Post-transcriptional control of gene expression in Eukaryotes plays a pivotal role in determining intricated networks defining physiological and pathological conditions among each organism. RNA Binding Proteins (RBPs), by exploiting RNA-protein and protein-protein interactions, have been recognized as the main actors in modulating these processes. As a consequence, RBPs aberrant expression, modulation or mis-localization, leads to the insurgence of complex phenotypes and diseases. Therefore, targeting and modulating the activity of RBPs found associated to different pathologies represents a new promising therapeutic strategy. During my PhD I aimed at identify, characterize and refine inhibitors targeting the RNA binding protein HuR. HuR belongs to the ELAVL protein family, it is ubiquitously expressed in the cells and among tissues and highly conserved throughout mammalian evolution. By binding AU/U rich elements (ARE) in the 3’UTRs of mRNAs, HuR mainly stabilizes its target transcripts, enhancing their translation. ARE sequences are found in 7% of the human mRNAs, coding for protein involved in key cellular processes as: immune response and inflammation, cell division and proliferation, angiogenesis, senescence and apoptosis. Hence, dysregulation in HuR expression and in its subcellular localization have been associated with the insurgence of several pathologies, mostly cancers and inflammation diseases. Notably, malignant transformations and poor prognosis in patients have been found characterized by highly nuclear or cytosolic HuR expression in a significant number of human cancers. Indeed, the majority of HuR regulated transcripts encode for protein responsible for the appearance of several cancerogenic traits. In particular, critical crosstalk established between cancer cells and inflammation processes play a pivotal role in worsening and compromising cancers development and onset. Moreover, considering that 90% of mRNAs coding for cytokines and chemokines contains repeated AREs sites in the 3’UTR, HuR plays a strong regulatory role in immune system (innate and adaptive) development and homeostasis as well as in pathogenic mechanisms. The searching for HuR inhibitors represents a challenging area, in the drug discovery field, due to its pleiotropic functions and its intrinsic structural complexity, which presents unfolded regions and sequences prone to aggregation. HuR disruptors have been reported in the literature, but without systematic studies, thus the identification of a new class of small molecules is still at the beginning. Among the molecules discovered so far, in 2015 our group identified through a High-throughput Screening a natural compound, DHTS, as a bona fide HuR inhibitor. Following that finding, we, me included, ascribed to the molecule a well-defined mechanism of action, identifying the specific binding sites on which HuR:DHTS interaction is based, defining that upon the mRNA binding DHTS interplays with HuR maintaining the protein in a closed conformation, thus inhibiting its function. Furthermore, we demonstrated DHTS anti-cancer activity in vitro, in cellular context and in vivo, in an HuR-dependent manner. In this way, DHTS represented the molecular scaffold, for the generation of a new class of highly potent HuR inhibitors, called Tanshinone Mimics (TMs). A functional oriented approach was applied for the synthesis of new molecules harboring only DHTS chemical elements responsible for HuR targeting, leading to a completely new molecular scaffold, not previously described in the literature, with respect to the ancestor molecule. I have characterized and identified more potent molecules, describing their anticancer properties, through the evaluation of their capabilities of downregulating the total expression level of well-known HuR targets, coding for proteins involved in tumor insurgence and progression, as VEGF, ERBB2 and CTNNB1, and reducing cancer cell migration, cell cycle progression in a minor extent. On the other end, I have explored TMs anti-inflammatory properties, counteracting the inflammatory response mediated by macrophages, directly impairing the binding between HuR and its pro-inflammatory targets, diminishing their expression and related protein secretion. Moreover, I have put evidences on TMs activity in vivo in acute inflammation mouse models. Lastly, I have evaluated TMs activity in affecting T-cells proliferation, on which HuR it is known to play a regulatory role. In conclusion, we identified TMs with Structure-Activity Relationships (SARs) towards HuR inhibition and its biological implications, aimed at ameliorating their specificity and bioavailability suitable for in vivo therapeutic strategies.

