Inactivation of Hes1 in Skeletal Undifferentiated Cells Increases Bone Volume
Ernesto Canalis, Emily Denker, Lauren Schilling
TL;DR
Deleting the Hes1 gene in certain bone cells increases bone volume in mice by reducing bone resorption.
Contribution
This study identifies Hes1 as a key gene in LepR+ cells that influences bone mass through reduced RANKL and osteoclast activity.
Findings
Hes1 deletion in LepR+ cells increases femoral bone volume due to higher trabecular number.
Hes1 inactivation reduces osteoclast number and bone resorption without affecting osteogenesis or adipogenesis.
Single-cell RNA sequencing showed minimal differences in cell composition or gene expression between Hes1Δ/Δ and control mice.
Abstract
Leptin receptor positive (LepR+) cells are multipotent stromal cells and a source of osteogenic and adipogenic cells. Inactivation of Notch signaling in LepR+ cells increases bone mass in mature mice, but the target gene responsible was not identified. Because in LepR+ cells the expression of the Notch target gene Hes1 prevails over that of other genes, we explored the role of the Hes1 deletion in LepR+ cells. To this end, LepR-Cre;Hes1Δ/Δ mice were compared to Hes1loxP/loxP littermates. Male and female 5-month-old LepR-Cre;Hes1Δ/Δ mice exhibited an increase in femoral bone volume/total volume due to an increase in trabecular number; vertebral (L3) and cortical bone was not affected. Bone histomorphometry demonstrated decreased osteoclast number and eroded surface, decreased osteoblast number only in male mice, and no changes in bone formation. Neither osteogenesis nor adipogenesis was…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9| Gene | Strand | Sequence | Amplicon size (bp) |
|---|---|---|---|
|
| WT Forward | 5′-CTAGGCCACAGAATTGAAAGATCT-3′ | WT = 324 |
| WT Reverse | 5′-GTAGGTGGAAATTCTAGCATCATCC-3′ | ||
| Cre Forward | 5′-GCGGTCTGGCAGTAAAAACTATC-3′ | ||
| Cre Reverse | 5′-GTGAAACAGCATTGCTGTCACTT-3′ | ||
|
| Forward | 5′-CAGCCAGTGTCAACACGACACCGGACAAAC-3′ | WT = 224 |
| Reverse | 5′-TCGCCTTCGCCTCTTCTCCATGATA-3′ | ||
|
| Forward | 5′- CAGCCAGTGTCAACACGACACCGGACAAAC-3′ | Recombined = 291 |
| Reverse | 5′-GGTGGGGCTTGAAATTCATGTAGTTTG-3′ | ||
| Reverse | 5′-CCTGAGTAAGGACAGACAAATGAAGGT-3′ |
| Gene | Strand | Sequence | GenBank accession number |
|---|---|---|---|
|
| Forward | 5′-GTGAGCCTCTTCAAGAAG-3′ |
|
| Reverse | 5′-TCCTGATACTGGTCGTAG-3′ | ||
|
| Forward | 5′-TGGTATGGGCGTCTCCACAGTAACC-3′ |
|
| Reverse | 5′-CTTGGAGAGGGCCACAAAGG-3′ | ||
|
| Forward | 5′-GACTCCGGCGCTACCTTGGGTAAG-3′ |
|
| Reverse | 5′-CCCAGCACAACTCCTCCCTA-3′ | ||
|
| Forward | 5′-AATGGAGACGGCGATAGTTCC-3′ | |
| Reverse | 5′-ACCCGAGAGTGTGGAAAGTGTG-3′ | ||
|
| Forward | 5′-CTGTCCCCAGGACGGATGACGACTCTG-3′ |
|
| Reverse | 5′-CATCGCTTGTAGGGTTCAGG-3′ | ||
|
| Forward | 5′-ACCAAAGACGGCCTCTGAGCACAGAAAGT-3′ |
|
| Reverse | 5′-ATTCTTGCCCTTCGCCTCTT-3′ | ||
|
| Forward | 5′-TTCCAGAAGACCTACAACA-3′ |
|
| Reverse | 5′-ACCCTTCTCATAGGCATT-3′ | ||
|
| Forward | 5′-TGGAGCCTAAGTTTGAGTTTGCTGT-3′ |
|
| Reverse | 5′-GGGCGGTCTCCACTGAGAAT-3′ | ||
|
| Forward | 5′-AGAACAAGGATAATGTGAAGTTCAAGGTTC-3′ |
|
| Reverse | 5′-CTGCTTCAGCTTCTCTGCCTTT-3′ | ||
|
| Forward | 5′-TATAGAATCCTGAGACTCCATGAAAAC-3′ | |
| Reverse | 5′-CCCTGAAAGGCTTGTTTCATCC-3′ | ||
|
| Forward | 5′-CAGAAAGGAAATGCAACACATGACAAC-3′ |
|
| Reverse | 5′-GCCTCTTCACACAGGGTGACATC-3′ |
| Males | Females | |||
|---|---|---|---|---|
| Control |
| Control |
| |
| Distal femur trabecular bone | ||||
| Bone volume/total volume (%) | 12.6 ± 2.0 | 16.0 ± 2.3 | 3.7 ± 1.4 | 6.3 ± 1.6 |
| Trabecular separation (µm) | 224 ± 29 | 196 ± 8 | 328 ± 15 | 276 ± 17 |
| Trabecular number (1/mm) | 4.5 ± 0.5 | 5.1 ± 0.2 | 3.1 ± 0.2 | 3.7 ± 0.2 |
| Trabecular thickness (µm) | 48 ± 7 | 47 ± 4 | 38 ± 8 | 39 ± 4 |
| Connectivity density (1/mm3) | 135 ± 43 | 200 ± 30 | 57 ± 13 | 95 ± 19 |
| Structure model index | 2.0 ± 0.3 | 1.7 ± 0.3 | 3.1 ± 0.4 | 2.6 ± 0.3 |
