HOX11 is a developmental regulator that plays a crucial role in the normal development of the spleen and is also aberrantly activated by the t(10;14)(q24;q11) and variant t(7;10)(q35;q24) translocations in a subset of T-cell acute lymphoblastic leukaemias (TALLs). The recent finding that HOX11 is deregulated in up to 40% of childhood TALLs when abnormalities not detected by cytogenetics are included, suggests that the over-expression of HOX11 and subsequent deregulation of downstream target genes are critical events in the progression of this tumour type. To date, three candidate HOX11 target genes have been reported, two of which are Aldehyde Dehydrogenase 1a1 (ALDH1A1) and Four and a Half LIM domain Protein 1 (FHL1). This investigation focused on two aspects of HOX11 function, namely its roles as a transcriptional regulator and as a nuclear oncoprotein capable of inducing neoplastic transformation. More specifically, we sought to further understand the role of HOX11 in tumorigenesis by 1) Confirming target gene status of ALDH1A1 and FHL1 by assessing whether their proximal promoter regions are transcriptionally regulated by HOX11, 2) Investigating the regulatory elements/transcriptional complexes involved in the response of ALDH1A1 to HOX11 in both a T-cell and an erythroid cell line in order to gain an insight into the mechanism(s) responsible for mediating a HOX11 activity and 3) Assessing the ability of ALDH1A1 and FHL1 to perturb normal patterns of haematopoiesis, on the basis that the transforming capabilities of HOX11 are thought to derive from its ability to affect haematopoietic cell differentiation. To confirm ALDH1A1 and FHL1 as target genes, they were both characterised in terms of the ability of their proximal promoters to be transcriptionally regulated by HOX11 using luciferase reporter assays. Significant repression of the proximal promoters of ALDH1A1 and FHL1 by HOX11 was observed in PER-117 T-cells which provided further evidence for their status as target genes. In the case of ALDH1A1, a CCAAT box (-74/-70bp) was identified as the primary cis-regulatory element involved in ALDH1A1 transcription and repression by HOX11 appeared to occur, either directly or indirectly, via interactions at the CCAAT box. Electromobility shift assays (EMSAs) revealed the disruption of a specific complex at this site by HOX11, which also altered the formation of complexes at a non-canonical TATA box (a GATA box at -34/-29bp). Significantly, HOX11 was shown to have the potential to interact with TFIIB, a member of the basal transcriptional complex. This, together with the presence of a TFIIB responsive element immediately 5’ of the GATA box, suggested that HOX11 may repress transcription by interfering with members of a preinitiation complex on the ALDH1A1 promoter. The transcriptional repression by HOX11 demonstrated in T-cells was dependent on DNA binding helix 3 of the homeodomain, suggesting that repression may require DNA binding. Alternatively, this region may be required for stable protein-protein interactions. In support of this, the in vitro association of HOX11 with TFIIB was disrupted upon deletion of helix 3, and the HOX11∆H3 mutant switched from a transcriptional repressor to a potent activator of transcription. Together, this data supports a model whereby HOX11 represses transcription by interfering with activation complexes at the CCAAT box and at the GATA box possibly via protein-protein interactions involving the homeodomain helix 3, whereas deletion of the region disables repressor-specific interactions, resulting in potent activation by HOX11. Luciferase reporter gene assays investigating the response of nested deletions of the ALDH1A1 promoter to HOX11 in the HEL900 erythroleukaemic cell line, also identified the CCAAT box (-74/-70bp) as the primary cis-regulatory element involved in ALDH1A1 transcription. However, in stark contrast to the its effect in T-cells, HOX11 was shown to activate transcription in the HEL cell line, both from the empty pGL3Basic luciferase reporter vector and from the ALDH1A1 promoter, in a manner independent of the homeodomain DNA binding helix 3. HOX11 thus appears to be a dichotomous regulator, capable of both transcriptional activation and repression depending on the circumstances. The mechanisms underlying these two functions are also appear to be distinct, with repression but not activation requiring the presence of homeodomain helix 3. ALDH1A1 encodes an enzyme involved in the irreversible conversion of retinaldehyde to the biologically active metabolite, retinoic acid (RA) and appears to be physiologically regulated by Hox11 in the developing spleen. Since RA is a potent modulator of cellular differentiation, proliferation and apoptosis, the dysregulation of RA synthesis is likely to have severe consequences for the cell and may constitute a mechanism whereby overexpression of HOX11 predisposes T-cells to malignant transformation. FHL1 also appears to have potential relevance to tumorigenesis, given that it encodes protein isoforms with suspected roles in transcriptional regulation. As a starting point to investigate a possible link between these HOX11 target genes and leukaemogenesis, the effect of overexpressing ALDH1A1 and FHL1 on murine haematopoiesis was assessed following reconstitution of lethally irradiated mice with retrovirally-transduced primary murine bone marrow cells. The enforced expression of ALDH1A1 in bone marrow was associated with a marked increase in myelopoiesis and a decrease in B and T-lymphopoiesis. By contrast, overexpression of FHL1 was not associated with perturbations in myelopoiesis or lymphopoiesis, although a slight increase in erythropoiesis was observed in the bone marrow. While further work is required to clarify the possible oncogenic roles of both of these HOX11 target genes, these findings have served to identify ALDH1A1 in particular, as a gene which could potentially be involved in HOX11-mediated tumorigenesis.

