Les chiffres des statistiques du cancer colorectal en France et dans le mondemontrent la nécessité de développement de plateformes technologiques plus rapides,sensibles et spécifiques pour assurer le diagnostic du cancer. Un diagnostic rapide va ainsiaider à améliorer l’état de santé et réduire le temps d’attente des résultats qui peut être ungrand facteur de stress pour les patients. L’analyse des biomarqueurs dans le sang, lesurines et autres fluides corporels est l’une des méthodes appliquées pour la détectionprécoce de la maladie. Dans le cadre de ce projet des nucléosides urinaires ont été identifiéscomme biomarqueurs pour le cancer colorectal. Financée par l’Agence Nationale de laRecherche (ANR), à travers le projet CancerSensor (programme TECSAN), cette thèse s’estdéroulée au sein de l’équipe MDA (Microsystèmes de Détection Acoustique) du laboratoireIMS. Dans le cadre de ce projet, nous avons proposé une solution technologique dedétection et de suivi de biomarqueurs du cancer colorectal. Notre choix de la stratégie dedétection s’est porté sur les polymères à empreintes moléculaires comme élément dereconnaissance des biomarqueurs. Celui-ci sera associé à un transducteur acoustique àondes de Love mis au point lors de travaux précédents au sein de l’équipe MDA. Lebiocapteur ainsi développé va cibler les nucléosides mis en évidence pour le cancercolorectal. / Colorectal cancer statistics in France and all over the world demonstrate theneed for fast, sensitive and specific technological platforms development for cancerdiagnosis. A rapid diagnosis will improve the patients’ health status and reduce the resultswaiting time which could be a great stress factor. Biomarkers analysis in blood, urine andother body fluids is recognized as one of the applied methods for early cancer detection. Inframe of this project, urinary nucleosides have been identified as colorectal cancerbiomarkers. Funded by the National Research Agency (ANR), through the cancer sensorproject (TECSAN program), this thesis was carried out in IMS laboratory. Hence, a colorectalcancer biomarkers detection and monitoring technological solution has been proposed. Inour detection strategy, Molecularly Imprinted Polymers (MIP) has been identified asbiomarker recognition element. The MIP layer has been associated to Love Wave acoustictransducer. This biosensor will sense the identified colorectal cancer nucleosides.

Killer strains of the yeast Kluyveromyces lactis secrete a heterotrimeric protein toxin (zymocin) to inhibit the growth of sensitive yeasts. The cytotoxicity of zymocin resides in the γ subunit (γ-toxin), however the mechanism of cytotoxicity caused by γ-toxin was previously unknown. This thesis aimed to unravel the mode of γ-toxin action and characterize the interaction between γ-toxin and its substrates. Previous studies suggested a link between the action of γ-toxin and a distinct set of transfer RNAs. In paper I, we show that γ-toxin is a tRNA anticodon endonuclease which cleaves tRNA carrying modified nucleoside 5-methoxycarbonylmethyl-2-thiouridine (mcm5s2U) at position 34 (wobble position). The cleavage occurs 3’ to the wobble uridine and yields 2’, 3’-cyclic phosphate and 5´-hydroxyl termini. In paper II, we identified the determinants in tRNA important for efficient γ-toxin cleavage. The modifications present on the wobble uridines have different effects on tRNA cleavage by γ-toxin. The Saccharomyces cerevisiae wobble modification mcm5 group has a strong positive effect, whereas the Escherichia coli wobble modification 5-methylaminomethyl group has a strong inhibitory effect on tRNA cleavage. The s2 group present in both S. cerevisiae and E. coli tRNAs has a weaker positive effect on the cleavage. The anticodon stem loop (ASL) of tRNA represents the minimal structural requirement for γ-toxin action. Nucleotides U34U35C36A37C38 in the ASL are required for optimal cleavage by γ-toxin, whereas a purine at position 32 or a G at position 33 dramatically reduces the reactivity of ASL. Screening for S. cerevisiae mutants resistant to zymocin led to the identification of novel proteins important for mcm5s2U formation (paper III). Sit4p (a protein phosphatase), Sap185p and Sap190p (two of the Sit4 associated proteins), and Kti14p (a protein kinase) are required for the formation of mcm5 side chain. Ncs2p, Ncs6p, Urm1p, and Uba4p, the latter two function in a protein modification (urmylation) pathway, are required for the formation of s2 group. The gene product of YOR251C is also involved in the formation of s2 group. The involvement of multiple proteins suggests that the biogenesis of mcm5s2U is very complex.

