Nutritional control of mRNA isoform expression during developmental arrest and recovery in C. elegans.

Nutrient availability profoundly influences gene expression. Many animal genes encode multiple transcript isoforms, yet the effect of nutrient availability on transcript isoform expression has not been studied in genome-wide fashion. When Caenorhabditis elegans larvae hatch without food, they arrest development in the first larval stage (L1 arrest). Starved larvae can survive L1 arrest for weeks, but growth and post-embryonic development are rapidly initiated in response to feeding. We used RNA-seq to characterize the transcriptome during L1 arrest and over time after feeding. Twenty-seven percent of detectable protein-coding genes were differentially expressed during recovery from L1 arrest, with the majority of changes initiating within the first hour, demonstrating widespread, acute effects of nutrient availability on gene expression. We used two independent approaches to track expression of individual exons and mRNA isoforms, and we connected changes in expression to functional consequences by mining a variety of databases. These two approaches identified an overlapping set of genes with alternative isoform expression, and they converged on common functional patterns. Genes affecting mRNA splicing and translation are regulated by alternative isoform expression, revealing post-transcriptional consequences of nutrient availability on gene regulation. We also found that phosphorylation sites are often alternatively expressed, revealing a common mode by which alternative isoform expression modifies protein function and signal transduction. Our results detail rich changes in C. elegans gene expression as larvae initiate growth and post-embryonic development, and they provide an excellent resource for ongoing investigation of transcriptional regulation and developmental physiology.

Sea Urchin Body Plan Development and Evolution: An Integrative Transcriptomic Approach

My dissertation work integrates comparative transcriptomics and functional analyses to investigate gene expression changes underlying two significant aspects of sea urchin evolution and development: the dramatic developmental changes associated with an ecologically significant shift in life history strategy and the development of the unusual radial body plan of adult sea urchins.

In Chapter 2, I investigate evolutionary changes in gene expression underlying the switch from feeding (planktotrophic) to nonfeeding (lecithotrophic) development in sea urchins. In order to identify these changes, I used Illumina RNA-seq to measure expression dynamics across 7 developmental stages in three sea urchin species: the lecithotroph Heliocidaris erythrogramma, the closely related planktotroph Heliocidaris tuberculata, and an outgroup planktotroph Lytechinus variegatus. My analyses draw on a well-characterized developmental gene regulatory network (GRN) in sea urchins to understand how the ancestral planktotrophic developmental program was altered during the evolution of lecithotrophic development. My results suggest that changes in gene expression profiles occurred more frequently across the transcriptome during the evolution of lecithotrophy than during the persistence of planktotrophy. These changes were even more pronounced within the GRN than across the transcriptome as a whole, and occurred in each network territory (skeletogenic, endomesoderm and ectoderm). I found evidence for both conservation and divergence of regulatory interactions in the network, as well as significant changes in the expression of genes with known roles in larval skeletogenesis, which is dramatically altered in lecithotrophs. I further explored network dynamics between species using coexpression analyses, which allowed me to identify novel players likely involved in sea urchin neurogenesis and endoderm patterning.

In Chapter 3, I investigate developmental changes in gene expression underlying radial body plan development and metamorphosis in H. erythrogramma. Using Illumina RNA-seq, I measured gene expression profiles across larval, metamorphic, and post-metamorphic life cycle phases. My results present a high-resolution view of gene expression dynamics during the complex transition from pre- to post-metamorphic development and suggest that distinct sets of regulatory and effector proteins are used during different life history phases.

Collectively, my investigations provide an important foundation for future, empirical studies to investigate the functional role of gene expression change in the evolution of developmental differences between species and also for the generation of the unusual radial body plan of sea urchins.