| Density of material (mg HA/cm3) | 947 ± 17 | 957 ± 14 | 834 ± 91 | 874 ± 83 |
| Femoral midshaft cortical bone | ||||
| Bone volume/total volume (%) | 92.6 ± 0.3 | 92.4 ± 0.4 | 92.9 ± 0.2 | 92.9 ± 0.4 |
| Porosity (%) | 7.4 ± 0.3 | 7.6 ± 0.4 | 7.1 ± 0.2 | 7.1 ± 0.4 |
| Cortical thickness (µm) | 184 ± 8 | 183 ± 7 | 197 ± 3 | 192 ± 7 |
| Total area (mm2) | 2.2 ± 0.2 | 2.2 ± 0.2 | 1.7 ± 0.1 | 1.7 ± 0.1 |
| Bone area (mm2) | 1.0 ± 0.1 | 1.0 ± 0.1 | 0.9 ± 0.1 | 0.9 ± 0.1 |
| Marrow area (mm2) | 1.2 ± 0.1 | 1.2 ± 0.1 | 0.89 ± 0.02 | 0.86 ± 0.03 |
| Periosteal perimeter (mm) | 5.2 ± 0.2 | 5.3 ± 0.2 | 4.6 ± 0.1 | 4.6 ± 0.1 |
| Endocortical perimeter (mm) | 3.8 ± 0.2 | 3.9 ± 0.2 | 3.2 ± 0.1 | 3.3 ± 0.1 |
| Density of material (mg HA/cm3) | 1230 ± 34 | 1231 ± 21 | 1243 ± 21 | 1252 ± 39 |
| pMOI (mm4) | 0.520 ± 0.09 | 0.552 ± 0.08 | 0.347 ± 0.01 | 0.342 ± 0.02 |
| Males | Females | |||
|---|---|---|---|---|
| Control |
| Control |
| |
| Osteoblasts/bone perimeter (1/mm) | 5.6 ± 1.8 | 3.1 ± 1.3 | 9.4 ± 3.0 | 6.2 ± 2.9 |
| Osteoblast surface/bone surface (%) | 7.6 ± 3.1 | 3.5 ± 1.6 | 12.8 ± 4.5 | 9.8 ± 4.4 |
| Osteocyte number | 177 ± 67 | 245 ± 43 | 105 ± 29 | 156 ± 39 |
| Osteocytes/bone area | 638 ± 138 | 563 ± 24 | 758 ± 128 | 891 ± 204 |
| Osteoid surface/bone surface (%) | 0.1 ± 0.1 | 0.1 ± 0.1 | 0.2 ± 0.2 | 0.1 ± 0.2 |
| Osteoclasts/bone perimeter (1/mm) | 2.2 ± 0.7 | 1.0 ± 0.3 | 4.5 ± 1.8 | 1.5 ± 0.5 |
| Osteoclast surface/bone surface (%) | 4.4 ± 1.6 | 2.1 ± 0.5 | 9.7 ± 3.9 | 2.9 ± 1.3 |
| Eroded surface/bone surface (%) | 1.0 ± 0.5 | 0.3 ± 0.1 | 2.3 ± 0.7 | 0.5 ± 0.2 |
| Mineral apposition rate (µm/day) | 1.1 ± 0.6 | 0.8 ± 0.1 | 1.3 ± 0.4 | 1.2 ± 0.3 |
| Mineralizing surface/bone surface (%) | 2.1 ± 1.2 | 1.5 ± 0.6 | 2.8 ± 1.0 | 2.6 ± 0.7 |
| Bone formation rate (µm3/µm2/day) | 0.02 ± 0.01 | 0.01 ± 0.01 | 0.04 ± 0.02 | 0.03 ± 0.01 |
| Sample | Cells loaded | Estimated number of cells | % recovery | Median genes per cell | Median UMI count per cell | Total detected coding genes | Number of reads from cells called from this sample |
|---|---|---|---|---|---|---|---|
| Control | 6580 | 2813 | 42.8 | 2315 | 8803 | 21 777 | 152 491 742 |
|
| 6528 | 3601 | 57.5 | 2090 | 10 128 | 22 339 | 201 970 474 |
| Total | 12 838 | 6414 | 2203 | 22 058 | 177 231 108 |
- —National Institute of Arthritis and Musculoskeletal and Skin Diseases10.13039/100000069
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsRegulation of Appetite and Obesity · MicroRNA in disease regulation · Bone Metabolism and Diseases
Notch receptors (Notch 1 to 4) are determinants of cell fate decisions and cell function that are activated following interactions with ligands of the Jagged and Delta-like families (1). These interactions lead to the exposure of the negative regulatory region, making it accessible to ADAM metalloproteases and the γ-secretase complex for proteolytic cleavage (2). Nicastrin is a component of the γ-secretase complex that, following the initial cleavage of the negative regulatory region, binds to the newly formed amino terminus and recruits it to the complex to facilitate proteolysis (3). As a consequence of the Notch ligand interactions, the Notch intracellular domain is released into the cell and translocates to the nucleus, where it binds recombination signal-binding protein for Ig of k (RBPJk and mastermind-like (MAML) to induce the transcription of target genes (2, 4-6). Genes induced by this canonical pathway include members of the hairy and enhancer of split (Hes) and Hes-related with YRPW motif (Hey) families (7-9). Although there is a degree of overlap in the function of Notch receptors, each receptor plays a unique role in physiology and disease (10, 11). This is related to structural differences and specific cellular patterns of expression. Notch1, 2, 3 and low levels of Notch4 mRNA are detected in skeletal cells, where the prevalent ligand is Jag1 (10).
