. In addition, I have engineered sense and antisense riboregulators consisting of the small RNA reverse complement of a known riboregulator. This riboregulator can transcribe two small RNAs from the same DNA template depending on the direction of transcription. These two small RNAs independently trans-activate translation of their cognate target genes and both RNAs also silence each other by antisense interaction. We have also engineered an RNA-based tunable antiterminator, a TNA-derived adaptor that acts as a signal converter in a genetic circuit, converting a translation signal to a transcription signal (unpublished). I have engineered a minimum alphabet riboregulator that has only three nucleotides (GCU) that currently validating (unpublished). In order to explore the use of directed evolution for the engineering of RNA switches, I am developing an evolution-based system for generation and selection of new biomolecules. These evolved new biomolecules could be used in future medical applications such as molecular sensing. I have been using T7 and P2 bacteriophages as the basis for this evolutionary system. I have engineered the genome of the T7 phage with (regazyme, Riboswitch and riboregulators) using homologous recombination with marker-based selection. These engineered phages can be used to evolve new biomolecules such as other regazymes, riboswitches and riboregulators.

No description available.

Haemopoietic stem cells (HSC) maintain lifelong haemopoiesis through the monitoring and production of cells from multiple haemopoietic cell lineages. A key property of HSC is their ability to maintain quiescence. Quiescence refers to a state of inactivity in which the cell is not dividing and remains dormant. It is this property of the HSC that is thought to maintain genomic integrity and to allow the HSC to sustain haemopoiesis over the period of a lifetime. However, the regulation of quiescence in this context is not well understood. Numerous studies have aimed to understand the molecular mechanisms underlying HSC quiescence using high-throughput approaches. A previous microarray study by our group aimed to understand the transcriptional differences between quiescent and proliferating human HSC. Data from this microarray showed that the most up regulated group of genes in quiescent compared to proliferating human HSC were chemokine ligands, specifically within the CXC family. Although this was a novel finding at the time, the biological function of these chemokine genes was not studied until the current work presented here. In this thesis, we aimed to extend foregoing research and importantly, investigate the role of CXC chemokines in HSC properties, using both human and mouse systems. First, we validated the results from the microarray study using gene expression analyses to show that chemokine ligands CXCL1 and CXCL2 were significantly up regulated in quiescent HSC (CD34+CD38-) in comparison to more proliferative progenitors (CD34+CD38+). Focusing on CXCL1, we showed positive expression of the ligand protein in human stem/progenitor cells using immunofluorescence and western blotting on human primary CD34+ cells. In addition, we identified positive expression of receptor CXCR2 by gene and protein analyses on CD34+ cells, indicating the presence of an autocrine chemokine signalling loop. To determine the biological function of CXCL1/CXCR2 signalling in human HSC, we used shRNA to reduce CXCL1 expression and a commercially available inhibitor (SB-225002) to block CXCR2 receptor signalling. Experiments on cell lines expressing CXCL1 and CXCR2 (HT 1080) showed that reduction of CXCL1 and over-expression reduced or increased cell viability and proliferation respectively. Experiments on human primary CD34+ cells revealed that reduction of CXCL1 induced apoptosis and reduced colony formation. Similarly, inhibition of CXCR2 signalling in CD34+ cells using SB-225002 induced apoptosis and reduced colony formation in a dose dependent manner. However, due to human sample availability and technical challenges, experiments need repeated in order for a valid conclusion to be made and statistical analysis could not be carried out for some primary experiments. In addition, further experimental work is required to conclusively prove that human stem/progenitors express CXCL1 and CXCR2 as different techniques showed varying results. In summary, we provide some evidence that CXCL1 and CXCR2 is expressed by human HSC and may be an important survival pathway in normal human HSC which requires further experimental data to provide valid conclusions. In order to gain a deeper understanding of the biological function of chemokine signalling in HSC biology, we used an in vivo murine system. First, we examined mRNA transcripts of CXC chemokines in mouse HSC populations. We screened a small selected group of CXC chemokines using primitive mouse HSC and single cell quantitative PCR using the Fluidigm™ platform. Gene expression analyses identified that Cxcr2 and Cxcl4 