L’enzyme méthionine synthase (MTR) catalyse la reméthylation de l’homocystéine en méthionine, le précurseur du donneur universel de groupe méthyle S-Adenosylmethionine (SAM), impliqué dans des mécanismes de régulations épigénétiques. Des polymorphismes de MTR sont associés à des défauts métaboliques et des défauts de développement embryonnaire. Afin d’étudier les conséquences d’une déficience en MTR, nous avons généré des modèles murins d’invalidation conditionnelle du gène MTR de manière constitutive ou inductible dans le foie et dans le cerveau. L’invalidation constitutive ou inductible ciblée dans le foie pendant l’embryogenèse n’est pas viable, suggérant un rôle limitant de la méthionine synthase sur le développement précoce et l’organogenèse en lien probable avec les conséquences sur la prolifération cellulaire. Dans les périodes post-natales, nous avons utilisé le modèle inductible complété par une hépatectomie pour étudier les altérations de la régénération hépatique liée aux effets sur le stress cellulaire ainsi que l’expression et l’activation des cyclines. Le KO dans le cerveau induit principalement une perte des fonctions de mémorisation de l’apprentissage hippocampo-dépendant. Au total, nos résultats illustrent les effets différents de l’invalidation de MTR en fonction de l’organe considéré. Le foie est un organe très plastique avec une capacité de régénération très importante. Les effets sur les étapes de l’organogénèse et sur l’inhibition de la régénération confirment l’hypothèse du rôle majeur et limitant de la méthionine synthase dans la régulation du cycle cellulaire. Le modèle d’invalidation au niveau du cerveau confirme le rôle très important de la voie de reméthylation de l’homocystéine catalysée par la méthionine synthase, rôle qui a déjà été illustré par d’autres travaux sur les rats carencés en donneur de méthyle et sur la souris transgénique KO cd320 / The enzyme methionine synthase (MTR) catalyzes the remethylation of homocysteine to methionine, the precursor of the methyl donor S-universal Adenosylmethionine (SAM), involved in epigenetic regulation mechanisms. We generated mouse models with conditional invalidation of the mtr gene in a constitutive or inducible manner to delete the gene expression specifically in the liver and brain. Constitutive invalidation during embryonic life is not sustainable when targeted to the liver, suggesting a limiting role of methionine synthase in early organogenesis and probably on cell proliferation. We performed hepatectomy to study regeneration-related effects on the cellular stress and found dramatic effects on cell proliferation through altered expression and activation of cyclins. The constitutive model in brain highlighted the behavioral anomalies related to a loss of learning and memory. This suggested major effects in the hippocampus. Overall, our findings highlighted the specific effects of the invalidation of methionine synthase in both organs. The liver is a plastic member with a very high regenerative capacity. The effects on organogenesis and inhibition of regeneration confirm the hypothesis for a major role of methionine synthase in cell cycle regulation. The invalidation model in the brain confirms the important role of the remethylation pathway catalysed by methionine synthase, a role which has been shown by other studies in rats deprived in methyl donors and in cd320 KO transgenic mice

The ZNF9 gene on chromosome 3 encodes the cellular nucleic acid binding protein (CNBP), a ubiquitously expressed, 177 amino acid (≈19.5kDa) protein that is highly conserved among vertebrates. The function of the protein is largely unknown, however an expansion in the first intron of the protein results in myotonic dystrophy type 2 (DM2), a multisystemic disease featuring cardiac arrhythmia, muscle wasting, cataracts, and a range of neuropathologies. Remarkably, we recently discovered that CNBP is involved in regulating the activity of β-secretase, the enzyme that produces the first cleavage event in the generation of the amyloid-β peptide (Aβ). The progressive fibrillization and deposition of Aβ is widely believed to be the primary causal factor in the development of Alzheimer’s disease (AD), and AD-like pathology in individuals with Down syndrome (DS). DS provides a unique model for evaluating how these factors change in the aged brain as compared to young brain, and how such changes affect the proportion of DS patients with AD. In the AD brain, both BACE1 and BACE2 increased from an early stage of disease; in DS brains, BACE1 significantly decreased (p<0.04) with age, whereas BACE2 was unchanged, even though the gene for BACE2 is located within the DS obligate region of chromosome 21. BACE1 and BACE2 activity levels were highly correlated in this series (r2 = 0.95), indicating that there may be a higher degree of shared regulation than previously believed. This implicates regulators of BACE as potentially critical for the development of AD, and our data suggests that CNBP may be one such regulator. In AD, CNBP increases early in the disease process, a change that does not occur in the normal aging process or in DS. CNBP and BACE protein levels were correlated in these cases (p<0.001), while there was no relationship between CNBP and age, or CNBP and Aβ, in either the human or mouse brain, indicating that CNBP does not increase as a consequence of normal aging. Thirty day overexpression of CNBP following adeno-associated viral delivery in murine gastrocnemius muscle resulted in an increase in BACE1 protein (p<0.01) and a consequential increase in Aβ production (p<0.01). Other experiments indicated that CNBP overexpression did not affect the half-life of BACE1 mRNA or protein, but resulted in an increase in BACE1 translation. These data indicate that CNBP is an important regulator of β-secretase, and may play an important role in the onset and progression of AD.

