The IRCM
Research
The Clinic
Studying at IRCM
Core facilities
Conferences
Careers
Molecular Genetics
Skip navigation links
Director
Team
Photos
Projects
Publications
Contact us


Research unit director
Projects

2009 - To be updated soon

Gene expression in eukaryotes
In complex organisms, several thousand genes need to be expressed in specific cells and at specific times during the course of development in order for a viable organism to develop. We now know of many signals that are required for the stringent control of gene expression. Some of these signals are built into the cell differentiation program, others derive from cell:cell interactions, from interactions between cells and extracellular matrix, or they are produced in response to growth factors, hormones or neurotransmitters. We have been using a hormone-coding gene to understand at the molecular level the integration of these various developmental and hormonal signals in the control of gene expression. The pro-opiomelanocortin (POMC) gene encodes hormones, which are vital to the survival of most mammals, including man. Indeed, POMC is the polypeptide precursor of ACTH, a pituitary hormone fundamental to the maintenance of essential body functions mediated by glucocorticoids. These adrenal steroids affect intermediate metabolism, immune and inflammatory responses and they mediate the body's response to stress. In addition, the analysis of the complex regulatory pathways operating at the POMC gene has led our laboratory to study, in addition to transcriptional regulation, the mechanisms of hormone action, the mechanisms for pituitary cell differentiation as well as for early development of limbs and craniofacial structures.

Cell specificity of POMC gene expression
The same POMC gene is expressed in two different cell populations of the pituitary gland, anterior pituitary corticotroph cells found in the anterior lobe and intermediate lobe melanotrophs. However, the processing of the POMC precursor differs in the two cell types, thus leading to the secretion of different hormones by each cell. Accordingly, the control of POMC gene expression is specific and different for each cell. Using transgenic animals and gene transfer into cells, we have identified control sequences of the POMC gene promoter that account for cell-specific transcription and hormonal regulation (Figure 1).

Figure 1:
This analysis has led us to clone novel transcription factors. One is Pitx1 (Ptx1), a homeobox factor related to the bicoid family. Pitx1 is essential for POMC gene transcription and in its absence, the complex of transcription factors associated with the POMC promoter is inefficient. This role of "angular stone" played by Pitx1 within a complex of transcription factors was also found for other pituitary promoters (e.g. the LHb promoter), and it appears that this role results from direct interaction between proteins. Indeed, the other transcription factor that is essential for cell-specific transcription of POMC is the Tbox factor Tpit. This factor is only present in the two cell lineages of the pituitary that express POMC and it is critical for POMC expression as well as for differentiation of POMC-expressing pituitary cells. Moreover, we have identified mutations in the human TPIT gene that cause isolated ACTH deficiency (IAD). Another factor for cell-specific expression of POMC is the bHLH factor, NeuroD1/BETA2. This factor is only present in POMC-expressing corticotrophs in the anterior pituitary whereas it is not present in POMC-expressing melanotroph cells of the intermediate pituitary. It is also found in pancreas insulin cells, some gut cells and in neurons. On the POMC promoter, corticotroph specificity of transcription results from the protein:protein interaction between bHLH factors (including NeuroD) and Pitx1. This direct interaction takes place between the DNA binding domains of each protein.


Pituitary cell differentiation
The pituitary gland is an interesting model to identify regulatory mechanisms for cell differentiation because each cell line of this endocrine gland is easily recognized by the hormone that it produces. Indeed, the adult gland contains six different cell types, each dedicated to the production of a different hormone. Prior work by different investigators had shown a common origin for differentiation of somatotroph cells (producing growth hormone GH), of thyrotrophs (producing TSH) and of lactotrophs (producing prolactin PRL).

Figure 2:
We showed by gene knockout of the Tpit gene that corticotroph cells (that process POMC into ACTH) and melanotroph cells (that process POMC into;MSH in the intermediate lobe) share a common precursor with gonadotroph cells (that produce the gonadotropins LH and FSH). By establishing a relation between these latter cell lineages and by demarcating those from the GH, TSH and PRL lineages, we were led to propose a simple binary model that accounts for differentiation of all pituitary cells (Figure 2).