mRNA transcripts were detected including in the most rare, primitive HSC fraction. To elucidate the mechanism of action, we used a transgenic reporter and knock out mouse models for both genes of interest. Analysis of a Cxcr2 null mice model (Cxcr2-/-) validated previous research in which animals lacking Cxcr2 show disrupted haemopoiesis with an expansion of myeloid cells in the haemopoietic organs. Interestingly, within the current work, analysis of steady state haemopoiesis revealed an expansion of the most primitive HSC in the BM of animals lacking Cxcr2 and enhanced mobilisation demonstrated by an increase in the stem/progenitor activity in the spleen and PB. HSC functional analyses using BM reconstitution assays with wildtype (WT) or Cxcr2-/- HSC showed that there was a trend towards a reduction in engraftment in animals transplanted with HSC lacking Cxcr2. However, this result was not statistically significant due to high sample variability and due to time constraints and the length of this assay, this was not repeated. The data suggests that Cxcr2 expressing HSC may be important for stem cell maintenance via a cell autonomous mechanism however experiments are required to be repeated to draw valid conclusions. Cxcl4-Cre transgenic mice containing a RFP construct under the control of the Rosa26 promoter (Cxcl4-Cre) showed RFP expression in HSC and progeny. RFP expression in HSC populations was in accordance with Cxcl4 mRNA transcripts therefore suggesting RFP expression was correlated with endogenous Cxcl4 expression. Interestingly, flow cytometry analysis identified that not all (~50%) HSC showed positive expression for RFP. Flow cytometry sorting of positive and negative populations revealed that cells with enhanced colony formation potential reside within the RFP (Cxcl4) positive fraction. To extend this data, we aimed to knock out and reduce Cxcl4 expression and examine the phenotype. Targeted deletion of Cxcl4 in vitro using a Cxcl4 shRNA vector demonstrated that Cxcl4 reduction in vitro diminished colony formation in primary and secondary replating assays. Since data for human CXCL4 mRNA were not conclusive from the original microarray, we reassessed the relevance of CXCL4 in the human system. Gene expression analyses showed that CXCL4 transcripts were indeed detected and furthermore, up regulated in primitive HSC (CD34+CD38-CD90+) compared with proliferative progenitors (CD34+CD38+). Collectively, the data indicates that CXCL4 may play an important role in mouse and human HSC biology, however further experimental work is required to address this. In summary, the data presented in this thesis demonstrate that several chemokines including CXCL1, CXCL4 and receptor CXCR2 may have key roles in HSC survival and maintenance, both in the mouse and human systems. However, increased biological replicates and further experiments are required to draw valid conclusions. Enhanced understanding of the regulation of stem cell properties is critical for improving our ability to manipulate normal stem cells in vitro and in vivo. Furthermore, understanding normal stem cell regulation is fundamental for the research of diseases such as leukaemia in which leukaemic stem cells are less sensitive to drug treatment.

Histone modifications form an important part of the epigenetic landscape that controls many aspects of cellular function, including regulation of gene expression and cell differentiation. The persistence and inheritance of many of these modifications through the cell cycle and differentiation are still unknown. Here, I show global epigenetic karyotypes of metaphase chromosomes labelled to highlight specific marks. Metaphase is transcriptionally inactive and so epigenetic marks here are not simply reflective of gene transcription. I found that histone marks such as H3K27me3 are inherited through differentiation, whereas others such as H4K20me3 have re-organised distributions. FISH analysis allowed the alignment of genetic features with H3K4me3 and H3K27me3 distributions, showing that these marks are correlated with increased gene density, revealing a deeply intertwined distribution in ES cells, indicating bivalency. Focusing on the Hoxa cluster using N-ChIP in ES cells allowed the analysis of histone modification prevalence at the single gene level in ES cells. Most histone modifications remain stable between G1/S phase and G2/M phase, although H3K9ac decreases in ES cells at G2/M. Results for bivalent modifications show permissive chromatin environments denoted by high H3K4me3 and low H3K27me3 methylation at gene promoters that are expressed soon after the onset of differentiation, denoting a predictive chromatin signature. This signature was altered after five days of differentiation, where H3K4me3 increases and H3K27me3 decreases at most Hoxa promoters, concomitant with the rise in expression of some Hoxa genes, displaying the dynamic properties necessary to represent a mechanism for control of transcription during differentiation.