No description available.

Mémoire numérisé par la Division de la gestion de documents et des archives de l'Université de Montréal.

L’épithélium intestinal est une monocouche de cellules qui tapisse la lumière intestinale, constitué d’un compartiment différencié, les villosités dans l’intestin grêle et les plateaux épithéliaux dans le colon, et d’un compartiment prolifératif, les cryptes de Lieberkühn. Ce tissue se renouvelle de façon rapide et continue tout au long de la vie de l’individu, grâce à la présence de cellules souches adultes dans le fond des cryptes. Ces cellules s’autorenouvellent et donnent naissance à des progéniteurs prolifératifs (capables d’engendrer les différents cytotypes épithéliaux) qui se différencient tout en migrant vers le compartiment différencié. Mon travail de these a porté sur l’étude d’une marqueur putatif de ces cellules souches épithéliales intestinales: Musashi1 (Msi1).Dans ce contexte, mon premier axe d’étude s’est focalisé sur l’isolement et la caractérisation des cellules souches épithéliales intestinales chez la souris. Pour cela, nous avons généré des souris transgéniques exprimant la protéine fluorescente GFP sous le contrôle du promoteur de Msi1. Les cellules souches intestinales de ces souris coexpriment donc Msi1 et la GFP. Ce modèle a été validé et nous à permis de isoler les cellules GFP/Msi1 positives dans l’intestin. A l'aide de différentes approches cellulaires et moléculaires, nous avons confirmé leur nature de cellules souches et nous avons apporté des nouvelles données sur la composition de la zone proliférative de l’épithélium intestinal murin.Le second axe de mes travaux de thèse a porté sur l’étude de la fonction de Msi1 dans l'homéostasie de l’épithélium intestinal chez la souris, par son sur-expression tous au long de l’épithélium. Nous avons montré que la sur-expression de cette protéine, qui est un régulateur des voies Wnt et Notch, perturbe l’architecture intestinale, a propriétés pro-prolifératives et un potentiel tumorigènique. / The intestinal epithelium is a monolayer of cells surrounding the intestinal lumen. It consists of a differentiated compartment, the villi in the small intestine and a flat surface in the colon, and a proliferative compartment, the crypts of Lieberkühn. This tissue self-renews rapidly and continuously throughout life, due to the presence of adult stem cells in the bottom of the crypts. These cells are capable of self-renewing and give rise to proliferating progenitors (capable of generating all the different epithelial cytotypes) that differentiate and migrate toward the differentiated compartment. My thesis focused on the study of the intestinal epithelial stem cells marker Musashi1 (Msi1).In this context, the first part of my thesis work focused on the isolation and characterization of the intestinal epithelial stem cells that express Msi1 in the mouse. For this, we generated transgenic mice expressing the fluorescent protein GFP under the control of the promoter of Msi1. The intestinal stem cells of these mice co-express Msi1 and GFP. This model has been validated and allowed us to isolate GFP+/Msi-expressing cells in the intestine. By using different cellular and molecular approches, we confirmed their nature of stem cells and provided new data on the composition of the proliferative zone in the murine intestinal epithelium.The second part of my thesis has focused on the study of the function of Msi1 in the intestinal epithelium homeostasis in the mouse, by its over- and ectopic expression all along the epithelium. We have shown that the over-expression of this protein, which is a regulator of the Wnt and Notch pathways, perturbs the intestinal architecture, has pro-proliferative properties and tumorigenic potential.