The studies reported in this thesis address firstly, the biology of a modified nucleoside, Queuosine (Q) and secondly, the properties of mitochondrial translation initiation factor 2. A summary of the relevant literature on both these topics is presented in Chapter 1. Section I of this ‘General Introduction’ summarizes the literature on biosynthesis and physiological importance of Queuosine. Section II is a brief review of the current understanding of translation initiation in Eubacteria. Information about the mitochondrial translation initiation apparatus also features as a subsection. The next chapter (Chapter 2), describes the ‘Materials and Methods’ used throughout the experimental work presented in the thesis. It is followed by three chapters containing experimental work as described below:- i) Biosynthesis of Queuosine (Q) in Escherichia coli Q is a hypermodification of guanosine found at the wobble position of tRNAs with GUN anticodons. Q is thought to be produced via a complex multistep pathway, the details of which are not known. It was found in our laboratory that a naturally occurring strain of E. coli B105 lacked Q modification in the tRNAs. As the known enzymes of Q biosynthesis were functional in this strain, it presented us with the opportunity to uncover novel component(s) of Q biosynthetic pathway. In the present work, a genetic screen was developed to map the defect in E. coli B105 to a previously uncharacterised gene, ybaX, predicted to code for a 231 amino acid long protein with a pI of 5.6. Further genetic analyses showed that YbaX functions at a step leading to production of preQ0, the first known intermediate in the generally accepted pathway that utilizes GTP as the starting molecule. The gene ybaX has been renamed as queC. Using a combination of bioinformatics based prediction and gene knockouts, we have also been able to place two more genes, queD and queE at the initial step in Q biosynthesis, suggesting that the initial reaction of Q biosynthesis might be more complex and mechanistically different than what has been proposed earlier. ii) Caenorhabditis elegans as a Model System to Study Queuosine Metabolism in Metazoa Animals are thought to obtain Q (or its analogs) as a micronutrient from dietary sources such as gut microflora, and the corresponding base is then inserted in the substrate tRNAs by tRNA guanine transglycosylase (TGT). In animal cells, changes in the abundance of Q have been shown to correlate with diverse phenomena including stress tolerance, cell proliferation and tumor growth but the precise function of Q in animal tRNAs remains unknown. A major obstacle in the study of Q metabolism in higher organisms has been the requirement of a chemically defined medium to cause Q depletion in animals. Having discovered that E. coli B105 has a block in the initial step of Q biosynthesis, we reasoned that this strain could be used as a Q- diet for organisms like C. elegans, which naturally feed on bacteria. An analysis of C. elegans tRNA revealed that as in the other higher animals, tRNAs in the worm C. elegans, are modified by Q and its sugar derivatives. When the worms were fed on Q deficient E. coli B105, Q modification was absent from the worm tRNAs suggesting that C. elegans lacks a de novo pathway of Q biosynthesis. The inherent advantages of C. elegans as a model organism, the speed and simplicity of conferring a Q deficient phenotype on it, make it an ideal system to investigate the function of Q modification in tRNA. By microinjecting tgt-1-gfp constructs into C. elegans, we could also demonstrate that a major form of TGT is localised to the nucleus, suggesting that insertion of Q into the tRNAs could be occurring in the nucleus. iii) Dual Function of Bovine Mitochondrial Initiation Factor 2 as Initiation Factors 1 and 2 in Escherichia coli Translation initiation factors 1 and 2 (IF1 and IF2) are known as ‘universal translation initiation factors’ due to the presence of their homologs in all living organisms. Homologs of these factors are also present in the chloroplast, however, a unique situation exists in the mitochondria where IF2 homolog (IF2mt) is known to occur but an IF1 like factor is not found. We have engineered a system of E. coli knockouts to allow the study of IF2mt in a prokaryotic milieu. We found that the bovine IF2mt complements an E. coli strain wherein the gene for IF2 is knocked out, providing the first proof of a mitochondrial translation initiation factor working in a eubacterial system. This conservation of function is especially interesting in light of the recent reports revealing significant differences between the mitochondrial and eubacterial ribosomes. Further, we found that the IF2mt can also support a double knockout of IF1 and IF2 genes in E. coli, suggesting that IF2mt possesses both IF1 and IF2 like activities in E. coli. This finding offers an explanation for the lack of an IF1 like factor in mitochondria. Molecular modeling of bovine IF2mt indicated that a conserved insertion found in all mitochondrial IF2s, may form a protruding α-helix that could stabilize IF2mt on ribosomes. This insertion could in principle function as IF1 and we have explored the role of this conserved insertion both in vivo and in vitro, by generating mutants of IF2mt and EcoIF2, to lose or gain the conserved insertion respectively.