Gene regulatory network
We also use the pituitary model to define globally regulatory interactions that implicate the different transcription factors for the complex developmental program of pituitary organogenesis and function. Towards this end, we use genomic expression profiling techniques and we have also developed a ChIP-on-chip (chromatin immunoprecipitation analysed on DNA chip) technology for identification of all genes targeted by specific pituitary transcription factors. We are applying these genomic approaches to the pituitary cell-specific factors such as Tpit, NeuroD and Pitx1/2/3 as well as to transcription factors mediating hormone action such as GR and NGFI-B.

Isolated ACTH deficiency
The discovery of Tpit and its highly restricted expression in POMC cells of the pituitary led us to make the prediction that mutations in the human TPIT gene would lead to a specific deficit in pituitary POMC. At the onset, this appeared to be an extremely rare human condition as only a few children had been described in the clinical literature with a similar condition. However, the analysis of the DNA from two of these children led us to discover TPIT gene mutations that cause IAD. Immediately after we published these first mutations, we received patient DNAs from pediatric endocrinologists of around the world who had become aware of this condition. It now appears that IAD is a very homogenous disease, that it had been under diagnosed until we described it and that one reason for this under-diagnosis might have been the neonatal death of many children carrying the disease.

Figure 3:
We have now identified over a dozen TPIT mutations causing IAD (Figure 3). The diagnosis of this inherited hormone deficiency has clear and immediate benefit to patients and family since hormone replacement therapy can prevent neonatal death and ensure perfectly normal development and life for the affected children.



Activation of POMC transcription by CRH

Both POMC gene transcription as well as ACTH secretion are stimulated by the hypothalamic hormone CRH. In corticotroph cells of the pituitary, CRH action on its membrane receptor is mediated through activation of the cyclic AMP/protein kinase A pathway. These signals activate expression and activity of three transcription factors of the orphan nuclear receptor family, namely the factors of the Nur77 (NGFI-B) subfamily.

Figure 4:
This subfamily also has two other members, Nurr1 and NOR1. These three orphan nuclear receptors act as mediators for the transcriptional effect of CRH and they do so by binding, either as homo- or heterodimers, to a unique regulatory element of the POMC gene (Figure 4). We were the first to document this mode of action that also appears to be important for the mediator role of Nur transcription factors in control of T lymphocytes apoptosis (programmed cell death).

 

 

Role of the tumour suppressor Rb in hormone responsiveness
The tumor suppressor Rb plays a particularly limiting role in melanotrophs of the mouse pituitary since the loss of one allele of this gene leads to the formation of melanotroph tumors in all carrier mice. As a first step towards understanding the reason for this limiting role, we studied Rb and the related protein p107 for their implication in POMC cell transcription and found that these two proteins act as modulators of the cellular response to CRH.

Figure 5:
Indeed, the level of Rb proteins in pituitary POMC cells changes the gain in the hormone response to signals elicited by CRH: this modulatory action of Rb results for multiple protein interactions between Rb and NGFI-B as well as between Rb and SRC2, the NGFI-B coactivator. We extended this paradigm to a subset of nuclear receptors (each of these transcription factors is inducible by a specific ligand) (Figure 5). Rb also modulates the transcriptional activity of NeuroD.


 

 

 

Mechanisms of transcriptional repression by steroid hormones
Pituitary ACTH stimulates adrenal steroidogenesis and glucocorticoid production. Upon their release into the blood stream, glucocorticoids exert a negative feedback on pituitary ACTH secretion and on transcription of the POMC gene. This classical negative feedback loop was one of the first to be described and it provides a most relevant paradigm to understand the mechanisms by which some genes are repressed by the same factors (steroid receptors) which activate other genes.

Figure 6:
Our analysis of the POMC promoter has led us to identify a negative glucocorticoid response element (nGRE) that can bind three subunits of the glucocorticoid receptor. The nGRE provides one mechanism for repression of POMC. The glucocorticoid receptors are also uses another mechanism to repress POMC gene transcription (Figure 6). 