Characterization of the RNA transcriptome by next-generation sequencing can produce an unprecedented yield of information that provides novel biologic insights. I describe four approaches for sequencing different aspects of the transcriptome and provide computational tools to analyze the resulting data. Methods that query the dynamic range of gene expression, low expressing transcripts, micro RNA levels, and start-site usage of transcripts are described.

La dystrophie musculaire d'Emery-Dreifuss (EDMD) est caractérisée par des retractions précoces, une faiblesse et atrophie musculaire lentement progressive, et une atteinte cardiaque. Les mutations des gènes EMD et LMNA sont respectivement responsables de formes liées à l'X et de formes autosomiques de l'EDMD. Ces deux gènes codent pour des protéines de l'enveloppe nucléaire, l'émerine et les lamines A/C. Les mutations du gène FHL1 ont été impliquées dans d'autres cas d'EDMD liée à l'X. FHL1 codent pour FHL1A, FHL1B et FHL1C, protéines jamais décrites comme localisées à l'enveloppe nucléaire. Nous avons cherché à enrichir les connaissances sur la distribution subcellulaire des différentes isoformes de FHL1 dans les muscles squelettiques humains sains et malades. Nous avons mis en évidence que les isoformes FHL1 présentent à la fois une localisation cytoplasmique et nucléaire dans les myoblastes humains. Au noyau, FHL1B est fortement accumulé au niveau de l'enveloppe nucléaire où il interagit avec les lamines A/C et l'émerine. Cette localisation à l'enveloppe nucléaire est indépendante de l'expression de l'émerine ou des lamines A/C. La différenciation des myoblastes entraîne une forte réduction de l'expression de FHL1B et de son exclusion progressive du noyau, n'impliquant pas la protéine CRM-1. Nous avons mis en évidence l'augmentation de l'expression de FHL1B dans les myoblastes de deux patients atteints d'EDMD, l'un porteur d'une mutation dans le gène LMNA, l'autre dans le gène FHL1. En conclusion, la localisation spécifique de FHL1B et sa modulation dans les myoblastes de patients confirment les cas d'EDMD liés à FHL1 comme des pathologies de l'enveloppe nucléaire. / Emery-Dreifuss muscular dystrophy (EDMD) is characterized by the triad of early contractures, slowly progressive muscle wasting and weakness, and cardiac disease. Mutations in EMD and LMNA, encoding for the nuclear envelope (NE) proteins emerin and lamin A/C, are associated with X-linked and autosomal form of EDMD, respectively. The discovery that FHL, encoding FHL1A, FHL1B and FHL1C, is implicated in the pathogenesis of EDMD, raises the question of how a non-NE protein can be linked to emerin and lamin A/C. We aimed to provide knowledge of the subcellular distribution and expression of the various FHL1 isoforms in healthy and diseased human skeletal muscle. We found that FHL1 isoforms display a dual cytoplasmic and nuclear localization in human myoblasts. In addition, FHL1B strongly accumulated at the NE where it interacted with both lamin A/C and emerin. NE localization of FHL1B was independent of emerin and lamin A/C expression. Myoblast differentiation resulted in greatly reduced FHL1B protein expression and in the progressive nuclear exclusion of FHL1 protein isoforms. We have shown that chromosome region maintenance 1 (CRM1)-mediated nuclear export was not involved in the progressive decrease of nucleoplasmic FHL1B. Finally, we detected increased FHL1B protein levels in myoblasts of two patients with LMNA- and FHL1-related EDMD. Altogether, the specific localization of FHL1B and its modulation in disease-patient’s myoblasts confirmed FHL1-related EDMD as a NE disease.