Leptin receptor positive (LepR^+^) cells have emerged as multipotent stromal cells and a source of osteogenic and adipogenic cells (12, 13). LepR^+^ cells are closely related to CXCL12^+^ cells and represent 0.3% of bone marrow cells and 85% to 95% of the fibroblast colony forming unit cells in the bone marrow (13-17). The adipogenic cells are related to perisinusoidal CXCL12^+^ or Adipo-CAR cells and to the recently discovered marrow adipogenic lineage precursors (18). The osteogenic cells are osteolectin-positive cells that are periarteriolar and colocalize with type I collagen-expressing cells (19, 20). Single-cell RNA sequencing (scRNA-seq) of bone marrow stroma has demonstrated not only coexpression of LepR^+^ and CXCL12^+^ cells but also an overlap in LepR^+^ with Adipoq and Prxx1 gene expression (13, 21).
Loss of Notch signaling following the deletion of Ncstn (nicastrin) in LepR^+^ cells was found to increase bone mass in mature/aging mice, but the Notch target gene responsible for the phenotype was not established (22). Whereas the products of the Hey family of genes could mediate selected actions of Notch signaling, their expression in skeletal cells, including LepR^+^ cells, is quite low, and the misexpression of Hey1, 2, and l results in minimal skeletal phenotypes (22-25). HES1 is the prevalent Notch target gene in LepR^+^ cells and has been shown to mediate the effects of Notch signaling in cells of the myeloid lineage and osteoclasts (22, 26). Consequently, HES1 seems to be the most promising canonical target potentially responsible for the effects of Notch in bone and LepR^+^ cells.
In the present study, we explored the role of HES1 in LepR^+^ cells. To this end, LepR-Cre mice were crossed with conditional Hes1 mice, where sequences coding for HES1 are flanked by loxP sites and their phenotype examined at 5 months of age, since the expression of LepR in bone increases sharply with age (13). The phenotype was established by determining femoral bone microarchitecture using microcomputed tomography (µCT) and femoral bone histomorphometry. Possible mechanisms involved were explored in cultures of bone marrow stromal cells from control and experimental mice and by transcriptomic analysis of femurs from Hes1 inactivated mice at a single-cell resolution.
Materials and Methods
Mouse Lines
Hes1^loxP/loxP^ conditional mice, where loxP sites flank exons 2 and 4, were obtained from R. Kageyama and were backcrossed into a C57BL/6 background (27). To inactivate Hes1, Hes1^loxP/loxP^ mice were crossed with LepR-Cre (Lepr^tm2(Cre)Rck/J^; Jackson 0008320) mice to create LepR-Cre^+/−^; Hes1^loxP/loxP^ mice, and these were bred with Hes1^loxP/loxP^ mice to obtain male and female experimental LepR-Cre;Hes1^Δ/Δ^ and control Hes1^loxP/loxP^ sex-matched littermates. Experimental Hes1^Δ/Δ^ and control Hes1^loxP/loxP^ mice were compared at 5 months of age, because the contribution of LepR^+^ cells to bone increases with age and is low in the first few months of life (13). Genotyping of LepR-Cre and Hes1 conditional mice was carried out by PCR in tail DNA extracts (Table 1). Deletion of the loxP-flanked sequences by Cre-dependent recombination was documented by PCR in DNA from tibiae using specific primers (Table 1), and the downregulation of Hes1 mRNA was documented by quantitative RT-PCR in primary cell cultures from the control and experimental mouse lines. Studies were approved by the Institutional Animal Care and Use Committee of UConn Health.
µCT
Microarchitecture was determined in femurs from control and experimental mice using a μCT instrument (Scanco μCT 40, Scanco Medical AG, Bassersdorf, Switzerland), calibrated periodically with a phantom obtained from the manufacturer (28, 29). Scanning of femurs and vertebral lumbar (L3) from control and experimental mice was performed in 70% ethanol at an energy level of 55 peak kilovoltage, intensity of 145 μA, and integration time of 200 ms as reported (30). One hundred slices at midshaft or 160 slices at the distal metaphysis were acquired from femurs, and 400 to 450 slices from L3 at an isotropic voxel size of 216 μm^3^ and a thickness of 6 μm were chosen for analysis. Images of cancellous bone architecture were evaluated starting ∼1.0 mm proximal from the femoral condyles. For L3 vertebral, the entire vertebral body was analyzed. To define the region for analysis, contours were drawn manually every 10 slices, a few voxels away from the endocortical boundary, and the remaining slice contours were iterated automatically. Bone volume fraction [bone volume/total volume (BV/TV)], trabecular thickness, trabecular number, connectivity density, structure model index, and material density were measured in trabecular regions using a Gaussian filter (σ = 0.5) and defined thresholds (28, 29). A threshold of 260 permil equivalent to 377.3 mg of hydroxyapatite (HA)/cm^3^ was used for femoral cancellous bone, and a threshold of 290 permil equivalent to 465 mg of HA/cm^2^ was used for L3. For cortical femoral bone, analysis contours were iterated across 100 slices along the cortical shell of the femoral midshaft, excluding the marrow cavity. Analysis of BV/TV, porosity, cortical thickness, total cross-sectional and cortical bone area, segmented bone area, periosteal and endosteal perimeter, material density, and polar moment of inertia were conducted using a Gaussian filter (σ = 0.8, support = 1) with a threshold of 400 permil equivalent to 682.9 mg of HA/cm^3^.