The thesis is divided in to five chapters. Chapter I: Synthesis of New thiolevomannosan derivatives In this chapter, a general and efficient method for the synthesis of conformationally locked thiosugars has been discussed. An unprecedented synthesis of a novel thioorthoester and its synthetic utility in glycosylation has been demonstrated. Chapter II: Studies on -Mannosidase Inhibitors The synthesis and evaluation of novel, conformationally locked glycomimic, thiolevomannosan and its analogs sulfoxide and sulfone starting from readily available D-mannose is discussed in this chapter. This is the first report of thiosugar derivatives with enhanced potency compared to kifunensine. Docking and biochemical studies have been discussed. Chapter III: Studies on Novel Cyclic Tetraselenides of Mannose In this chapter, the syntheses and structural properties of novel cyclic tetraselenides starting from mannose have been discussed. These tetraselenides are the first of their kind where all four selenium atoms are arranged in a cyclic manner as the backbone of mannose. X-ray structures have been reported for the tetraselenides. Chapter IV: Novel Chalcogenides of Uridine and Thymidine: Synthesis and Applications An efficient and simple method to synthesise disulfides and diselenides of thymidine and uridine derivatives has been demonstrated in this chapter. The utility of these disulfides in various ring opening reactions as well in Michael addition reactions has been demonstrated. Chapter V: Studies on New, Potent Urease Inhibitors In this chapter, a facile one-pot synthesis of thio and selenourea derivatives under mild conditions by the reaction of amines and commercially available Viehe’s iminium salt in the presence of benzyltriethylammonium tetrathiomolybdate as sulfur transfer reagent and tetraethylammonium tetraselenotungstate as selenium transfer reagent has been discussed. A few of the urea derivatives have shown potent inhibitor activity in the nanomolar range for jackbean urease.

