Final Report Summary - RXR IN MYELOID CELLS (Contribution of RXRs to myeloid cell biology: from hematopoietic stem cells to osteoclastogenesis)
Hematopoietic stem cells (HSCs) yield blood precursors devoted to unilineage differentiation and production of mature blood cells. Precursors of the myeloid lineage differentiate not only into monocytes and granulocytes but also they can differentiate from various stages of maturity into multinucleated osteoclasts (OCs). Consequently, any problem in myelopoiesis can be directly linked to deficiencies in osteclastogenesis and bone diseases.
Retinoic X receptors (RXR-alpha, RXR-beta and RXR-gamma) are members of the nuclear receptor family. They appear to contribute to hematopoiesis and osteoclastogenesis via their role as heterodimeric partner of several nuclear receptors. Because RXRs can also function as homodimers, it is unclear whether the role of RXRs on hematopoiesis and osteoclastogenesis occurred through the same pathway or whether it involves the regulation of parallel pathways. Additionally, while the RXR isoform is functionally the most important RXR isotype during morphogenesis, analyses of different knock-out mice demonstrate a partial redundancy in RXR functions.
In this context, our main goal was to characterise the role of RXR in HSC function and osteoclastogenesis and the two main objectives addressed in the project were:
(1) to test the hypothesis that RXR is compensating the in vivo effects of conditional inhibition of RXR in the murine adult hematopoietic system;
(2) to clarify the contribution of RXR signalling during myelomonocytic differentiation and osteoclastogenesis processes; our hypothesis is that RXR influences myelopoiesis and osteoclastogenesis exerting direct actions as homodimer.
To address these questions, we have carried out a cross-disciplinary research project (involving gene engineered animal models, in vivo phenotyping, ex vivo cell biology techniques and in vitro molecular biology assays) that allowed us to obtain a complete overview of how RXR signaling contributes to bone physiology; from the molecular mechanisms controlled by RXR activation in cells to the impact that RXR deficiency has in rodents.
In this project we worked with a knock-out mouse model in which both RXRa and RXRß were selectively ablated in the hematopoietic system (dKO mice). These mice presented a phenotype aimed to develop osteopetrosis that was not observed in mouse lacking only the RXR isoform. Thus, in these mice, RXR at basal levels of expression is compensating for the loss of RXR and therefore, deletion of both isoforms is needed in order to study the role of RXR in hematopoiesis. The characterisation of the osteopetrotic phenotype presented by the dKO mice at the age of five months revealed a defect in bone remodelling process whose origin appear to be the presence of gigantic non-resorbing OC attached to dKO bones. According to these in vivo observations, differences in the size of dKO OC cultured in vitro were also detected. Indeed, in vitro differentiated dKO OC were bigger in size and less active when compared with WT OC. OC are highly motile cells cycling between a migratory non-resorptive phase and a fixed and resorptive phase, thus the change from one phase to the other is determined by cell morphology reorganisation. In fact, cytosqueleton studies demonstrated that in dKO OC, the migratory phase of the cells dominates over the resorptive phase and as a consequence, the dKO OC presents an altered size and a deficient resorptive function. The murine colony stimulating factor (M-CSF) has been proposed as one of the main signals inducing OC migration therefore we next explored the possibility of an exacerbated response to M-CSF as the main cause of the dKO OC behaviour. Indeed, myeloid OC progenitors presented an altered response to M-CSF associated to an altered capacity to proliferate that maintained dKO OC progenitors in a less differentiated state compare to WT OC progenitors. Interestingly, the expression of Mafb, a negative modulator of osteoclastogenesis and a transcription factor (TF) whose expression in macrophages seems to be linked to the loss of their proliferative state among differentiation, is less expressed in dKO OC progenitors than in WT cells. Moreover, Mafb expression in dKO cells was not modulated during osteoclastogenesis. This lack of Mafb regulation was shown to affect OC differentiation by impairing the normal kinetics of the main OC-driven TFs, cFOS and NFATc1. As a result, the kinetics of Acp5, CathepsinK and MMP9, all of them target genes for NFATc1 was also altered. The relationship between mafb and RXR levels suggest the modulation of Mafb expression by the RXR signalling pathway. Indeed, we have shown not only that RXR directly induce Mafb expression, but also that RXR is bound to certain regions from the proximal promoter of Mafb. Thus, although direct mutagenesis studies are needed to recognise the specific sequence of interaction, Mafb is probably a target gene for the homodimer RXRs in myeloid cells.
