mTORC1 plays an important role in osteoblastic regulation of B-lymphopoiesis

Skeletal osteoblasts are important regulators of B-lymphopoiesis, serving as a rich source of factors such as CXCL12 and IL-7 which are crucial for B-cell development. Recent studies from our laboratory and others have shown that deletion of Rptor, a unique component of the mTORC1 nutrient-sensing complex, early in the osteoblast lineage development results in defective bone development in mice. In this study, we now demonstrate that mTORC1 signalling in pre-osteoblasts is required for normal B-lymphocyte development in mice. Targeted deletion of Rptor in osterix-expressing pre-osteoblasts (Rptorob−/−) leads to a significant reduction in the number of B-cells in the bone marrow, peripheral blood and spleen at 4 and 12 weeks of age. Rptorob−/− mice also exhibit a significant reduction in pre-B and immature B-cells in the BM, indicative of a block in B-cell development from the pro-B to pre-B cell stage. Circulating levels of IL-7 and CXCL12 are also significantly reduced in Rptorob−/− mice. Importantly, whilst Rptor-deficient osteoblasts are unable to support HSC differentiation to B-cells in co-culture, this can be rescued by the addition of exogenous IL-7 and CXCL12. Collectively, these findings demonstrate that mTORC1 plays an important role in extrinsic osteoblastic regulation of B-cell development.


Conditional Ablation of Rptor in Osteoblast Precursors leads to Impaired B Lymphopoiesis.
Rptor, a unique and essential component of mTORC1, was ablated in osteoblast precursor cells using transgenic mice in which Cre recombinase is driven by the osterix promoter, a transcription factor expressed early in osteoblastogenesis (Osx1-GFP::Cre) 23 . These mice were mated with mice harbouring loxP sites flanking exon 6 of the Rptor gene (Rptor fl/fl ) 24 and a floxed stop R26-stop-EYFP reporter 25 to produce heterozygous (eYFP-Rptor ob +/− ) and homozygous (eYFP-Rptor ob −/− ) Rptor knockout mice 20 . In light of the distinct skeletal phenotype of Osx1-GFP::Cre mice 20,26-28 , eYFP-Osx:Cre transgenic mice (as opposed to wild type littermates) were used as controls for all experiments. Furthermore, as we have previously shown a gene dosage effect in mice with deletion of one or both alleles of Rptor in pre-osteoblasts 20 , heterozygous littermates were also included in all analyses.
To determine whether the loss of Rptor affects the ability of osteoblasts to support haematopoietic development, we analysed the frequency of mature haematopoietic lineages in the BM of heterozygous (eYFP-Rptor ob +/− ) and homozygous (eYFP-Rptor ob −/− ) knockout mice using flow cytometry. For these studies, animals were analysed at both 4 and 12 weeks of age as they represent early and later phase of long-bone growth in mice 29 . As eYFP-Rptor ob −/− mice are significantly smaller than eYFP-Osx:Cre controls at both 4 and 12 weeks of age 20 , the distribution of each lineage was calculated as a percentage of total BM cells in order to account for the reduced skeletal size and bone marrow cellularity of eYFP-Rptor ob −/− mice. At both time points, a significant reduction in B220 + B-cells was observed in the BM of eYFP-Rptor ob −/− knockout mice compared to eYFP-Osx:Cre controls (Fig. 1A,B). At 4 weeks of age, no significant difference in CD3 + T-cells was observed in the BM of eYFP-Rptor ob −/− mice, whereas CD11b + /Gr1 + myeloid cells were increased (Fig. 1A). In contrast, a significant reduction in CD3 + T-cells and a significant increase in CD11b + /Gr1 + myeloid cells was observed in 12-week old eYFP-Rptor ob −/− mice compared to controls (Fig. 1B). Notably, the B-cell defect observed in the BM of eYFP-Rptor ob −/− knockout mice was accompanied by a significant reduction in the number of B220 + B-cells in the peripheral blood circulation (Fig. 1C) and the spleen (Fig. 1D) at both 4 and 12 weeks of age. A significant reduction in B220 + B-cells in the peripheral blood circulation and spleen of heterozygous eYFP-Rptor ob +/− mice relative to controls was also observed at 12 weeks of age, suggestive of a gene dose effect (Fig. 1C,D).