Trypanosoma brucei causes African sleeping sickness in humans and Nagana in cattle. There are no vaccines available against the disease and the current treatment is also not satisfactory because of inefficacy and numerous side effects of the used drugs. T. brucei lacks de novo synthesis of purine nucleosides; hence it depends on the host to make its purine nucleotides. T. brucei has a high affinity adenosine kinase (TbAK), which phosphorylates adenosine, deoxyadenosine (dAdo), inosine and their analogs. RNAi experiments confirmed that TbAK is responsible for the salvage of dAdo and the toxicity of its substrate analogs. Cell growth assays with the dAdo analogs, Ara-A and F-Ara-A, suggested that TbAK could be exploited for drug development against the disease. It has previously been shown that when T. brucei cells were cultivated in the presence of 1 mM deoxyadenosine (dAdo), they showed accumulation of dATP and depletion of ATP nucleotides. The altered nucleotide levels were toxic to the trypanosomes. However the salvage of dAdo in trypanosomes was dramatically reduced below 0.5 mM dAdo. Radiolabeled dAdo experiments showed that it (especially at low concentrations) is cleaved to adenine and converted to ATP. The recombinant methylthioadenosine phosphorylase (TbMTAP) cleaved methylthioadenosine, dAdo and adenosine into adenine and sugar-1-P in a phosphate-dependent manner. The trypanosomes became more sensitive to dAdo when TbMTAP was down-regulated in RNAi experiments. The RNAi experiments confirmed that trypanosomes avoid dATP accumulation by cleaving dAdo. The TbMTAP cleavage-resistant nucleoside analogs, FANA-A and Ara-A, successfully cured T. brucei-infected mice. The DNA building block dTTP can be synthesized either via thymidylate synthase in the de novo pathway or via thymidine kinase (TK) by salvage synthesis. We found that T. brucei and three other parasites contain a tandem TK where the gene sequence was repeated twice or four times in a single open reading frame. The recombinant T. brucei TK, which belongs to the TK1 family, showed broad substrate specificity. The enzyme phosphorylated the pyrimidine nucleosides thymidine and deoxyuridine, as well as the purine nucleosides deoxyinosine and deoxyguanosine. When the repeated sequences of the tandem TbTK were expressed individually as domains, only domain 2 was active. However, the protein could not dimerize and had a 5-fold reduced affinity to its pyrimidine substrates but a similar turnover number as the full-length enzyme. The expressed domain 1 was inactive and sequence analysis revealed that some active residues, which are needed for substrate binding and catalysis, are absent. Generally, the TK1 family enzymes form dimers or tetramers and the quaternary structure is linked to the affinity for the substrates. The covalently linked inactive domain-1 helps domain-2 to form a pseudodimer for the efficient binding of substrates. In addition, we discovered a repetition of an 89-bp sequence in both domain 1 and domain 2, which suggests a genetic exchange between the two domains. T. brucei is very dependent on de novo synthesis via ribonucleotide reductase (RNR) for the production of dNTPs. Even though T. brucei RNR belongs to the class Ia RNR family and contains an ATP-binding cone, it lacks inhibition by dATP. The mechanism behind the RNR activation by ATP and inactivation by dATP was a puzzle for a long time in the ~50 years of RNR research. We carried out oligomerization studies on mouse and E. coli RNRs, which belongs to the same family as T. brucei, to get an understanding of the molecular mechanism behind overall activity regulation. We found that the oligomerization status of RNRs and overall activity mechanism are interlinked with each other. / Targeting the nucleotide metabolism of the mammalian pathogen Trypanosoma brucei.

The origin of the first RNA polymers is central to most current theories regarding the origin of life. However, difficulties associated with the prebiotic formation of RNA have lead many researchers to conclude that simpler polymers, or proto-RNAs, preceded RNA. These earlier polymers would have been replaced by RNA over the course of evolution. A remaining difficulty for this theory is that the de novo synthesis of a feasible proto-RNA has not yet been demonstrated by plausible prebiotic reactions. This thesis focuses on two problems associated with prebiotic proto-RNA synthesis: The formation of nucleosides and the necessity of reversible backbone linkages for error correction in nucleic acid polymers. "The Nucleoside Problem", or the lack of success in forming pyrimidine nucleosides by plausible prebiotic reactions, represents a significant stumbling block to the RNA world hypothesis. Nearly four decades ago Orgel and coworkers demonstrated that the purine nucleosides adenosine and inosine are synthesized by heating and drying their respective bases and ribose in the presence of magnesium, but these reaction conditions do not yield the pyrimidine nucleosides uridine or cytidine from their respective bases. In this thesis a potential solution to The Nucleoside Problem is hypothesized based upon a proposed chemical mechanism for nucleoside formation. This hypothesis is supported by the successful synthesis of 2-pyrimidinone nucleosides by a plausible prebiotic reaction in good yield, demonstrating that pyrimidine nucleosides could have been available in the prebiotic chemical inventory, but that uridine and cytidine were likely not abundant. Reversible backbone linkages are necessary to provide a mechanism for error correction in non-enzymatic template-directed syntheses of proto-RNAs. In this thesis, acetals are explored as low-energy, reversible linkage groups for nucleosides in polymers. The synthesis of glyoxylate-acetal nucleic acids (gaNAs) through simple heating-drying reactions from neutral aqueous solutions is demonstrated, and these linkages are shown to be hydrolytically stable under a considerable range of solution conditions. Computational models demonstrate that the glyoxylate linkage is an excellent electronic and isosteric replacement for phosphate. Molecular dynamics simulations also indicate that a gaNA duplex would have structural properties that closely match a phosphate-linked RNA helix, suggesting the possibility for cross-pairing between gaNAs and RNAs, allowing for sequence transfer and genetic continuity through the evolution from proto-RNAs to RNA. The principles illustrated in this thesis by 2-pyrimidinone nucleoside and gaNA synthesis can be extended to other prebiotic condensation reactions. Factors affecting condensation yield, such as thermodynamics, kinetics, reactant solubility, and salt effects, are summarized herein.