La dynamique chromatinienne joue un rôle central dans les contrôles développementaux, la différenciation cellulaire ou les réponses des organismes à l’environnement. Chez les animaux, les protéines du groupe Polycomb sont impliquées dans l’établissement d’états chromatiniens silencieux. Chez les plantes, des données récentes suggèrent que la protéine LIKE HETEROCHROMATIN PROTEIN 1 (LHP1) participerait à un complexe de type Polycomb. Nous nous sommes intéressés aux complexes LHP1 en étudiant un de ses partenaires, LHP1 INTERACTING FACTOR 2 (LIF2). L’objectif de ce travail de thèse a été de poursuivre la caractérisation de LIF2. LIF2 se caractérise par la présence de domaines de liaison à l’ARN, suggérant la participation d’une composante ARN dans les complexes LHP1. Nous avons recherché des ARN ligands de LIF2 et étudié les interactions LIF2/ARN par différentes approches dont la technologie Biacore. En analysant le transcriptome du mutant lif2, nous avons remarqué un enrichissement pour des gènes impliqués dans la réponse aux stress biotiques et abiotiques. Nous avons étudié la fonction de LIF2 dans la réponse aux pathogènes et avons pu mettre en évidence que LIF2 joue un rôle dans l’immunité innée des plantes et est essentiel pour réguler négativement les réactions de défenses en l’absence de pathogènes. / Chromatin dynamics play a central role in developmental control, cell differentiation or responses of the organisms to environment. In animals, Polycomb group proteins are involved in the establishment of silent chromatin states. In plants, recent data suggest that LIKE HETEROCHROMATIN PROTEIN 1 (LHP1) participates to a Polycomb-like complex. We focused on LHP1 complexes by studying one of its partners, LHP1 INTERACTING FACTOR 2 (LIF2). The aim of this thesis was to pursue the characterization of LIF2. LIF2 is composed of RNA-binding domains, suggesting the participation of an RNA component in LHP1 complexes. We have searched for LIF2 RNA-ligands and studied LIF2/RNA interactions with different approaches including Biacore technology. By analyzing the transcriptome profile of lif2, we have noticed an enrichment for genes involved in responses to abiotic and biotic stresses stimuli. We investigated the LIF2 functions in response to pathogens infection and we have been able to highlight that LIF2 plays a role in plant innate immunity and is essential to negatively regulate defense responses in the absence of pathogens.

Les lymphomes anaplasiques à grandes cellules (LAGC) sont des lymphomes non-hodgkiniens dits de type T ou nul, représentant la majorité des lymphomes T pédiatriques (20 à 30%). Dans plus de 80% des cas, une translocation chromosomique réciproque aboutissant à l'expression anormale de protéines chimères de type X-ALK qui arborent constitutivement et de manière anormale l'activité tyrosine-kinase ALK (Anaplastic Lymphoma Kinase) est le moteur de la tumorigenèse (LAGC dits "ALK+ "). L'une des particularités de ces lymphomes, mise en évidence par mon équipe, est le fait que B-Cell Lymphoma-2 (BCL-2) demeure indétectable dans les cas ALK+ contrairement aux cas ALK-. Ce point est d'autant plus surprenant que BCL-2, oncogène largement établi comme prototype de protéines anti-apoptotiques ainsi que régulateur clé de l'autophagie, est fortement surexprimé dans la majorité des lymphomes. A l'inverse, Human Antigen R (HuR) est surexprimée dans les LAGC (comme dans la plupart des cancers). Il a été démontré que cette protéine de liaison aux ARN participait au maintien du phénotype tumoral, et que sa localisation subcellulaire et ses fonctions dépendaient étroitement de son statut de phosphorylation, lequel est régulé par ALK dans les LAGC ALK+. Au niveau du cytoplasme, HuR permet de stabiliser et d'augmenter la traduction d'ARNm possédant, dans leur région 3'-UTR, des séquences riches en adénine et uridine (AU-rich elements, "ARE"). De manière plus générale, HuR a la capacité de dialoguer avec les microARNs (miARN), soit en empêchant leur action par compétition, soit à l'inverse en coopérant avec ces derniers et en promouvant ainsi leur fonction de régulateur négatif sur certains transcrits cibles communs. Le transcrit Bcl-2, dont l'expression est réprimée dans les LAGC ALK+, fait partie des cibles potentielles de HuR. Au cours de ma thèse, j'ai ainsi cherché à comprendre les mécanismes moléculaires mis en jeu dans la répression de l'expression de Bcl-2, en me focalisant sur le rôle de HuR et de miARN "partenaires" dans ce processus. Mes données semblent indiquer que ce mécanisme implique le recrutement par HuR du miR-34a sur l'ARNm Bcl-2, conduisant à la mise en silence de ce dernier. A l'inverse, quand l'activité tyrosine- kinase de ALK est inhibée, l'interaction