Bone Histomorphometry
Bone histomorphometry was conducted in control and experimental mice injected with calcein 20 mg/kg and demeclocycline 50 mg/kg at a 5- to 7-day interval and sacrificed 24 to 48 hours following the administration of demeclocycline. Dissected femurs were fixed in 70% ethanol, embedded in methyl methacrylate, sectioned along the sagittal plane at a thickness of 5 µm on a Microm microtome (Richards-Allan Scientific, Kalamazoo, MI), and stained with 0.1% toluidine blue. Static and dynamic parameters of bone morphometry were measured in a defined area between 0.35 mm and 2.16 mm from the growth plate at a magnification of 100× using an OsteoMeasure morphometry system (Osteometrics, Atlanta, GA). Stained sections were used to draw bone tissue contours and to measure osteoid and eroded surface, as well as to count osteoblast, osteocyte, and osteoclast number. Mineralizing surface per bone surface and mineral apposition rate were measured on unstained sections visualized under UV light and a triple diamidino-2-phenylindole/fluorescein/Texas red set long pass filter, and bone formation rate was calculated (31).
Cross-linked C-telopeptide of Type I Collagen
Cross-linked C-telopeptide of type I collagen (CTX) was measured in serum samples following an overnight fast using an ELISA EEL219 kit from Thermo Fisher Scientific (Invitrogen, Carlsbad, CA; RRID: AB_3720130).
Bone Marrow Stromal Cell Cultures
Bone marrow stromal cells were obtained from femurs and tibiae of 5-month-old control and experimental mice. Bones were dissected aseptically, the epiphysis was removed, and bone marrow cells were recovered by flushing in α-minimum essential medium (α-MEM; Life Technologies, Thermo Fisher, Grand Island, NY). Following removal of tissue debris and cellular aggregates by filtration, cells were seeded at a density of 1.25 × 10^6^ cells/cm^2^ in α-MEM containing 15% heat-inactivated fetal bovine serum (Atlanta Biologicals, Norcross, GA) and grown in a humidified 5% CO_2_ incubator at 37 °C. Cells in suspension were removed by replacing culture medium 48 hours after seeding, and adherent cells were considered bone marrow stromal cells (32). To induce osteoblast cell differentiation, cells were grown to confluence, trypsinized, and plated in α-MEM supplemented with 10% fetal bovine serum, 100 µg/mL ascorbic acid, and 5 mM β-glycerophosphate (both from Sigma-Aldrich, St. Louis, MO). To induce adipogenesis, cells were grown to confluence, trypsinized, and seeded in α-MEM containing an adipogenic cocktail consisting of dexamethasone 1 µM, insulin 100 nM, and 3-isobutyl-1-methylxanthine 0.5 mM (Sigma-Aldrich) and cultured for 14 days in the absence of β-glycerophosphate (33).
Quantitative RT-PCR
Total RNA was extracted from cells with the RNeasy kit (Qiagen, Valencia, CA), in accordance with the manufacturer's instructions, as previously reported (34-37). Equal amounts of RNA were reverse transcribed using the iScript RT-PCR kit (Bio-Rad, Carlsbad, CA) and amplified in the presence of specific primers (Integrated DNA Technologies, Coralville, IA) (Table 2) with SsoAdvanced Universal SYBR Green Supermix (Bio-Rad) at 60 °C for 35 cycles. Transcript copy number was estimated by comparison with a serial dilution of cDNA for Alpl, Bglap, Bsp1, Cfd, Hes1, Pparg (American Type Tissue Culture Collection, Manassas, VA), Tnfsf11 (Source BioScience, Nottingham, UK), and Tnfrsf11b (Thermo Scientific, Waltham, MA). Amplification reactions were conducted in a CFX96 quantitative RT-PCR detection system (Bio-Rad), and fluorescence was monitored during every PCR cycle at the annealing step. Data are expressed as copy number corrected for Rpl38 (from American Type Tissue Culture Collection) except for Adipoq and Plin1, which are expressed as relative expression values corrected for Rpl38 (38).
scRNA-seq and Computational Analysis
scRNA-seq was conducted in femoral bone samples from 5-month-old male LepR-Cre;Hes1^Δ/Δ^ and sex-matched Hes1^loxP/loxP^ controls. Intact femoral bones were excised, the epiphyseal ends removed, and the marrow flushed and discarded; bones were crushed and cells dispersed by digesting bones with collagenase, dispase, and DNase (39, 40). Dispersed cells were recovered and incubated with ACK lysis buffer to remove red blood cells followed by immune and erythroid cell depletion by double-negative (CD45^−^; TER119^−^) selection by fluorescence-activated cell sorting. Following sorting, cells were suspended and live cells counted. Transcriptomes were analyzed on a cell-by-cell basis through the use of microfluidic partitioning to capture single cells, and barcoded libraries were prepared using a ChromiumX instrument and a Chromium Single Cell GEM-X Universal 3′ gene expression v4 4-plex library and a GEM-X OCM3′ Chip v4 4-plex kit (10× Genomics, Pleasanton, CA). cDNA amplification and library generation were performed according to the manufacturer's protocol. cDNAs from a single cell had the same barcode, allowing the sequencing reads to be mapped back to the cell of origin (41). Libraries were validated for length and adapter dimer removal using the Agilent TapeStation D1000 High Sensitivity assay, quantified, and sequenced as 1 sample pool on an AVITI Platform using 2 × 7.5 high-output flow cell (AmpSeq, Gaithersburg, MD). A target read depth of 50 000 transcriptome reads per cell was achieved in each sample.