The Snail protein functions as a transcriptional regulator to establish early mesodermal cell fate in Drosophila. Later, in germ band-extended embryos, Snail is considered a pan-neural protein based on its extensive expression in neuroblasts. The evidence presented in thesis links snail expression and function in CNS. Cloning and functional characterization of a novel snail homologue, in Drosophila, are also described here. Cloning of this gene, worniu (Chinese for snail), revealed that the neural function of snail is masked by this and another closely related gene escargot. Both Escargot and Worniu contain zinc finger domains that are highly homologous to that of Snail. These three members of Snail protein family are redundantly required for CNS development. Although not affecting formation of neuroblasts, the loss of expression of these three members correlates with disruption of Nb asymmetry and division. Downstream targets of Snail protein family, in these processes, are inscuteable and string. In mutant embryos, which have the three genes deleted, the RNA expression of inscuteable and string is significantly lowered. Consistent with the gene expression defects, the mutant embryos have loss of asymmetric localization of prospero RNA in neuroblasts and nuclear localization of Prospero protein in ganglion mother cells. Transgenic expression of inscuteable and string together, in the snail family deletion mutant, efficiently restores the Prospero expression in GMC, demonstrating that the two genes are key targets of Snail in Nbs. Like in the mesoderm, in CNS Snail function depends on interaction with dCtBP co-repressor. These results suggest that Sna [Snail] family of proteins control both asymmetry and cell division of neuroblasts by activating, perhaps indirectly, the expression of inscuteable and string.

Calcium ions (Ca2+) are involved in almost all neuronal functions, providing the link between electrical signals and cellular activity. This work examines the mechanisms by which a neuron can regulate the movement and sequestration of Ca2+ through specific channels such that this ubiquitous ion can encode specific functions. My initial focus was using intracellular calcium ([Ca2+]i) imaging techniques to study the influence of the inhibition of specific voltage gated calcium channels (VGCC) by ethanol on a depolarization induced rise in [Ca2+]i in neurohypophysial nerve terminals. This research took an unexpected turn when I observed an elevation of [Ca2+]i during perfusion with ethanol containing solutions. Control experiments showed this to be an artifactual result not directly attributable to ethanol. It was necessary to track down the source of this artifact in order to proceed with future ethanol experiments. The source of the artifact turned out to be a contaminant leaching from I.V. drip chambers. Due to potential health implications stemming from the use of these drip chambers in a clinical setting as well as potential artifactual results in the ethanol field where these chambers are commonly used, I choose to investigate this phenomenon more rigorously. The agent responsible for this effect was shown to be di(2-ethylhexyl)phthalate (DEHP), a widely used plasticizer that has been shown to be carcinogenic in rats and mice. The extraction of this contaminant from the I.V. drip chamber, as measured by spectrophotometry, was time-dependent, and was markedly accelerated by the presence of ethanol in the solution. DEHP added to saline solution caused a rise in [Ca2+]i similar to that elicited by the contaminant containing solution. The rise in calcium required transmembrane flux through membrane channels. Blood levels of DEHP in clinical settings have been shown to exceed the levels which we found to alter [Ca2+]i. This suggests that acute alterations in intracellular calcium should be considered in addition to long-term effects when determining the safety of phthalate-containing plastics. As part of a collaboration between Steven Treistman and Robert Messing's laboratory at UCSF, I participated in a study of how ethanol regulates N-type calcium channels which are known to be inhibited acutely, and upregulated in the chronic presence of ethanol. Specific mRNA splice variants encoding N-type channels were investigated using ribonuclease protection assays and real-time PCR. Three pairs of N-type specific α-subunit Cav2.2 splice variants were examined, with exposure to ethanol observed to increase expression of one alternative splice form in a linker that lacks six bases encoding the amino acids glutamate and threonine (ΔET). Whole cell electrophysiological recordings that I carried out demonstrated a faster rate of channel activation and a shift in the voltage dependence of activation to more negative potentials after