HDAC inhibitors cause site-specific chromatin remodeling at PU.1-bound enhancers in K562 cells

BACKGROUND: Small molecule inhibitors of histone deacetylases (HDACi) hold promise as anticancer agents for particular malignancies. However, clinical use is often confounded by toxicity, perhaps due to indiscriminate hyperacetylation of cellular proteins. Therefore, elucidating the mechanisms by which HDACi trigger differentiation, cell cycle arrest, or apoptosis of cancer cells could inform development of more targeted therapies. We used the myelogenous leukemia line K562 as a model of HDACi-induced differentiation to investigate chromatin accessibility (DNase-seq) and expression (RNA-seq) changes associated with this process. RESULTS: We identified several thousand specific regulatory elements [~10 % of total DNase I-hypersensitive (DHS) sites] that become significantly more or less accessible with sodium butyrate or suberanilohydroxamic acid treatment. Most of the differential DHS sites display hallmarks of enhancers, including being enriched for non-promoter regions, associating with nearby gene expression changes, and increasing luciferase reporter expression in K562 cells. Differential DHS sites were enriched for key hematopoietic lineage transcription factor motifs, including SPI1 (PU.1), a known pioneer factor. We found PU.1 increases binding at opened DHS sites with HDACi treatment by ChIP-seq, but PU.1 knockdown by shRNA fails to block the chromatin accessibility and expression changes. A machine-learning approach indicates H3K27me3 initially marks PU.1-bound sites that open with HDACi treatment, suggesting these sites are epigenetically poised. CONCLUSIONS: We find HDACi treatment of K562 cells results in site-specific chromatin remodeling at epigenetically poised regulatory elements. PU.1 shows evidence of a pioneer role in this process by marking poised enhancers but is not required for transcriptional activation.

Contrasting transcriptome response to thermal stress in two key zooplankton species, Calanus finmarchicus and C. glacialis

Climate change has already led to the range expansion of warm-water plankton assemblages in the northeast Atlantic and the corresponding range contraction of colder-water species. The temperate copepod Calanus finmarchicus is predicted to shift farther northward into polar waters traditionally dominated by the arctic copepod C. glacialis. To identify temperaturemediated changes in gene expression that may be critical for the thermal acclimation and resilience of the 2 Calanus spp., we conducted a whole transcriptome profiling using RNA-seq on an Ion Torrent platform. Transcriptome responses of C. finmarchicus and C. glacialis from Disko Bay, west Greenland, were investigated under realistic thermal stresses (at + 5, +10 and +15°C) for 4 h and 6 d. C. finmarchicus showed a strong response to temperature and duration of stress, involving up-regulation of genes related to protein folding, transcription, translation and metabolism. In sharp contrast, C. glacialis displayed only low-magnitude changes in gene expression in response to temperature and duration of stress. Differences in the thermal responses of the 2 species, particularly the lack of thermal stress response in C. glacialis, are in line with laboratory and field observations and suggest a vulnerability of C. glacialis to climate change.

Genetic and genomic analyses of musculoskeletal differences between BEH and BEL strains

Berlin high (BEH) and Berlin low (BEL) strains selected for divergent growth differ 3-fold in body weight. We aimed at examining muscle mass, which is a major contributor to body weight, by exploring anatomical characteristics of the soleus muscle, its fiber numbers and their cross sectional area (CSA), by analysing transcriptome of the gastrocnemius and by initiating quantitative trait locus (QTL) mapping. BEH muscles were 4-to-8 times larger compared to BEL strain. In sub-strain BEH+/+, mutant myostatin was replaced with a wild type allele, however, BEH+/+muscles still were 2-to-4 times larger compared to the BEL strain. BEH soleus contained 2-times more (P<0.0001) and 2-times larger in CSA (P<0.0001) fibers compared to BEL strain. In addition, soleus femoral attachment anomaly (SFAA) was observed in all BEL mice. One significant (chromosome 1) and four suggestive (chromosomes 3, 4, 6 and 9) muscle weight QTLs were mapped in 21-day old F2 intercross (n=296) between BEH and BEL strains. The frequency of SFAA incidence in the F2 and in the backcross to BEL strain (BCL) suggested the presence of more than one causative gene. Two suggestive SFAA QTLs were mapped in BCL, however, their peak markers were not associated with the phenotype in F2. RNA-Seq analysis revealed 2,148 differentially expressed (P<0.1) genes and 45,673 SNPs and >2,000 indels between BEH+/+ and BEL males. In conclusion, contrasting muscle traits, genomic and gene expression differences between BEH and BEL strains provide a promising model for the search of genes involved in muscle growth and musculoskeletal morphogenesis.

Small RNAs from plants, bacteria and fungi within the order Hypocreales are ubiquitous in human plasma

Background

The human microbiome plays a significant role in maintaining normal physiology. Changes in its composition have been associated with bowel disease, metabolic disorders and atherosclerosis. Sequences of microbial origin have been observed within small RNA sequencing data obtained from blood samples. The aim of this study was to characterise the microbiome from which these sequences are derived.