FASTQ files generated using reference genome mouse (GRCM39) 2024-A were analyzed using Cell Ranger Multi 9.0.1 (10× Genomics) for sample demultiplexing, barcode processing, and transcript counting and output HTMLs for sample visualization and data quality assessment. Output files were used for secondary analyses, such as dimensionality reduction, cell clustering, and differential gene expression using the Seurat R Package (v4.3) (42, 43). Cells expressing fewer than 250 transcripts, more than 10% mitochondrial RNA, and >10 000 genes/cell were excluded manually, and the data set was normalized by employing the NormalizeData function in Seurat. Highly variable functions were identified, and nonlinear dimensional reduction was performed by uniform manifold approximation and projection (UMAP) (44). Doublets were identified and excluded, and cell clusters were then identified by the FindClusters function in Seurat. Trajectory analysis was constructed by analyzing the Seurat object in Monocle 3 (45). A Monocle data set was constructed, and the learn principal graph from the reduced dimensional space was applied using learn_graph. Cells were organized along their trajectory using Monocle plot cells and order_cells; marker genes for each cluster were identified using Seurat's FindAllMarkers function. Ingenuity pathway analysis (IPA; Qiagen, Germantown, MD) of differentially expressed genes was conducted on select clusters, and genes were prioritized based on ontology and relevance to transcriptional control and signaling pathways (46). IPA was performed to analyze canonical pathways under the Gene and Chemicals category.
Statistics
Data are expressed as individual sample values and means ± SD. Data represent biological replicates except for cell cultures, which represent technical replicates. Statistical differences were determined by an unpaired Student's t-test for pairwise comparisons.
Results
General Appearance of LepR-Cre;Hes1Δ/Δ Mice
To study the inactivation of Hes1 in Lepr^+^ cells, LepR-Cre^+/−^;Hes1^loxP/loxP^ mice were crossed with Hes1^loxP/loxP^ mice to generate LepR-Cre^+^;Hes1^Δ/Δ^ experimental and Hes1^loxP/loxP^ littermate controls. The general appearance of LepR-Cre;Hes1^Δ/Δ^ mice was comparable to that of Hes1^loxP/loxP^ controls, and the weight and femoral length of male and female mice were not different from control mice, except for a slight and not significant decrease in weight in female mice (Fig. 1). Cre-dependent recombination was detected in tibiae from LepR-Cre;Hes1^Δ/Δ^ but not in control mice.
*General appearance, weight, and femoral length are not affected in Lepr-Cre;Hes1Δ/Δ mice. Weight, femoral length, and identification of 5-month-old LepR-Cre;Hes1Δ/Δ (black bars) and sex-matched Hes1loxP/loxP controls (white bars); males n = 6 control and n = 5 Hes1Δ/Δ; females n = 5 control and n = 9 Hes1Δ/Δ. DNA was obtained from tibiae of Hes1Δ/Δ and control mice in the upper panel for genotyping and in the lower panel to document Hes1loxP (342 bp) and Cre-mediated recombination of loxP flanked sequences in Hes1Δ/Δ (291 bp). Bars represent means ± SD, and individual values are shown as open circles. Significantly different between LepR-Cre;Hes1Δ/Δ and control by unpaired t-test, P < .05.
Skeletal Microarchitecture and Bone Histomorphometry of LepR-Cre;Hes1Δ/Δ Mice
Femoral microarchitecture, determined by µCT, of 5-month-old mice confirmed a substantially higher trabecular BV/TV in male than in female mice (29, 32). Femoral microarchitecture of LepR-Cre;Hes1^Δ/Δ^ mice revealed an ∼30% increase in trabecular BV/TV in male mice and a 70% increase in BV/TV in female mice. This increase was associated with a higher number of trabeculae but not with an increase in trabecular thickness; the structure model index was decreased revealing a tendency toward flat-like trabeculae (Fig. 2, Table 3). In contrast to the changes in cancellous bone, cortical bone architecture did not reveal significant differences between male and female LepR-Cre;Hes1^Δ/Δ^ mice and sex-matched littermate controls, indicating that the phenotype was limited to cancellous femoral bone. µCT analysis of L3 vertebrae of male control (Hes1^loxP/loxP^) mice revealed a (mean ± SD; n = 7) BV/TV of 24.2 ± 2.5% and in Lepr-Cre;Hes1^Δ/Δ^ (n = 5) of 26.2 ± 2.1 (not significant). µCT analysis of L3 in female mice revealed a BV/TV of 19.6 ± 0.8 in control and 20.4 ± 0.4 (both n = 7; not significant) in LepR-Cre;Hes1^Δ/Δ^ mice.