The expression of genes is tightly controlled, predominantly at the point of transcription. RNA polymerase (RNAP) must first bind to a deoxyribonucleic acid (DNA) promoter upstream of a gene to transcribe it. However, the ability of RNAP binding is dictated by the core promoter DNA sequence, the presence of transcription activator or repressor proteins and numerous other factors. The strength of promoters has been indirectly measured. Only a few studies have attempted to directly address the RNAP flux through transcription units, and further studies are still required. In the current study, the aim was to directly correlate RNAP gene transcription with the strength of core promoter elements. To do that, I employed the direct method of chromatin immunoprecipitation (ChIP), followed by quantification of immunoprecipitated DNA. For promoter regions, this method directly measures the occupancy by RNAP; for regions within transcription units, the flux of the RNAP was deduced. A range of semisynthetic promoters, with different combinations of core promoter elements to obtain different levels of expression, was used to validate our method. This direct method enabled the calculation of “promoter competitivity”, “promoter occupancy index” (POI), RNAP “escape index” (EI), “fragment occupancy percentage” (FOP) and the time interval between transcribing RNAPs (Tint). On the basis of Tint, the number of RNAPs crossing any DNA sequence of interest per second (polymerase per second; PoPS) was calculated. Surprisingly, the results of the present study revealed that the RNAPs are well separated during transcription of the \(lac\) operon.

In the post-genomic era the availability of genome-wide datasets has revealed an unexpected complexity of transcriptional regulation. In this context, where most enhancer predictions are based on computational analyses, functional validations are lacking. This thesis investigated the utility of the transgenic zebrafish embryo as an in vivo vertebrate model to study the function of candidate human enhancers, and detect subtle changes in enhancer function caused by disease-associated variants. Our functional validations indicated that despite the evolutionary distance between human and fish, 60% of the conserved enhancers predicted by a combination of chromatin signatures, TF binding events and bidirectional transcription, lead to reporter expression that recapitulates the patterns of either zebrafish or human genes. To improve the reliability of zebrafish transgenesis, a targeted integration system mediated by PhiC31 integrase was validated for enhancer testing. I demonstrated that this method overcomes position effect variation commonly found in transposon-based assays. However, enhancer-driven expression could not be detected when I attempted to quantitate TCF7L2-associated enhancer variants, indicating the need for further studies to understand the limitations of the zebrafish model. Taken together, my results provide strong support for zebrafish as a valuable in vivo model to study the function of mammalian transcriptional regulatory elements.

Human essential hypertension is a highly heritable disorder with complex aetiology and is a major risk factor for cardiovascular events such as ischaemic heart disease and stroke. A combination of multiple environmental and lifestyle factors contribute to blood pressure variation alongside a strong genetic component. Only a small proportion of the genetic factors that regulate blood pressure in the population are currently known, although there is strong evidence that the adrenal cortex and the steroid hormones it produces contribute. Several research strategies have been utilised to dissect the genetic aetiology of hypertension, including candidate gene studies and association studies. Two recent genome-wide association studies aimed to identify variations associated with altered blood pressure and hypertension. A total of ten variants were identified with genome-wide significance after a combined analysis between the two consortia, including a polymorphism located within intron 3 of the CYP17A1 gene. This variant was reported to be associated with a systolic blood pressure increase of 1.16 mmHg. The CYP17A1 gene codes for a dual-function enzyme (17α-hydroxylase/17,20 lyase) expressed primarily in the adrenal cortex and gonads that plays a key role in the steroidogenic pathway. Mutations in its coding region and splice sites are known to cause a rare form of congenital adrenal hyperplasia, suggesting that more common genetic variations at this locus might result in more subtle effects on blood pressure. A detailed examination of the variation across the CYP17A1 locus