As a summary, here we propose that the lack of RXR implies low levels of Mafb in myeloid cells and this deficiency is determinant for the phenotype presented in dKO mice: Due to the low level of Mafb in dKO myeloid cells, these cells present an altered response to M-CSF which is associated to an altered myelopoyesis, thus, dKO OC progenitors present a less differentiate phenotype when undergo into terminal OC differentiation and osteoclastogenesis process is altered. As a result, Mafb deficiency impacts OC morphology and activity and determine the bone structure of RXR dKO mice. Finally, we provide evidence that RXR regulates the transcription of Mafb in myeloid cells independently of heterodimeric partners and possibly acting as an RXR / RXR homodimer.
The results presented here contribute to increase the knowledge of the molecular mechanisms implicated in myeliod cell differentiation and osteoclastogenesis. Characterisation of the intrinsic and extrinsic pathways that regulate HSC biology and differentiation has rapidly evolved within the last few years and different clinical methods to expand human HSC based on very new discoveries are currently being tested. More specifically, imbalances in bone remodeling result in severe perturbations in skeletal structure and function leading to conditions such as osteoporosis, osteosclerosis, or osteopetrosis. HSC yields active OCs. These cells, together with osteoblasts, are the cellular players controlling long-term bone homeostasis. Thus, the description of the RXR transcriptional mechanisms responsible for the expression of OC specific genes such as Mafb, establish a starting point to look for new pharmacological or stem cell based strategies targeting osteoporosis, diabetic osteopenia, as well as impaired bone regeneration ability. Clinical settings should evaluate the potential benefits in patient populations including diabetics, postmenopausal osteoporotic women, patients with severe metabolic alterations accompanied by increased bone fragility and impaired bone healing. These disorders are currently resulting in disabilities, reduced life expectancy and enormous health costs for our society underlying the relevance of the results presented here.
Retinoic X receptors (RXR-alpha, RXR-beta and RXR-gamma) are members of the nuclear receptor family. They appear to contribute to hematopoiesis and osteoclastogenesis via their role as heterodimeric partner of several nuclear receptors. Because RXRs can also function as homodimers, it is unclear whether the role of RXRs on hematopoiesis and osteoclastogenesis occurred through the same pathway or whether it involves the regulation of parallel pathways. Additionally, while the RXR isoform is functionally the most important RXR isotype during morphogenesis, analyses of different knock-out mice demonstrate a partial redundancy in RXR functions.
In this context, our main goal was to characterise the role of RXR in HSC function and osteoclastogenesis and the two main objectives addressed in the project were:
(1) to test the hypothesis that RXR is compensating the in vivo effects of conditional inhibition of RXR in the murine adult hematopoietic system;
(2) to clarify the contribution of RXR signalling during myelomonocytic differentiation and osteoclastogenesis processes; our hypothesis is that RXR influences myelopoiesis and osteoclastogenesis exerting direct actions as homodimer.
To address these questions, we have carried out a cross-disciplinary research project (involving gene engineered animal models, in vivo phenotyping, ex vivo cell biology techniques and in vitro molecular biology assays) that allowed us to obtain a complete overview of how RXR signaling contributes to bone physiology; from the molecular mechanisms controlled by RXR activation in cells to the impact that RXR deficiency has in rodents.