Abundant non-human small RNA sequences were identified in plasma and plasma exosomal samples. Assembly of these short sequences into longer contigs was the pivotal novel step in ascertaining their origin by BLAST searches. Most reads mapped to rRNA sequences. The taxonomic profiles of the microbes detected were very consistent between individuals but distinct from microbiomes reported at other sites. The majority of bacterial reads were from the phylum Proteobacteria, whilst for 5 of 6 individuals over 90% of the more abundant fungal reads were from the phylum Ascomycota; of these over 90% were from the order Hypocreales. Many contigs were from plants, presumably of dietary origin.  In addition, extremely abundant small RNAs derived from human Y RNAs were detected.

A characteristic profile of a subset of the human microbiome can be obtained by sequencing small RNAs present in the blood. The source and functions of these molecules remain to be determined, but the specific profiles are likely to reflect health status. The potential to provide biomarkers of diet and for the diagnosis and prognosis of human disease is immense.

The mitochondrial and autosomal mutation landscapes of prostate cancer

BACKGROUND: Prostate cancer (PCa) is the most common cancer in men. PCa is strongly age associated; low death rates in surveillance cohorts call into question the widespread use of surgery, which leads to overtreatment and a reduction in quality of life. There is a great need to increase the understanding of tumor characteristics in the context of disease progression.

OBJECTIVE: To perform the first multigenome investigation of PCa through analysis of both autosomal and mitochondrial DNA, and to integrate exome sequencing data, and RNA sequencing and copy-number alteration (CNA) data to investigate how various different tumor characteristics, commonly analyzed separately, are interconnected.

DESIGN, SETTING, AND PARTICIPANTS: Exome sequencing was applied to 64 tumor samples from 55 PCa patients with varying stage and grade. Integrated analysis was performed on a core set of 50 tumors from which exome sequencing, CNA, and RNA sequencing data were available.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Genes, mutated at a significantly higher rate relative to a genomic background, were identified. In addition, mitochondrial and autosomal mutation rates were correlated to CNAs and proliferation, assessed as a cell cycle gene expression signature.

RESULTS AND LIMITATIONS: Genes not previously reported to be significantly mutated in PCa, such as cell division cycle 27 homolog (Saccharomyces cerevisiae) (CDC27), myeloid/lymphoid or mixed-lineage leukemia 3 (MLL3), lysine (K)-specific demethylase 6A (KDM6A), and kinesin family member 5A (KIF5A) were identified. The mutation rate in the mitochondrial genome was 55 times higher than that of the autosomes. Multilevel analysis demonstrated a tight correlation between high reactive-oxygen exposure, chromosomal damage, high proliferation, and in parallel, a transition from multiclonal indolent primary PCa to monoclonal aggressive disease. As we only performed targeted sequence analysis; copy-number neutral rearrangements recently described for PCa were not accounted for.

CONCLUSIONS: The mitochondrial genome displays an elevated mutation rate compared to the autosomal chromosomes. By integrated analysis, we demonstrated that different tumor characteristics are interconnected, providing an increased understanding of PCa etiology.

Sequence and expression of complement factor H gene cluster variants and their roles in age-related macular degeneration risk

Purpose: To investigate how potentially functional genetic variants are coinherited on each of four common complement factor H (CFH) and CFH-related gene haplotypes and to measure expression of these genes in eye and liver tissues.

Methods: We sequenced the CFH region in four individuals (one homozygote for each of four common CFH region haplotypes) to identify all genetic variants. We studied associations between the haplotypes and AMD phenotypes in 2157 cases and 1150 controls. We examined RNA-seq profiles in macular and peripheral retina and retinal pigment epithelium/choroid/sclera (RCS) from eight eye donors and three liver samples.

Results: The haplotypic coinheritance of potentially functional variants (including missense variants, novel splice sites, and the CFHR3–CFHR1 deletion) was described for the four common haplotypes. Expression of the short and long CFH transcripts differed markedly between the retina and liver. We found no expression of any of the five CFH-related genes in the retina or RCS, in contrast to the liver, which is the main source of the circulating proteins.