was required to establish patterns of linkage disequilibrium and is presented in Chapter 3. Some information on the polymorphic variation in this region was already available in public databases but precise details on linkage disequilibrium and the corresponding haplotype blocks were lacking. The entire CYP17A1 gene was directly sequenced, including approximately 2.5kb upstream from the transcriptional start site, in 62 subjects drawn from a normotensive population. Polymorphic variations were identified mainly in the promoter and introns. Two seemingly unrelated blocks of SNPs were identified as being worthy of follow-up investigations, particularly those located in the promoter region, as these could be responsible for alterations in the transcriptional activity of the gene. A total of seven promoter polymorphisms were then genotyped in a larger hypertensive population where the relationship between SNPs was less clear. In Chapter 4 the effect of CYP17A1 genotype on intermediate corticosteroid phenotype is explored in a hypertensive population. Corticosterone, cortisol and androgen production were not significantly altered in the population when stratified by genotype for each polymorphism. However when further split by gender, increased cortisol excretion rates were found to associate with the minor allele at position -362 in males and at positions -1204 and -2205 in females. Ratios of various corticosteroid intermediary metabolites were also compared as indices of CYP17A1 enzymatic activity. Ratios of THDOC:THS were significantly reduced in the presence of the minor allele at positions -34, -1204 and -2205, suggesting increased 17α-hydroxylase efficiency. In addition, aldosterone excretion was significantly elevated in individuals with CC genotype at position -1877; an indirect genotype-dependent effect has been speculated. A bioinformatic search was conducted to identify putative transcription factor binding sites at the polymorphic locations. This is presented in Chapter 5. This confirmed the hypothesis that single base changes at each of the seven polymorphic sites could lead to altered transcriptional activity. Using reporter gene assays in vitro, the G allele at position -362 (rs248658) associated with greater transcriptional activity than the A allele. The T allele at position -1877 (rs138009835) was transcriptionally less active than its alternative C allele. Similarly, the C allele at position -2205 (rs2150927) showed lower activity than the T allele. These data provide strong evidence that common variation at this locus may be of functional significance. The studies in Chapter 6 investigate a potential regulatory role of microRNA (miRNA) at the CYP17A1 locus. MiRNAs are a class of small non-coding RNA molecules that have recently emerged as novel post-transcriptional regulators of gene expression. They function by targeting the 3’ untranslated region (3’UTR) of specific mRNAs and cause repression either through mRNA destabilisation followed by degradation, or by mRNA translational repression. Previous research utilised a siRNA approach to knock down Dicer, a protein required for miRNA maturation, and noted significantly increased CYP17A1 mRNA levels in the H295R human adrenocortical cell line. The investigation presented here cross-referenced bioinformatic analysis with microarray expression data in order to predict which adrenal miRNAs are most likely to regulate CYP17A1 expression. Predicted miRNAs also shown to be differentially expressed between normal and diseased adrenal tissue were then selected for further analysis. In vitro investigation involved artificial manipulation of the specific miRNA levels in H295R cells followed by measurement of CYP17A1 mRNA levels. Increased amounts of hsa-miR-320a significantly raised CYP17A1 mRNA levels, although subsequent reporter construct assays showed that this was not due to direct miRNA binding of the CYP17A1 3’UTR. The studies in this chapter are the first to demonstrate miRNA-mediated regulation of CYP17A1 expression. In summary, this work aimed to investigate polymorphic variation in the human CYP17A1 gene and its association with hypertension. Patterns of linkage disequilibrium across the CYP17A1 gene were identified and the association of several polymorphisms with intermediate corticosteroid phenotype examined. The functional effects of candidate polymorphisms have also been assessed in vitro. Further studies will be required to determine whether observed changes in transcriptional activity are the direct result of altered transcription factor binding at polymorphic sites. Finally, the role of miRNA in the post-transcriptional regulation of CYP17A1 has been confirmed.