*(A) Representative µCT of cancellous bone from femurs of 5-month-old male and female Hes1loxP/loxP control and LepR-Cre;Hes1Δ/Δ mice, demonstrating increased bone volume in Hes1-deleted mice. Bar in the right corner of the image represents 100 µm. (B) Cancellous bone volume/total volume (%), trabecular number (1/mm), connectivity density (1/mm3), and structure model index in control (white bars) and Hes1Δ/Δ (black bars) femurs. Bars represent means ± SD, and individual values are shown as open circles; n = 5 to 9 biological replicates. Significantly different between LepR-Cre;Hes1Δ/Δ and control by unpaired t test, P < .05. A full data set is shown in Table 3.Abbreviations: µCT, microcomputed tomography.
To gain an understanding of the cause behind the cancellous bone phenotype, bone histomorphometric analysis was performed and demonstrated that there were no changes in osteocyte number or osteoid surface in either male or female LepR-Cre;Hes1^Δ/Δ^ mice (Table 4). Osteoclast number/perimeter was decreased by 55% to 65% and eroded surface by 70% to 80% in LepR-Cre;Hes1^Δ/Δ^ mice, of both sexes, indicative of decreased bone resorption (Fig. 3). Osteoblast number/perimeter and osteoblast surface/bone surface were decreased in male but not in female LepR-Cre;Hes1^Δ/Δ^, whereas bone formation parameters, including mineral apposition rate and mineralizing surface, were not affected by the Hes1 deletion in either sex. Serum CTX levels in male mice were (means ± SD; n = 5-7) 18.2 ± 9.9 ng/mL in control and 24.1 ± 7.0 ng/mL in LepR-Cre;Hes1^Δ/Δ^ mice (not significant); in female mice serum CTX levels were (n = 3-4) 16.8 ± 1.5 ng/mL in control and 21.2 ± 4.9 ng/mL (not significant) in LepR-Cre;Hes1^Δ/Δ^ mice.
*(A) Representative histomorphometry of cancellous bone from femurs of 5-month-old male and female Hes1loxP/loxP control and LepR-Cre;Hes1Δ/Δ mice, demonstrating osteoclasts and areas of eroded surface (arrows) in control but not in Hes1-deleted mice. (B) Osteoblasts/bone perimeter (1/mm), osteoclasts/bone perimeter (1/mm3), eroded surface/bone surface, and bone formation rate (µm3/µm2/day) in control (white bars) and Hes1Δ/Δ (black bars) femurs. Bars represent means ± SD, and individual values are shown as open circles; n = 4 to 7 biological replicates. Significantly different between LepR-Cre;Hes1Δ/Δ and Hes1loxP/loxP control by unpaired t test, P < .05. A full data set is shown in Table 4.
Bone Marrow Stromal Cell Cultures from LepR-Cre;Hes1Δ/Δ Mice
Bone marrow stromal cells from LepR-Cre;Hes1^Δ/Δ^ mice and littermate controls cultured under osteogenic conditions for 21 days revealed a significant decrease in Hes1 mRNA expression in Hes1^Δ/Δ^ cultures but no substantial differences in the expression of Alpl, Bglap, or Bsp1, except for a minimal and transient decrease in Alpl, demonstrating no alterations in osteogenesis by the Hes1 inactivation (Fig. 4). There was a significant decrease in Tnfsf11 (encoding receptor activator of nuclear factor k ligand, RANKL) mRNA levels and modest and not significant changes in Tnfrsf11b (encoding osteoprotegerin) mRNA in LepR-Cre;Hes1^Δ/Δ^ cells compared to control cells.
*Deletion of Hes1 in LepR+ cells has no effect on osteoblastogenesis but suppresses Tnfsf11 (RANKL). Bone marrow stromal cells harvested from 5-month-old LepR-Cre;Hes1Δ/Δ (black bars) and Hes1loxP/loxP control littermate mice (white bars) cultured for 21 days under osteogenic conditions following confluence (day 0). Total RNA was extracted and gene expression determined by quantitative RT-PCR. Data are expressed as Hes1, Alpl, Bglap, Bsp1, Tnfsf11, and Tnfrsf11b copy number corrected for Rpl38. Bars represent means ± SD, and individual values are shown as open circles; n = 4 technical replicates. Significantly different between LepR-Cre;Hes1Δ/Δ and control by unpaired t test, P < .05.
Bone marrow stromal LepR-Cre;Hes1^Δ/Δ^ and control cells cultured under adipogenic conditions displayed a substantial suppression of Hes1 mRNA in Hes1^Δ/Δ^ cells and a slight increase in Adipoq transcripts (Fig. 5). There were no changes in Cfd (encoding adipsin), Plin1 (perilipin), Pparg, or Dlkl (Pref1) expression in LepR-Cre;Hes1^Δ/Δ^ cells compared to control cells, demonstrating no effect on adipogenesis by the Hes1 inactivation. Tnfsf11 mRNA was not different between control and Hes1^Δ/Δ^ cells.