 


 

Figure 7:
Specifically, glucocorticoids blunt the induction of Nur receptors in response to CRH, and GR blunts the action of Nur receptors on their transcriptional targets. This antagonism results from protein: protein interactions and the mechanism of these interactions appears to be very similar to another protein: protein antagonism that we have described in the past, which is observed between the glucocorticoid receptor and the AP1 transcription factors, jun and fos (Figure 7).

 

 

Glucocorticoid resistance, corticotroph adenomas and Cushing disease
The investigation of the mechanism of trans-repression exerted between the glucocorticoid receptor (GR) and transcription factors of the NGFI-B/Nur77 family led us to identify proteins that are absolutely required for these transcriptional antagonisms. Indeed, we found that Brg1, the ATPase subunit of the Swi/Snf chromatin remodeling complex, acts as an essential scaffolding protein for formation of a protein complex required for trans-repression. In this protein complex, Brg1 is also required to recruit and interact with the histone deacetylase HDAC2. The activity of both Brg1 and HDAC2 for chromatin remodeling appear to be essential for their implication in trans-repression and thus, we have proposed a model for this active trans-repression that implicates chromatin remodeling (Figure 8).

Figure 8:

Expression of either Brg1 or HDAC 2 was found to be abnormal in about half of the human corticotroph adenomas that were studied following surgery in patients affected by Cushing disease. Most strikingly, some tumors have cytoplasmic Brg1 instead of the normal nuclear expression of this protein; other tumors do not express Brg1. The loss of Brg1 expression in the nuclei of corticotroph adenoma cells provides a molecular explanation for the resistance of these cells to glucocorticoids, this hormone resistance being the hallmark of corticotroph adenomas and Cushing disease. In addition, the loss of either Brg1 or HDAC2 may also contribute to the initiation of tumor formation in these patients by leading to corticotroph hyperplasia.

Control of the cell cycle during pituitary development
We also investigate the control of the cell cycle during the formation of the pituitary gland as well as in the adult tissue in order to understand the process of the pituitary tumor formation, such as in Cushing disease.

We are investigating a mechanism for cell cycle exit that precedes the initiation of cell differentiation during normal mouse pituitary development and we have identified markers of each of the steps involved in this transition. Some of these markers are reactivated in human pituitary tumors, and we are currently studying the contribution of each of these genes in the tumorigenic process.

Role of Pitx genes in limb specification
The Pitx1 gene is expressed very early in lateral mesoderm, in addition to its stomodeal expression. Within the lateral plate mesoderm, Pitx1 expression is exclusively observed in the posterior half of the embryo, resulting in expression in hindlimb buds but not in forelimb buds. The inactivation (knockout) of the Pitx1 gene in mice showed that Pitx1 is essential for proper specification of hindlimb character since hindlimbs partly resemble forelimbs in these mice. Gain-of-function experiments performed in chicken have confirmed this role showing that forced expression of Pitx1 in forelimb buds will result in chicken wings that are partially transformed into legs.

Figure 9:
This work also showed that Pitx1 is the upstream regulatory gene of a cascade that also includes Tbx4 for specification of hindlimb identity. The related factor Tbx5 may play a similar role in forelimbs (Figure 9). We currently investigate the action of Pitx1 in specification of limb identity as well as its genetic interaction with another gene of this subfamily, namely Pitx2. Indeed, both Pitx1 and Pitx2 act together in the control of early limb bud growth.

 

 

Role of Pitx3 transcription factor
The third member of the Pitx family, Pitx3, is expressed in the eyes and in only one population of neurons of the brain, the dopaminergic neurons of the midbrain. In man, Pitx3 gene mutations have been associated with a familial form of cataracts, which is consistent with one of the developmental roles of this gene. In the midbrain, Pitx3 expression starts just before terminal differentiation of dopaminergic neurons. The timing of this expression is consistent with a role of Pitx3 in the differentiation and/or maintenance of these neurons, which are otherwise responsible through their degeneration for Parkinson's disease (Figure 10). We are investigating the role of Pitx3 during dopaminergic neuron development and differentiation as well as the control of gene expression in these neurons.

Figure 10:

All rights reserved: IRCM 2011_110 avenue des Pins Ouest - Montréal (Québec) H2W 1R7 – Canada
.