Conclusions: We identified all genetic variants on common CFH region haplotypes and described their coinheritance. Understanding their functional effects will be key to developing and stratifying AMD therapies. The small scale of our expression study prevented us from investigating the relationships between CFH region haplotypes and their expression, and it will take time and collaboration to develop epidemiologic-scale studies. However, the striking difference between systemic and ocular expression of complement regulators shown in this study suggests important implications for the development of intraocular and systemic treatments.

Host-symbiont interactions in the deep-sea vent mussel Bathymodiolus azoricus : a molecular approach

Tese de Doutoramento, Ciências do Mar, especialidade de Biologia Marinha, 19 de Dezembro de 2015, Universidade dos Açores.

CLOCK-controlled polyphonic regulation of circadian rhythms through canonical and noncanonical E-boxes.

In mammalian circadian clockwork, the CLOCK-BMAL1 complex binds to DNA enhancers of target genes and drives circadian oscillation of transcription. Here we identified 7,978 CLOCK-binding sites in mouse liver by chromatin immunoprecipitation-sequencing (ChIP-Seq), and a newly developed bioinformatics method, motif centrality analysis of ChIP-Seq (MOCCS), revealed a genome-wide distribution of previously unappreciated noncanonical E-boxes targeted by CLOCK. In vitro promoter assays showed that CACGNG, CACGTT, and CATG(T/C)G are functional CLOCK-binding motifs. Furthermore, we extensively revealed rhythmically expressed genes by poly(A)-tailed RNA-Seq and identified 1,629 CLOCK target genes within 11,926 genes expressed in the liver. Our analysis also revealed rhythmically expressed genes that have no apparent CLOCK-binding site, indicating the importance of indirect transcriptional and posttranscriptional regulations. Indirect transcriptional regulation is represented by rhythmic expression of CLOCK-regulated transcription factors, such as Krüppel-like factors (KLFs). Indirect posttranscriptional regulation involves rhythmic microRNAs that were identified by small-RNA-Seq. Collectively, CLOCK-dependent direct transactivation through multiple E-boxes and indirect regulations polyphonically orchestrate dynamic circadian outputs.

Genomic and transcriptomic analysis of the streptomycin-dependent Mycobacterium tuberculosis strain 18b.

The ability of Mycobacterium tuberculosis to establish a latent infection (LTBI) in humans confounds the treatment of tuberculosis. Consequently, there is a need to discover new therapeutic agents that can kill M. tuberculosis both during active disease and LTBI. The streptomycin-dependent strain of M. tuberculosis, 18b, provides a useful tool for this purpose since upon removal of streptomycin (STR) it enters a non-replicating state that mimics latency both in vitro and in animal models. The 4.41 Mb genome sequence of M. tuberculosis 18b was determined and this revealed the strain to belong to clade 3 of the ancient ancestral lineage of the Beijing family. STR-dependence was attributable to insertion of a single cytosine in the 530 loop of the 16S rRNA and to a single amino acid insertion in the N-terminal domain of initiation factor 3. RNA-seq was used to understand the genetic programme activated upon STR-withdrawal and hence to gain insight into LTBI. This revealed reconfiguration of gene expression and metabolic pathways showing strong similarities between non-replicating 18b and M. tuberculosis residing within macrophages, and with the core stationary phase and microaerophilic responses. The findings of this investigation confirm the validity of 18b as a model for LTBI, and provide insight into both the evolution of tubercle bacilli and the functioning of the ribosome.

Alternative splicing by hnRNP L as a new modulator of hematopoietic cell differentiation, survival and migration