In this project we worked with a knock-out mouse model in which both RXRa and RXRß were selectively ablated in the hematopoietic system (dKO mice). These mice presented a phenotype aimed to develop osteopetrosis that was not observed in mouse lacking only the RXR isoform. Thus, in these mice, RXR at basal levels of expression is compensating for the loss of RXR and therefore, deletion of both isoforms is needed in order to study the role of RXR in hematopoiesis. The characterisation of the osteopetrotic phenotype presented by the dKO mice at the age of five months revealed a defect in bone remodelling process whose origin appear to be the presence of gigantic non-resorbing OC attached to dKO bones. According to these in vivo observations, differences in the size of dKO OC cultured in vitro were also detected. Indeed, in vitro differentiated dKO OC were bigger in size and less active when compared with WT OC. OC are highly motile cells cycling between a migratory non-resorptive phase and a fixed and resorptive phase, thus the change from one phase to the other is determined by cell morphology reorganisation. In fact, cytosqueleton studies demonstrated that in dKO OC, the migratory phase of the cells dominates over the resorptive phase and as a consequence, the dKO OC presents an altered size and a deficient resorptive function. The murine colony stimulating factor (M-CSF) has been proposed as one of the main signals inducing OC migration therefore we next explored the possibility of an exacerbated response to M-CSF as the main cause of the dKO OC behaviour. Indeed, myeloid OC progenitors presented an altered response to M-CSF associated to an altered capacity to proliferate that maintained dKO OC progenitors in a less differentiated state compare to WT OC progenitors. Interestingly, the expression of Mafb, a negative modulator of osteoclastogenesis and a transcription factor (TF) whose expression in macrophages seems to be linked to the loss of their proliferative state among differentiation, is less expressed in dKO OC progenitors than in WT cells. Moreover, Mafb expression in dKO cells was not modulated during osteoclastogenesis. This lack of Mafb regulation was shown to affect OC differentiation by impairing the normal kinetics of the main OC-driven TFs, cFOS and NFATc1. As a result, the kinetics of Acp5, CathepsinK and MMP9, all of them target genes for NFATc1 was also altered. The relationship between mafb and RXR levels suggest the modulation of Mafb expression by the RXR signalling pathway. Indeed, we have shown not only that RXR directly induce Mafb expression, but also that RXR is bound to certain regions from the proximal promoter of Mafb. Thus, although direct mutagenesis studies are needed to recognise the specific sequence of interaction, Mafb is probably a target gene for the homodimer RXRs in myeloid cells.
As a summary, here we propose that the lack of RXR implies low levels of Mafb in myeloid cells and this deficiency is determinant for the phenotype presented in dKO mice: Due to the low level of Mafb in dKO myeloid cells, these cells present an altered response to M-CSF which is associated to an altered myelopoyesis, thus, dKO OC progenitors present a less differentiate phenotype when undergo into terminal OC differentiation and osteoclastogenesis process is altered. As a result, Mafb deficiency impacts OC morphology and activity and determine the bone structure of RXR dKO mice. Finally, we provide evidence that RXR regulates the transcription of Mafb in myeloid cells independently of heterodimeric partners and possibly acting as an RXR / RXR homodimer.
The results presented here contribute to increase the knowledge of the molecular mechanisms implicated in myeliod cell differentiation and osteoclastogenesis. Characterisation of the intrinsic and extrinsic pathways that regulate HSC biology and differentiation has rapidly evolved within the last few years and different clinical methods to expand human HSC based on very new discoveries are currently being tested. More specifically, imbalances in bone remodeling result in severe perturbations in skeletal structure and function leading to conditions such as osteoporosis, osteosclerosis, or osteopetrosis. HSC yields active OCs. These cells, together with osteoblasts, are the cellular players controlling long-term bone homeostasis. Thus, the description of the RXR transcriptional mechanisms responsible for the expression of OC specific genes such as Mafb, establish a starting point to look for new pharmacological or stem cell based strategies targeting osteoporosis, diabetic osteopenia, as well as impaired bone regeneration ability. Clinical settings should evaluate the potential benefits in patient populations including diabetics, postmenopausal osteoporotic women, patients with severe metabolic alterations accompanied by increased bone fragility and impaired bone healing. These disorders are currently resulting in disabilities, reduced life expectancy and enormous health costs for our society underlying the relevance of the results presented here.