Whilst the spleens of 4-week old eYFP-Rptor ob −/− mice were markedly smaller than eYFP-Osx:Cre controls, this was not statistically significant (p = 0.64) when corrected for body weight ( Fig. 2A). Intriguingly, eYFP-Rptor ob −/− spleen weights were significantly reduced compared to the heterozygous eYFP-Rptor ob +/− knockout mice, due to a modest increase in weight-adjusted spleen weight in the eYFP-Rptor ob +/− mice compared to eYFP-Osx:Cre controls ( Fig. 2A). Whilst eYFP-Rptor ob +/− and eYFP-Rptor ob −/− mice displayed reduced spleen weights compared to eYFP-Osx:Cre controls at 12 weeks of age, this was not statistically significant (p = 0.42 and p = 0.55 respectively, Fig. 2A). Within the spleen, the proliferation and differentiation of B-lymphocytes occurs in lymphoid follicles, the major component of the white pulp (Fig. 2B,C). While histological analysis revealed no difference in splenic white pulp area in eYFP-Rptor ob +/− and eYFP-Rptor ob −/− knockout mice compared to controls (Fig. 2D), the number of white pulp follicles was significantly increased in eYFP-Rptor ob −/− at 4 weeks of age (Fig. 2E). As a result, the average size of the white pulp follicles in these mice was significantly reduced in 4-week old mice (Fig. 2F). prepro-B, pro-B and pre-B cells and ultimately, immature B-cells. These immature B-cells then leave the BM for the final stages of their maturation in secondary lymphoid organs to form end-stage antibody-producing plasma cells which, upon activation by antigen-presentation, return and colonise the BM 7-12 . To investigate whether impaired B-lymphopoiesis is responsible for the reduced numbers of B220 + B-cells in eYFP-Rptor ob +/− and eYFP-Rptor ob −/− mice, the relative proportion of prepro-B, pro-B, pre-B and immature B-cells in the BM, spleen and peripheral blood circulation was assessed using CD19, CD43, IgM and B220 phenotypic markers ( Fig. 3A; Supplementary Fig. 1). As shown in Fig. 3B, the percentage of pro-B cells, pre-B cells, and immature B-cells was significantly reduced in the BM of eYFP-Rptor ob −/− mice at both 4 and 12 weeks of age. This reduction in pre-B and immature B cells was also observed in the peripheral blood circulation and the spleen of eYFP-Rptor ob −/− mice albeit at 12 weeks of age only (Fig. 3C,D). Furthermore, this reduction was also evident in heterozygous eYFP-Rptor ob +/− mice, consistent with a gene dose effect. Unlike the BM, there was no significant change in the number of pro-B cells in the peripheral blood or spleen irrespective of genotype or age (Fig. 3C,D). Interestingly, there was a significant increase in the number of prepro-B cells in the spleens of both heterozygous and homozygous mice, relative to controls, at 12 weeks of age (Fig. 3D). These data suggest that mTORC1 signalling in pre-osteoblasts has a profound effect on B-cell development.

CXCL12 and IL-7 mRNA Expression is Decreased in Rptor-Deficient
Osteoblasts. Skeletal osteoblasts are a rich source of trophic factors which regulate B-lymphopoiesis. CXCL12, an important regulator of early B-cell development, is important for prepro-B cell development 9 . In contrast, IL-7 is an important regulator of late B-cell development and is crucial for the production of pro-B and pre-B cells 30,31 . To identify the potential mechanisms by which pre-osteoblast mTORC1 signalling regulates B-lymphopoiesis, we assessed transcript levels of Cxcl12 and Il7 in eYFP + cells (ie. osteoprogenitors, mature osteoblasts and osteocytes harbouring Cre-mediated recombination) recovered from the long bones of 4-week old eYFP-Osx:Cre, eYFP-Rptor ob +/− and eYFP-Rptor ob −/− mice. Cxcl12 and IL7 mRNA levels were significantly reduced in eYFP-Rptor ob −/− osteoblasts compared to controls ( Fig. 4A: mean decrease 39.5 ± 2.2%; and Fig. 4B: mean decrease 62 ± 6.1%). In contrast, for heterozygous eYFP-Rptor ob +/− mice, transcript levels of Cxcl12 were increased and no change in Il7 transcript levels, relative to controls, was observed (Fig. 4A,B). Despite the genotype-specific differences in transcript Rptor deficient osteoblasts fail to support HSC differentiation to B-cells in vitro, but this can be rescued by exogenous Flt3L, IL-7 and CXCL12. To confirm that the abnormal B-lymphopoiesis observed in the eYFP-Rptor ob −/− mice is directly attributable to Rptor deficiency in osteoblasts, we next examined the ability of wild type and Rptor-deficient osteoblasts to support the differentiation of purified haematopoietic stem cells (LSK cells) into B-cell precursors. To achieve this, primary calvarial MSCs were isolated from neonatal Rptor fl/fl mice and infected with a tamoxifen-inducible self-deleting Cre recombinase (CreER T2 ). CreER T2 -infected cells were then treated with or without tamoxifen for 8 days to induce Rptor deletion (RapKO) or vehicle control (WT) MSCs. These WT and RapKO MSCs were then cultured under osteoinductive conditions to produce RapKO and WT osteoblasts as previously described 6 .