Les modifications post-transcriptionnelles de l’ARN messager (ARNm), comme l’épissage alternatif, jouent un rôle important dans la régulation du développement embryonnaire, de la fonction cellulaire et de l’immunité. De nouvelles évidences révèlent que l’épissage alternatif serait également impliqué dans la régulation de la maturation et de l’activation des cellules du système hématopoïétique. Le facteur hnRNP L a été identifié comme étant le principal régulateur de l’épissage alternatif du gène codant pour le récepteur CD45 in vitro. Le récepteur CD45 est une tyrosine phosphatase exprimée par toutes les cellules du système hématopoïétique qui contrôle le développement et l’activation des lymphocytes T. Dans un premier temps, nous avons étudié la fonction du facteur hnRNP L dans le développement des lymphocytes T et dans l’épissage de l’ARNm de CD45 in vivo en utilisant des souris dont le gène de hnRNP L a été supprimé spécifiquement dans les cellules T. La délétion de hnRNP L dans les thymocytes résulte en une expression aberrante des différents isoformes de CD45 avec une prédominance de l'isoforme CD45RA qui est généralement absent dans le thymus. Une conséquence de la délétion de hnRNP L est une diminution de la cellularité du thymus causée par un blocage partiel du développement des cellules pré-T au stade DN4. Cette réduction du nombre de cellules dans le thymus n’est pas liée à une hausse de la mort cellulaire. Les thymocytes déficients pour hnRNP L démontrent plutôt une prolifération augmentée comparée aux thymocytes sauvages due à une hyper-activation des kinases Lck, Erk1/2 et Akt. De plus, la délétion de hnRNP L dans le thymus cause une perte des cellules T en périphérie. Les résultats des expériences in vitro suggèrent que cette perte est principalement due à un défaut de migration des thymocytes déficients pour hnRNP L du thymus vers la périphérie en réponse aux chimiokines. L’épissage alternatif de CD45 ne peut expliquer ce phénotype mais l’identification de cibles par RNA-Seq a révélé un rôle de hnRNP L dans la régulation de l’épissage alternatif de facteurs impliqués dans la polymérisation de l’actine. Dans un second temps, nous avons étudié le rôle de hnRNP L dans l’hématopoïèse en utilisant des souris dont la délétion de hnRNP L était spécifique aux cellules hématopoïétiques dans les foies fœtaux et la moelle osseuse. L’ablation de hnRNP L réduit le nombre de cellules progénitrices incluant les cellules progénitrices lymphocytaires (CLPs), myéloïdes (CMPs, GMPs) et mégakaryocytes-érythrocytaires (MEPs) et une perte des cellules hématopoïétiques matures. À l’opposé des cellules progénitrices multipotentes (MPPs) qui sont affectées en absence de hnRNP L, la population de cellules souches hématopoïétiques (HSCs) n’est pas réduite et prolifère plus que les cellules contrôles. Cependant, les HSCs n’exprimant pas hnRNP L sont positives pour l'Annexin V et expriment CD95 ce qui suggère une mort cellulaire prononcée. Comme pour les thymocytes, une analyse par RNA-Seq des foies fœtaux a révélé différents gènes cibles de hnRNP L appartenant aux catégories reliées à la mort cellulaire, la réponse aux dommages à l’ADN et à l’adhésion cellulaire qui peuvent tous expliquer le phénotype des cellules n’exprimant pas le gène hnRNP L. Ces résultats suggèrent que hnRNP L et l’épissage alternatif sont essentiels pour maintenir le potentiel de différenciation des cellules souches hématopoïétiques et leur intégrité fonctionnelle. HnRNP L est aussi crucial pour le développement des cellules T par la régulation de l’épissage de CD45 ainsi que pour leur migration.

Genomic variation in recombination patterns : implications for disease and cancer

Durant la méiose, il se produit des échanges réciproques entre fragments de chromosomes homologues par recombinaison génétique. Les chromosomes parentaux ainsi modifiés donnent naissance à des gamètes uniques. En redistribuant les mutations génétiques pour générer de nouvelles combinaisons, ce processus est à l’origine de la diversité haplotypique dans la population. Dans cette thèse, je présente des résultats décrivant l’implication de la recombinaison méiotique dans les maladies chez l’humain. Premièrement, l'analyse statistique de données de génotypage de familles québécoises démontre une importante hétérogénéité individuelle et sexe-spécifique des taux de recombinaisons. Pour la première fois chez l’humain, nous avons observé que le taux de recombinaison maternel diminue avec l'âge de la mère, un phénomène potentiellement impliqué dans la régulation du taux d’aneuploïdie associé à l’âge maternel. Ensuite, grâce à l’analyse de données de séquençage d’exomes de patients atteints de leucémie et de ceux de leurs parents, nous avons découvert une localisation anormale des évènements de recombinaison chez les enfants leucémiques. Le gène PRDM9, principal déterminant de la localisation des recombinaisons chez l’humain, présente des formes alléliques rares dans ces familles. Finalement, en utilisant un large spectre de variants génétiques identifiés dans les transcriptomes d’individus Canadiens Français, nous avons étudié et comparé le fardeau génétique présent dans les régions génomiques à haut et à faible taux de recombinaison. Le fardeau génétique est substantiellement plus élevé dans les régions à faible taux de recombinaison et nous démontrons qu’au niveau individuel, ce fardeau varie selon la population humaine. Grâce à l’utilisation de données génomiques de pointe pour étudier la recombinaison dans des cohortes populationnelles et médicales, ce travail démontre de quelle façon la recombinaison peut affecter la santé des individus.