Maternal antibodies to placental-specific antigens were not detected in third trimester maternal sera or in concentrated material adsorbed from maternal sera on a placental affinity column using immuno-fluorescence and gel precipitation techniques, A method was described for the dissociation of antigens of less than 100,000 daltons from soluble antibody-antigen complexes. This method was applied successfully to the dissociation of albumin-anti-albumin complexes formed in antigen excess. Application of this method to pregnancy serum followed by immunisation of rabbits with the dissociated material resulted in anti-SP1 response detectable on AACE. The molecular weight of SP1 is approx, 100,000 daltons and the bulk of SP1 was detectable in the retentate at the end of the dissociation. This argued against a conclusion that SP1 had been dissociated from immune complexes. Immunisation of rabbits uith soluble extract of a placenta which had been adsorbed over anti-human serum column induced AACE-detectable responses to very few serum proteins, PAPP-A, HPL, SP1 and several unidentified antigens found in placental tissue. Placental soluble extract which had in addition been adsorbed over anti-human lung column and anti-human amniotic fluid column induced responses to very few serum proteins, PAPP-A and HPL, Using these latter antisera two other antigens were identified in placental soluble extract (antigens a and b) which were not detectable in lung soluble extract or in non-pregnant serum. Antigen was also detected in pregnancy serum. By contrast, immunisation with unfractionated placental soluble extract induced extensive anti-serum response, anti-HPL and anti-SP1 but no detectable anti-PAPP-A response. The extent to which several serum proteins were denatured by a variety of commonly used dissociants of antibody-antigen complexes UBS assessed by changes in the protein's precipitating behaviour on one-dimensional AACE. AACE was found to be superior for this to double immunodiffusion. The smaller proteins studied were more resistant to irreversible denaturation, but the sample size was too small to seek a general relationship between molecular weight and denaturation. All proteins examined proved resistant to one hour exposure to 1o5 n KI-PBS and to pH 11 buffer. 1o5 n KI-PBS was used in the purification of PAPP-A by antibody affinity chromatography followed by DEAE-cellulose ion exchange and Sepharose 6B gel filtration. Double AACE arcs formed by PAPP-A in term maternal serum were also formed by partially-purified PAPP-A indicating that the antigenic variants had not been separated. The purification product contained PAPP-A which was 427-fold purified in terms of immunological reactivity on AACE, The high degree of purity was also suggested by studies with the radiolabelled purification product, 125I-PAPP-A was shown to have similar molecular weight and electro-phoretic mobility to maternal serum PAPP-A, Immune-precipitated [125]I-PAPP-A was analysed by 5% and 3% SDS-PAGE and found to contain a major radioactive component of approx, 180,000 daltons and a minor component of between 74,-93,000 daltons. The nature of the minor component has not been determined. It may be tentatively concluded that PAPP-A (mol. wt. 750,000) contains polypeptide subunits of approx. 180,000 daltons and may therefore be composed of up to four such subunits, A collaborative clinical study in which PAPP-A was assayed in the blood of women during the third trimester of pregnancy by AACE suggested that fetal sex may affect PAPP-A levels with males giving rise to higher levels than females. No significant difference in the levels of PAPP-A was detected when patients with babies affected with intrauterine growth retardation were compared with controls. No significant change in the mean concentration of PAPP-A was detected in the group of patients with gestational diabetes, but some very high values were found in this group. In the group of patients with insulin-controlled diabetes the mean concentration of PAPP-A was significantly reduced.

Although marine microorganisms drive the major biogeochemical cycles in marine ecosystems, there is a dearth of information on interactions between phototrophic and heterotrophic organisms co-occurring in oceanic waters. The aim of this project was to study these interactions using Synechococcus sp. as the model phototroph – a cosmopolitan and highly abundant member of the picophytoplankton. Heterotrophic bacteria most-frequently present in non-axenic Synechococcus sp. cultures, were identified by PCR screening using primers targeting the 16S rRNA gene. Members of the Nitratireductor, Rhodobacteraceae, Muricauda and Phyllobacteriacae genera were present in more than half of all the cultures tested (Chapter 3). Using a member of the Rhodobacteraceae as the model heterotroph, specific metabolites present in axenic cultures and co-cultures were analysed (Chapter 4). Much lower concentrations of these specific metabolites were present in the milieu of Synechococcus – Roseobacter co-cultures compared to axenic Synechococcus cultures as discovered by LC-MS. Natural product database searches suggest that these may be a group of novel compounds. A Synechococcus sp. WH7803 null mutant in the gene encoding a type III polyketide synthase was constructed (Chapter 5). A targeted exometabolomic analysis showed a decreased production of the metabolites identified above in the mutant strain compared to the wild type. Growth was considerably affected in the T3 PKS mutant and T3 PKS mutant culture supernatants had a stronger negative growth effect on a range of picocyanobacteria and green algal species than the wild type extract. Further research is required to establish the precise biological function of the observed molecules, their biosynthetic pathway and their function in the natural environment. Improving our understanding of interactions between environmentally important microorganisms not only helps us to learn more about how biogeochemical cycles in the ocean function, but can also provide new natural products for use in the pharmaceutical industry.