chronic alcohol exposure, consistent with increased expression of ΔET variants. These results demonstrate that chronic ethanol exposure not only increases the abundance of N-type calcium channels, but also increases the expression of a Cav2.2 splice variant with kinetics predicted to support a larger and faster rising intracellular calcium signal. This is the first demonstration that ethanol can up-regulate ion channel function through expression of a specific mRNA splice variant, defining a new mechanism underlying the development of drug addiction. Depolarizing a neuron opens voltage gated Ca2+ channels (VGCC), leading to an influx of Ca2+ ions into the cytoplasm, where Ca2+ sensitive signaling cascades are stimulated. How does the ubiquitous calcium ion selectively modulate a large array of neuronal functions? Concurrent electrophysiology and ratiometric calcium imaging were used to measure transmembrane Ca2+ current and the resulting rise and decay of [Ca2+]i, showing that equal amounts of Ca2+ entering through N-type and L-type voltage gated Ca2+ channels result in significantly different [Ca2+]i temporal profiles. When the contribution of N-type channels was reduced, a faster [Ca2+]i decay was observed. Conversely, when the contribution of L-type channels was reduced, [Ca2+]i decay was slower. Potentiating L-type current or inactivating N-type channels both resulted in a more rapid decay of [Ca2+]i. Channel-specific differences in [Ca2+]i decay rates were abolished by depleting intracellular Ca2+ stores suggesting the involvement of Ca2+-induced Ca2+ release (CICR). I was able to conclude that Ca2+ entering through N-type, but not L-type channels, is amplified by ryanodine receptor mediated CICR. Channel-specific activation of CICR generates a unique intracellular Ca2+ signal depending on the route of entry, potentially encoding the selective activation of a subset of Ca2+ -sensitive processes within the neuron.

Knowledge of how a polypeptide folds from a space-filling random coil into a biologically-functional, three-dimensional structure has been the essence of the protein folding problem. Though mechanistic details of DNA transcription and RNA translation are well understood, a specific code by which the primary structure dictates the acquisition of secondary, tertiary, and quarternary structure remains unknown. However, the demonstrated reversibility of in vitroprotein folding allows for a thermodynamic analysis of the folding reaction. By probing both the equilibrium and kinetics of protein folding, a protein folding mechanism can be postulated. Over the past 40 years, folding mechanisms have been determined for many proteins; however, a generalized folding code is far from clear. Furthermore, most protein folding studies have focused on monomeric proteins even though a majority of biological processes function via the association of multiple subunits. Consequently, a complete understanding of the acquisition of quarternary protein structure is essential for applying the basic principles of protein folding to biology. The studies presented in this dissertation examined the folding and assembly of two very different multimeric proteins. Underlying both of these investigations is the need for a combined analysis of a repertoire of approaches to dissect the folding mechanism for multimeric proteins. Chapter II elucidates the detailed folding energy landscape of HIV-1 protease, a dimeric protein containing β-barrel subunits. The folding of this viral enzyme exhibited a sequential three-step pathway, involving the rate-limiting formation of a monomeric intermediate. The energetics determined from this analysis and their applications to HIV-1 function are discussed. In contrast, Chapter III illustrates the association of a coiled coil component of L. terrestriserythrocruorin. This extracellular hemoglobin consists of a complex scaffold of linker chains with a central ring of interdigitating coiled coils. Allostery is maintained by twelve dodecameric hemoglobin subunits that dock upon this scaffold. Modest association was observed for this coiled coil, and the implications of this fragment to linker assembly are addressed. These studies depict the complexity of multimeric folding reactions. Chapter II demonstrates that a detailed energy landscape of a dimeric protein can be determined by combining traditional equilibrium and kinetic approaches with information from a global analysis of kinetics and a monomer construct. Chapter III indicates that fragmentation of large complexes can show the contributions of separate domains to hierarchical organization. As a whole, this dissertation highlights the importance of pursuing mulitmeric protein folding studies and the implications of these folding mechanisms to biological function.