Characterization of the Sterile20 kinase Slik : A regulator of growth in Drosophila

La prolifération cellulaire et la croissance tissulaire sont étroitement contrôlées au cours du développement. Chez la Drosophila melanogaster, ces processus sont régulés en partie par la kinase stérile-20 Slik (SLK et LOK chez les mammifères) et le suppresseur de tumeur Hippo (Hpo, MST1/2 chez les mammifères) dans les cellules épithéliales. La surexpression de la kinase Slik augmente la taille des tissus chez les mouches adultes. Cependant, les mutants slik-/- meurent avant d'avoir terminé leur développement. Lorsqu’elle est surexprimée dans les cellules épithéliales des ailes en voie de développement, cette protéine favorise la prolifération cellulaire. En outre, l'expression de Slik dans une population de cellules conduit à une surprolifération des cellules voisines, même quand elles sont physiquement séparées. Ceci est probablement dû à la sécrétion de facteurs de croissance qui stimulent la prolifération de manière paracrine. En utilisant des méthodes génétiques et transcriptomiques, nous essayons de déterminer les molécules et les mécanismes impliqués. Contrairement à ce qui a été publié, nous avons constaté que Slik ne transmet pas de signal prolifératif en inhibant le suppresseur de tumeur Merlin (Mer, NF2 chez les mammifères), un composant en amont de la voie Hippo. Plutôt, elle favorise la prolifération non-autonome et la croissance des tissus en signalisation par la kinase dRaf (la seule kinase de la famille Raf chez la drosophile). Nous prouvons que dRaf est nécessaire chez les cellules voisines pour conduire la prolifération chez ces cellules. De plus, nous avons utilisé le séquençage du transcriptome pour identifier de nouveaux effecteurs en aval de Slik. Ce qui permettra de mieux comprendre les effets de SLK et LOK chez les humains.

Expression des histones déméthylases dans les cellules hématopoïétiques humaines et les leucémies aiguës

L’importance des modificateurs de la chromatine dans la régulation de l’hématopoïèse et des hémopathies malignes est illustrée par l’histone méthyltransférase Mixed-Lineage Leukemia (MLL) qui est essentielle au maintien des cellules souches hématopoïétiques (CSH) et dont le gène correspondant, MLL, est réarrangé dans plus de 70% des leucémies du nourrisson. Les histones déméthylases (HDM), récemment découvertes, sont aussi impliquées dans le destin des CSH et des hémopathies malignes. Le but de ce projet est d’étudier l’expression des HDM dans les cellules hématopoïétiques normales et leucémiques afin d’identifier de potentiels régulateurs de leur destin. Nous avons réalisé un profil d'expression génique des HDM par qRT-PCR et par séquençage du transcriptome (RNA-seq) dans des cellules de sang de cordon (cellules CD34+ enrichies en CSH et cellules différenciées) et des cellules de leucémie aiguë myéloïde (LAM) avec réarrangement MLL. Les deux techniques montrent une expression différentielle des HDM entre les populations cellulaires. KDM5B et KDM1A sont surexprimés dans les cellules CD34+ par rapport aux cellules différenciées. De plus, KDM4A et PADI2 sont surexprimés dans les cellules leucémiques par rapport aux cellules normales. Des études fonctionnelles permettront de déterminer si la modulation de ces candidats peut être utilisée dans des stratégies d’expansion des CSH, ou comme cible thérapeutique anti-leucémique. Nous avons aussi développé et validé un nouveau test diagnostique pour détecter les mutations de GATA2 qui code pour un facteur de transcription clé de l’hématopoïèse impliqué dans les LAM. Ces travaux soulignent l’importance des facteurs nucléaires dans la régulation de l’hématopoïèse normale et leucémique.