entre HuR et le transcrit Bcl-2 diminue, ce qui limite le recrutement de miR-34a et conduit à une restauration de l'expression de cette oncogène majeur dans les cellules lymphomateuses. Dans le contexte des essais cliniques d'inhibiteurs ciblant l'activité tyrosine kinase de ALK tels que le Crizotinib, la question de cette ré-expression de BCL-2 éclaire d'une lumière nouvelle certains échecs thérapeutiques subis par cette molécule pourtant prometteuse. Je me suis donc également consacré, pendant ma thèse, à l'étude des conséquences de cette ré-expression de BCL-2 sur les LAGC ALK+ traités par le Crizotinib. Les résultats que j'ai obtenus in vitro et in vivo montrent que contrecarrer, par interférence à l'ARN, l'élévation du taux de BCL-2 consécutive au traitement par le Crizotinib, permet de potentialiser les effets de la drogue : cela se traduit en particulier par une potentialisation de la mort par apoptose induite par le traitement mais aussi, de manière fascinante, par une conversion de la réponse autophagique initialement cytoprotectrice et pro-tumorale en une autophagie incontrôlée et délétère, qui participe alors à l'effet thérapeutique accru de la drogue. De manière globale, ce travail permet d'envisager de nouvelles combinaisons et alternatives thérapeutiques pour les patients souffrants de LAGC ALK+, et illustre la complexité des régulations croisées entre processus apoptotiques et autophagiques. / Anaplastic large cell lymphoma (ALCL) are T/-null non-hodgkin lymphoma representing most of childhood T-cell lymphoma (up to 30%). More than 80% of cases bear reciprocal chromosomic translocation responsible for abnormal expression and constitutive activation of X-ALK type (Anaplastic Lymphoma Kinase) chimeric proteins (ALK+ ALCL). A striking characteristic of this lymphoma is that B-Cell Lymphoma-2 (BCL-2) remains undetectable in ALK+ cases compared to ALK- cases. This is all the more surprising as the BCL-2 oncogene, which is firmly established as a prototypic anti-apoptotic factor as well as a key autophagy regulator, has been shown to be overexpressed in a majority of lymphomas. On the other hand, the RNA-binding protein HuR (Human Antigen R) is overexpressed in ALCL (as in most cancers). It has been demonstrated that this protein was involved in the sustainability of the tumoral phenotype, and that its subcellular localization and functions were closely related to its phosphorylation status, which in turn heavily depends on ALK activity in ALK+ ALCL. In the cytoplasm, HuR has the ability to bind adenine and uridine-rich elements (ARE) located on the 3'-UTR of target mRNAs, and both protect them from degradation and increase their translation. From a general point of view, HuR is able to establish an interplay with microRNAs (miRNAs), either blocking them through competition, or actually cooperating with them and thus promote their function of negative regulators of gene expression on common target transcripts. The BCL-2 transcript, which expression seems to be silenced in ALK-expressing ALCL, has been described as a potential target of HuR. During my PhD work, I dedicated myself to understand the molecular mechanism at work in the silencing of BCL-2 expression with a focus on HuR and collaborating miRNA. The data I obtained point at a cooperation between HuR and miR-34a leading to the silencing of the BCL-2 transcript. However, when the ALK tyrosine kinase activity is inhibited, it appears the interaction between the BCL-2 mRNA diminishes, which limitates the miR-34a 's access to this transcript and ultimately results in a re-expression of the BCL-2 oncogene in these lymphoma cells. In the current context of clinical trials for ALK-targeting inhibitors, such as the Crizotinib, this BCL-2 re-expression observed upon ALK inhibition shed light on potential reasons behind some therapeutic failures that have recently been reported. Indeed, during my PhD work, I also studied the consequences of the BCL-2 re-expression observed in Crizotinib-treated cells. The data I obtained in vitro and in vivo show that, by blocking this re-expression using RNA interference, the Crizotinib anti-tumoral efficiency can be greatly potentiated. This potentiation took the form of an increase of apoptotic cell death induction and, interestingly, also affected the autophagic response triggered by the drug, making it switch from a cytoprotective- type, protumoral autophagic flux to an enhanced, deletary-type and tumor suppressive flux, adding to the therapeutic effect of the drug. This work in general provides insights for new therapeutic combinations that could potentially benefit to ALK+ ALCL patients, and illustrates the complex cross-regulations between apoptotic and autophagic pathway.