RNA interference (RNAi) is an evolutionarily conserved, sequence-specific gene silencing pathway found in eukaryotes, in which 21-nucleotide, small interfering RNAs (siRNAs) guide destruction of a corresponding target mRNA. RNAi is a natural mechanism for both genome surveillance and gene regulation. Moreover, siRNAs can be transfected into cultured mammalian cells, causing the sequence-specific ‘knock down’ of an mRNA. My work in the Zamore lab has centered around the Drosophilain vitro system and cultured mammalian cells to study the RNA interference (RNAi) pathway. small interfering RNAs (siRNAs) are incorporated into the RNA-induced silencing complex (RISC), which culminates in the cleavage of a complementary target mRNA. Previous work proved that certain structural features of siRNAs are essential for RNAi in flies, including the requirement for 5´ phosphates and 3´ hydroxyl groups. In cultured mammalian cells, the requirement for a 5´ phosphate also holds true, but we found no evidence to support the necessity for 3´ hydroxyls in either system. In addition, siRNAs can act as single strands entering the pathway downstream of double-stranded siRNAs, both of which are competent in directing the cleavage of its cognate mRNA at a single site. While these key features are a requirement for functional siRNAs, alone they do not determine the efficiency to which an siRNA can enter the RISC. In fact, both strands of an siRNA can enter RISC to a different degree as determined by the stabilities of the 5´ ends of the siRNA strand, a phenomenon termed ‘functional asymmetry’. This characteristic is also reflected in another class of small RNAs involved in gene silencing known as microRNAs (miRNAs), which are processed from long hairpin RNA structures into mature, single-stranded non-coding RNAs. The asymmetric loading of siRNAs suggests that miRNAs are initially generated from siRNA-like duplexes cleaved from the stem of the hairpins. The strand whose 5´ end is less tightly paired will be processed into the mature miRNA, while the other strand is destroyed. By applying the rules of siRNA asymmetry it is possible to predict which side of the stem will be processed into the mature miRNA, a finding verified experimentally by our lab and others. This discovery also has additional implications in designing highly effective siRNAs and in reducing siRNA off-target effects. We used these results to design siRNAs that target the single nucleotide polymorphism in superoxide dismutase that causes the familial form of amyotrophic lateral sclerosis (ALS), but leave the wild-type mRNA intact and functional. Our experiments have helped define the ‘rules’ for creating SNP-specific siRNAs. In particular, we found that only siRNAs with a purine:purine mismatch to the allele not intended for destruction show good discrimination. The placement of the mismatch in a tiled set of siRNAs shows that mismatches located in the 5´ region of the siRNA, a region shown to be responsible for siRNA binding, can not discriminate between alleles. In contrast, mismatches in the 3´ region of the siRNA, the region contributing to catalysis, discriminate between wild-type and mutant alleles. This work is an important step in creating allele-specific siRNAs as therapeutics for dominant negative genetic diseases. But how does RISC cleave its target? By isolating both the 5´ and 3´ cleavage products produced by RISC in the Drosophila in vitro system, we discovered that RISC acts as a Mg2+-dependent endonuclease that cleaves a single phosphodiester bond in the mRNA target, leaving 5´ phosphate and 3´ hydroxyl groups. These findings were a critical step in the demonstration that Argonaute, a protein known to be a component of RISC, is the RNAi endonuclease.