*Deletion of Hes1 in LepR+ cells has no effect on adipogenesis. Bone marrow stromal cells harvested from 5-month-old LepR-Cre;Hes1Δ/Δ (black bars) and Hes1loxP/loxP control littermate mice (white bars) cultured for 14 days under adipogenic conditions following confluence (day 0). Total RNA was extracted and gene expression determined by quantitative RT-PCR. Data are expressed at Hes1, Cfd (adipsin), Pparg, Tnfsf11, and Dlk1 copy number corrected for Rpl38 and as relative values for Adipoq and Plin1 corrected for Rpl38. Bars represent means ± SD, and individual values are shown as open circles; n = 4 technical replicates. Significantly different between LepR-Cre;Hes1Δ/Δ and control by unpaired t-test, P < .05.
scRNA-seq of Femurs from LepR-Cre;Hes1Δ/Δ Cells
Cells from the diaphysis and metaphysis of femurs from 5-month-old LepR-Cre;Hes1^Δ/Δ^ and control male Hes^loxP/loxP^ mice were obtained by enzymatic digestion following the removal of the epiphyses and bone marrow and negative selection of hematopoietic CD45^+^ and TER119^+^ cells. Approximately 12 838 cells from LepR-Cre;Hes1^Δ/Δ^ and control mice equally distributed among experimental and control groups were partitioned on a ChromiumX instrument using a 3′ library V4.4 plex kit (10× Genomics). An estimated 6414 cells or about 50% of cells were recovered, and 22 058 coding genes were detected (Table 5). Normalization excluded cells with >10% mitochondrial RNA, cells with <250 transcripts, and >10 000 genes/cell, as well as doublet filtering, reducing the number of cells analyzed to 2598 control cells and 3298 cells from LepR-Cre;Hes1^Δ/Δ^ femurs.
Cluster identification was conducted by the FindClusters function in Seurat, and analysis of pooled cells from control and Hes1^Δ/Δ^ femurs using UMAP for nonlinear dimensionality reduction defined 23 different cell clusters (Fig. 6). Analysis of the gene composition of the various clusters was used to verify the identity of the clusters, which were comprised of hematological cells, including myeloid and B cells and neutrophils, endothelial cells, and osteoblasts. LepR was detected solely in the osteoblast cluster. Notch1 and Notch2 were the prevalent Notch receptors and Hes1 the prevalent canonical target gene; these were expressed in multiple cell clusters including the osteoblast cluster. Trajectory analysis revealed a close association among the clusters within each cell category but not across cell classes (Fig. 6).
UMAP for dimensional reduction and trajectory analysis of scRNA-seq data of the diaphysis and metaphysis from femoral bone of Hes1loxP/loxP control and LepR-Cre;Hes1Δ/Δ mice reveals 23 cell clusters. (A) UMAP visualization of 23 cell clusters of pooled normalized data from control and LepR-Cre;Hes1Δ/Δ femoral bones. (B) Trajectory analysis of the 23 cellular clusters identified was performed using Monocle 3 in pooled data from control and LepR-Cre;Hes1Δ/Δ femoral bones. (C) Dot plot displaying the expression of genes associated with the cellular clusters. Red denotes higher and blue denotes lower than average expression, and the size of the circle represents the percentage of cells expressing each gene.Abbreviations: scRNA-seq, single-cell RNA sequencing; UMAP, uniform manifold approximation and projection.
Independent clustering analysis of transcriptome profiles using UMAP for nonlinear dimensionality reduction did not reveal substantial differences in the distribution of clusters or gene expression between cells from control and LepR-Cre;Hes1^Δ/Δ^ femurs (Fig. 7). Trajectory finding was constructed with Monocle 3 and used to predict the differentiation trajectory among cell clusters from control and Hes1^Δ/Δ^ femurs. Trajectory finding did not demonstrate substantial differences between clusters from control and LepR-Cre;Hes1^Δ/Δ^ cells (Fig. 8). IPA of the osteoblast cluster did not reveal differences in the pathways affected between control and Hes1^Δ/Δ^ cells (Fig. 9). In accordance with the absence of an effect by the Hes1 inactivation in either osteogenesis or adipogenesis, scRNA-seq data did not demonstrate obvious differences in the transcriptome profile of femurs from LepR-Cre;Hes1^Δ/Δ^ and control mice.
UMAP for dimensional reduction reveals similar cellular cluster distribution in the diaphysis and metaphysis of femurs from Hes1loxP/loxP control or LepR-Cre;Hes1Δ/Δ mice. (A) UMAP visualization of 23 cell clusters of normalized data from control and LepR-Cre;Hes1Δ/Δ femurs. (B) Bar graphs demonstrate cell number present in each individual cluster from control and LepR-Cre;Hes1Δ/Δ femurs. (C) Dot plot displaying the expression of genes associated with the various clusters in LepR-Cre;Hes1Δ/Δ or control femoral bones. Red denotes higher and blue denotes lower than average expression, and the size of the circle represents the percentage of cells expressing each gene.Abbreviation: UMAP, uniform manifold approximation and projection.
Trajectory analysis of cell clusters in the diaphysis and metaphysis of femurs reveals modest differences between Hes1loxP/loxP control and LepR-Cre;Hes1Δ/Δ femurs. Trajectory analysis of 23 cell clusters of normalized data from femurs from LepR-Cre;Hes1Δ/Δ (right panel) and control (left panel) mice.
IPA of the osteoblast cluster reveals modest differences between Hes1loxP/loxP control and LepR-Cre;Hes1Δ/Δ femurs. IPA of canonical pathways under the genes and chemical category was performed in the osteoblast cell cluster of control and LepR-Cre;Hes1Δ/Δ femurs. Upper panels show the 25 most activated (orange) or suppressed (blue) pathways in the osteoblast cluster from control and Hes1Δ/Δ cells. Lower panels show a graphical summary of the most activated (orange) or inactivated (blue) pathways. Solid lines lead to activation or inhibition, dotted lines represent inferred relationships, dashed lines represent indirect interactions, and solid lines represent direct interactions.Abbreviation: IPA, ingenuity pathway analysis.