When BM LSK cells from wild type C57BL/6 mice were added to these osteoblast monolayers, approximately 42% of the haematopoietic cells recovered from the WT osteoblast co-cultures were B220 + after 10 days compared to only 29% of the cells recovered from RapKO osteoblast co-cultures ( Fig. 5A: mean decrease 31.7 ± 1.5%). Importantly, the addition of exogenous IL-7 and CXCL12 to these co-cultures restored the ability of RapKO osteoblasts to support B lymphopoiesis, with 49% and 51% of the haematopoietic cells recovered from WT and RapKO osteoblast co-cultures found to be B220 + , respectively (Fig. 5A).
Using CD19, CD43 and IgM phenotypic markers, the relative proportion of prepro-B, pro-B, pre-B and immature B-cells within the B220 + cells isolated from the osteoblast-LSK co-cultures was also examined. As shown in Fig. 5B, in the absence of exogenous factors, the percentage of prepro-B cells was significantly increased in RapKO osteoblast co-cultures compared to WT co-cultures (mean increase: 115.47 ± 17%), whereas the percentages of pro-B, pre-B and immature-B cells were reduced. Importantly, the addition of exogenous IL-7 and CXCL12 to these co-cultures restored the ability of RapKO osteoblasts to support LSK differentiation into pre-B and immature B-cells as evidenced by a factor-dependent normalisation of prepro-B cell numbers and a significant increase in the percentage of pro-B, pre-B and immature B cells (Fig. 5B).

Discussion
Stromal cells within the BM microenvironment, such as osteoblasts, endothelial cells, adipocytes and fibroblasts, are crucial for HSC development. Beyond its support for HSC precursors, the BM stroma also provides a specific niche for lineage-committed haematopoietic cells such as B-lymphocytes 14  with bone-specific knockout of mTOR 32 and Rptor 20-22 leading to impaired postnatal bone acquisition and conversely, bone-specific knockout of tuberous sclerosis complex 2 (TSC-2, negative regulator of mTORC1) 32 causing a progressive increase in postnatal bone acquisition. Considering the well-established role that osteoblasts play in supporting haematopoiesis and B-cell development 3,6 , we hypothesised that skeletal mTORC1 signalling plays an important role in supporting B-cell development. We also showed that eYFP + bone cells isolated from eYFP-Rptor ob −/− KO mice express significantly lower levels of Cxcl12 and Il-7 transcript compared to eYFP-Osx:Cre controls, and display a concomitant decrease in CXCL12 and IL-7 protein levels in the peripheral blood circulation. Given the well-established role of IL-7 in the differentiation of pro-B cells 30,31 , the reduced levels of IL-7 expression observed in osteoblasts isolated from eYFP-Rptor ob −/− mice was consistent with the defect in B-lymphopoiesis at the pro-B to pre-B transition in these mice. The significant reduction in CXCL12 expression however, in the absence of any defect in prepro-B cell production in the BM, was somewhat surprising and raises the possibility that other factors may compensate for the loss of CXCL12 in the early stages of B-lymphopoiesis in these mice. CXCL12 is important for the formation of distinct niches in the bone marrow that support life-long generation of B lymphocytes by acting as a chemoattractant for CXCR4-expressing progenitors 33 . It is therefore interesting to note that the level of splenic prepro-B cells was elevated in 12-week old eYFP-Rptor ob −/− mice. As these cells are typically restricted to the BM, an increase in splenic prepro-B cells could reflect a change to the B cell supportive niche that allows extravasation of progenitor cells to the spleen. Moreover, modification of bone marrow niche may also account for the reduction in T cells and increase in myeloid cells in eYFP-Rptor ob −/− mice. Whilst this manuscript was under preparation, a study by Wang and colleagues showed that the deletion of Rptor in Osx-expressing pre-osteoblasts, and not mature Ocn-expressing osteoblasts, leads to impaired B-lymphopoiesis 34 20 . Phenotypes arising from the reduction (loss of one allele of Rptor) or complete (homozygous deletion of Rptor) loss of mTORC1 function in pre-osteoblasts are complex as mTORC1 plays an important role in negative feedback loops that control signal transduction pathways. In response to insulin, mTORC1 directly and indirectly phosphorylates insulin receptor substrate 1 (IRS-1), leading to its degradation 36 . mTORC1 also activates growth factor receptor-bound protein 10 (Grb10), which inhibits IRS1 from binding to activated insulin receptor (InsR) 37 . Thus perturbations in mTORC1 function can lead to hyper-stimulation of signalling pathways which contribute to phenotype complexity.