An investigation of amino acid metabolism in Streptomyces coelicolor, including the anabolism of tryptophan, histidine, the branched-chain amino acids and proline, as well as the catabolism of the latter, is reported. The experiments reported herein were conceptually conceived within a functional genomics framework. For this purpose the complete genome sequence of S. coelicolor was systematically exploited. Moreover, the current knowledge on the physiology of Streptomyces was taken onboard, as well as the prevailing and emerging notions on the evolution of proteins and metabolic pathways. Some of the results obtained using S. coelicolor as a model organism were expanded to other actinomycetes, such as Mycobacterium tuberculosis. This was aided by a comparative genomics analysis of the actinomycetes whose genomes have been sequenced. The theoretical principles that give support to this thesis are introduced in Chapter 1. This study was greatly facilitated by the development of a novel PCRtargeting mutagenesis method of which details can be found in Chapter VII. The discovery of a common isomerase for tryptophan and histidine biosynthesis is reported in Chapter II. This discovery arose from efforts aimed at reconstructing the tryptophan biosynthetic pathway of S. coelicolor, since the genome sequence project of this organism failed to identifiy a trpF gene coding for the enzyme phosphoribosyl anthranilate isomerase. The solution of this functional genomics discrepancy led to the discovery of a putative (~a)8-barrel enzyme, termed PriA, whose preliminary functional and structural characterisation is reported in Chapter III. The evolutionary implications of the discovery of PriA are discussed within Chapters III and N. A comparative genomics analysis of actinomycetes centred on the priA gene is presented in the latter Chapter, supporting the notion that this novel protein is spread across the high (0 + C) content Gram-positive organisms. Indeed, it was predicted that a priA orthologue accounts for the lack of a trpF gene from the genome of M tuberculosis, a hypothesis that proved to be correct. Finally, evidence to support the notion that the histidine and tryptophan biosynthetic pathways co-evolved is presented. In contrast to the isomerisation catalysed by PriA, in which an enzyme is shared by two amino acid biosynthetic pathways, several paralogous enzymes with the potential to account for the first step of tryptophan biosynthesis from chorismate were found on the genome of S. coelicolor. These chorismate-utilising enzymes are investigated in Chapter V. Mutational analysis of some of this paralogues is reported and it is anticipated that the analysis and results reported therein will serve to direct future experiments aimed at identifying the trpE paralogue encoding the enzyme anthranilate synthase. Chapter VI reports on the identification of the proC gene involved in the last step of proline biosynthesis in S. coelicolor. The pyrroline-5-carboxylate reductase activity of the enzyme encoded by the putative proC gene was extensively characterised, with particular emphasis on the interaction between primary and secondary metabolism. Furthermore, mutational analysis of proC suggested that paralogues of this gene are present on the genome of this organism, since its deletion did not lead to an auxotrophic phenotype. Investigation of this observation showed that two paralogous enzymes encoded by i1vC-like genes, involved in biosynthesis of the branched-chain amino acids, are capable of compensating for the lack of proC. This is the first example of a physiological link between the biosynthesis of proline and the branched-chain amino acids. To sum up, the results reported in this thesis represent an advancement towards understanding the physiology of S. coelicolor as a model actinomycete, within a functional and structural genomics framework. They also offer evidence on the evolutionary principles that lead to the appearance of novel proteins and metabolic pathways in bacteria.