Project I 2-Deoxystreptamine (2-DOS), the most conserved central scaffold of aminoglycosides, is known to specifically recognize the 5'-GU-'3 sequence step through highly conserved hydrogen bonds and electrostatic interactions within and without the context of aminoglycosides (Figure 1a). We proposed that a novel monomeric unnatural amino acid building block using 2-DOS as a template would allow us to develop RNA binding molecules with higher affinity and selectivity than those currently available. Conjugating two or more of the monomeric building blocks by an amide bond would introduce extra hydrogen bonding donors and acceptors that are absent in natural aminoglycosides and increase specificity of binding to a target RNA through a network of hydrogen bonds. In addition, the amide conjugation between the monomeric building blocks places two GU-base recognizing amines at 5 Å… distance, which is equal to the distance of neighboring base stacks in dsRNAs We hypothesized that targeting dsRNAs containing multiple consecutive 5'-GU-'3 sequence steps would become possible by connecting two or more of the monomeric building blocks by amide bonds. According to the proposed hypothesis, we designed three dimeric 2-DOS compounds connected by an amide bond. These three targets include the dimeric 2-DOS substrate connected by an amide bond, the dimeric 2-DOS containing the sugar moiety from Neamine, and a dimeric 2-DOS connected by a urea linker. These compounds were then tested for sequence specific binding against 8 different RNA strands, and for antibacterial activity against E. coli, actinobacter baumannii and klebsiella. Project II A dual optimization approach was used for to enhance the catalytic activity and chemoselectivity for nitro reduction. In this approach the composition of the nanoparticles and electronics effects of the polymer were studied towards nitro reduction. Bimetallic Ruthenium-Cobalt nanoparticles showed exceptional catalytic activity and chemoselectivity compared to monometallic Ruthenium nanoparticles. The electronic effects of the polymer also had a significant effect on the catalytic activity of the bimetallic nanoparticles. The electron-deficient poly(4-trifluoromethylstyrene) supported bimetallic nanoparticles undergo nitro reduction in 20 minutes at room temperature, whereas electron-rich poly(4-methylstyrene) and poly(4-methoxystyrene) supported bimetallic nanoparticles to longer reaction times to go to completion. Electronics of the polymers also effects the change in mechanism of nitroreduction. Polystyrene bimetallic Ruthenium-Cobalt nanoparticles showed excellent yields and chemoselectivity towards nitro functional group in the presence of easily reducible functional groups like alkenes, alkynes, allyl ethers, propargyl ethers. Monometallic ruthenium nanoparticles also showed excellent reactivity and chemoselectivity towards azide reduction in the presence of easily reducible functional groups. Interestingly monometallic ruthenium nanoparticles showed regioselective reduction of primary azides in the presence of secondary and benzylic azides, also aromatic azides can be selectively reduced in the presence of secondary azides. These polystyrene supported nanoparticles are heterogeneous and are easily separated from the reaction mixture and reused multiple times without significant of catalytic activity.

RNA recognition motifs (RRMs) are a ubiquitous class of proteins which bind RNA in a sequence-specific fashion, often with high affinity. The mechanisms through which this single protein domain recognizes diverse RNA sequences is not fully understood. High-resolution three-dimensional structures are particularly important in understanding the structural features required for RNA recognition and binding. This work presents the structure of an expanded RRM domain from the Drosophila melanogaster Bruno protein. The Bruno protein is involved in establishing proper body patterning during development. This is accomplished through the translational repression of several mRNAs, in particular, the oskar mRNA. Previous work has identified an expanded RRM domain within the Bruno protein. This RRM requires an additional forty amino acids prior to the start of the canonical RRM domain for high affinity RNA binding. The protein was found to contain a canonical RRM domain comprised of four anti-parallel [beta] strands and two [alpha] helices. The RRM is preceded by a ten amino acid loop that interacts with [alpha]₁ and [beta]₂, while the remaining amino acids are flexible in solution. Interestingly, the deletion of these residues does not alter the fold or stability of the RRM domain. Thus, these additional residues must be involved in RNA binding, as they are not required for structure. From these studies, the Bruno RRM represents a new example of protein features required for recognition and high affinity binding of RNA. / text