We have previously shown that osteoblasts derived from eYFP-Rptor ob −/− knockout mice have a reduced osteogenic potential and a transcriptional profile consistent with an immature osteoblast phenotype, indicating that Rptor deletion leads to a stall in early osteogenic differentiation 20 . These cells also have reduced protein synthetic capacity, which likely inhibits their capacity to differentiate into mature osteoblasts. Thus, we postulate that mTORC1-mediated translational control contributes to IL-7 and CXCL12 production in osteoblasts and is responsible for the impaired B-lymphopoiesis observed in response to osteoblastic mTORC1 inhibition.
In addition to playing an essential role in the maintenance and differentiation of HSCs, OBs have been shown to play a role in suppressing haematological malignancies. Osteoblast numbers are significantly decreased in patients with myelodysplasia and acute myeloid leukemia 38 , and genetic depletion of osteoblasts in murine models of acute leukemia results in an increase in circulating blasts and tumor engraftment 38 . Global disruption of gene expression by deletion of Dicer1 in osteoprogenitors results in impaired osteoblastic differentiation and the development of myelodysplasia with the propensity to develop acute myeloid leukemia (AML) 39 . Similarly, constitutive activation of β-catenin signaling, specifically in osteoblast progenitors alters the differentiation potential of myeloid and lymphoid progenitors leading to the development of AML 40 . These findings demonstrate that critical changes in the signaling pathways that govern normal osteoblast function can contribute to aberrant haematopoiesis. Therefore, gaining a greater understanding of osteoblastic signaling pathways that are important for maintaining normal haematopoiesis could reveal important therapeutic targets for haematopoietic malignancies.
In summary, our findings demonstrate that mTORC1 signalling in osteoprogenitor cells is crucial to the regulation of B-lymphopoiesis by cells of the osteoblast lineage. Given the propensity for malignant disorders of the B-cell lineage (e.g. multiple myeloma) to involve the skeleton, these findings lay the foundation for future studies involving the pathophysiology of these disorders and offer novel therapeutic approaches.

Materials and Methods
Transgenic mice. All mice were bred and housed at the SA Pathology Animal Care Facility and all studies performed with Institutional Ethics approval (SA Pathology/CALHN, #102/12) and were performed in accordance with relevant guidelines and regulations. Conditional knockout mice in which Rptor, a unique and essential component of mTORC1, was disrupted in early osteoprogenitor cells were generated using Osx1-GFP::Cre mice 23 , R26eYFP mice 25 and Rptor fl/fl mice 24 as previously described 20 . In all studies, Osx1-GFP::Cre (designated eYFP-Osx:Cre) mice were used as controls. Heterozygous (Osx1-GFP::Cre-Rosa26eYFP-Rptor +/fl , designated eYFP-Rptor ob +/− ) and homozygous (Osx1-GFP::Cre-Rosa26eYFP-Rptor fl/fl , designated eYFP-Rptor ob −/− ) knockout mice were born at the expected Mendelian frequencies and at birth, no gross phenotypic differences or difference in weight or length was evident in knockout animals, relative to age-matched eYFP-Osx:Cre controls 20 . For all studies, equal numbers of male and female mice were used in each group.