Discussion
Previous work demonstrated that the inactivation of Ncstn in LepR^+^ cells and consequent loss of Notch signaling results in increased bone mass attributed to an increase in osteogenesis and osteoblast function (22). In the present study, we explore the possible Notch target gene responsible for the effect by deleting Hes1 in LepR^+^ cells. Hes1 was considered the most likely canonical Notch target gene responsible for the effect of Notch signaling due to its prevalence in skeletal cells, including in LepR^+^ cells, in relation to members of the Hey family of genes, as well as demonstrating a greater role of HES1 in skeletal homeostasis than HEY1, HEY2, and HEYL (22-25). scRNA-seq of femoral bone extracts confirmed substantially greater expression of Hes1 than of Hey1, Hey2, and Heyl and revealed that the prevalent Notch receptors in femoral bone are Notch1 and Notch2. It also verified that LepR is primarily expressed in cells of the osteogenic lineage. To ensure that LepR is not expressed by osteoclasts, we reanalyzed published scRNA-seq data from bone marrow-derived macrophages as they differentiate into osteoclasts (47). LepR was not detected in any of the clusters, including osteoclast progenitors, precursors, or mature cells, verifying no expression of LepR in this lineage.
The conditional inactivation of Hes1 in LepR^+^ cells resulted in an increase in trabecular bone volume in cancellous femoral bone but did not cause a cancellous bone phenotype at vertebral (L3) sites or a cortical bone phenotype. This is consistent with the predominant expression of LepR in metaphyseal trabecular bone and not in other skeletal structures including cortical bone (13, 40). The bone microarchitectural phenotype of the Hes1 deletion in LepR^+^ cells is similar but not identical to the 1 observed following the deletion of Ncstn, resulting in decreased Notch signal activation (22). BV/TV and trabecular number were increased following both the Hes1 and Ncstn inactivation, although only the latter resulted in an increase in trabecular thickness. However, the mechanisms responsible for the phenotype observed in the context of the Hes1 inactivation appear to be different from those observed with the Notch signal downregulation since neither osteoblastogenesis in vitro nor bone formation rates in vivo were enhanced by the Hes1 deletion, as reported for the Ncstn inactivation. Indeed, a decrease in osteoblast number was observed following the Hes1 inactivation, although this was significant only in male Hes1^Δ/Δ^ mice. Therefore, a decrease in osteoclast number and bone resorption seems to be accountable for the Hes1^Δ/Δ^ phenotype. This is further supported by the higher trabecular number in Hes1^Δ/Δ^ than in control mice. Since the LepR is not expressed in the myeloid lineage, the decreased bone resorption is best explained by an inhibition of RANKL in osteogenic cells. The decreased number of osteoblasts in male Hes1^Δ/Δ^ mice may be a consequence of a lower bone remodeling state. Despite the inhibition in bone resorption as determined by bone histomorphometry, serum levels of CTX were not changed in LepR-Cre;Hes1^Δ/Δ^ mice. This is possibly due to the sensitivity of the serum CTX assay since serum CTX levels correlate significantly but only modestly with bone histomorphometric findings (48).
The discrepancy in the Ncst and Hes1 deletion phenotypes indicates that the effects of Notch in LepR^+^ cells are potentially mediated by alternate target genes. Although the present work does not exclude Hey1, 2, or l as possible candidates, their low level of expression and previous work showing modest phenotypes following their misexpression makes this possibility unlikely. Therefore, it is plausible that the effects of Notch in LepR^+^ cells involve alternate genes, regulated by either Notch canonical or noncanonical pathways (49, 50). Since the age of the animals studied was different between the Ncst (12 months) and the Hes1 (5 months) deletion, it is possible that this was a contributing factor to the differences observed.
Previously, we demonstrated that the Hes1 inactivation in Prrx1-expressing cells results in an increase in bone volume, but no changes in osteoblast function were found (51). The phenotype was attributed to an embryonic developmental effect since Prrx1 is expressed in the limb bud at day E9.5 (52). It is noteworthy that the expression of Prrx1 can overlap with the expression of LepR, and gene deletion under the control of either LepR or Prrx1 can result in analogous phenotypes (40). Inactivation of Hes1 in mature Bglap expressing osteoblasts caused an increase in BV/TV, and osteoclast number was temporarily decreased (51). Deletion of Hes1 in LepR^+^ cells replicates the decreased bone resorption observed secondary to an inhibition of Tnfsf11 (RANKL) in cells of the osteoblast lineage. The effect is in agreement with the dual role of NOTCH2 signaling on bone resorption, which is secondary to an induction of RANKL as well as to direct effects in cells of the myeloid lineage (26). The Hes1 inactivation did not result in alterations in adipogenesis, although HES1 has been shown to influence the adipogenic program in vitro (53). This may have been due to substantial differences in the experimental design and in the models used. It is of interest that Tnfsf11 (RANKL) expression was not modified in adipogenic cells by HES1, since recent work has demonstrated that adipogenic sources of RANKL can contribute to the regulation of bone resorption and remodeling (54, 55)
To explore the molecular processes occurring under the influence of HES1, we performed scRNA-seq analysis of femurs from LepR-Cre;Hes1^Δ/Δ^ and control littermates. The use of scRNA-seq is particularly important in the analysis of a complex tissue-like bone, allowing discernment of the transcriptome profile of various cell populations present in this tissue. It also allowed us to determine whether there were HES1-dependent changes in cell composition and transcriptomic profile.
The present study extends recent observations on scRNA-seq analysis of cells of the femoral bone and confirms the cellular heterogeneity of this tissue found in humans and mice (22, 56). We identified 23 cell clusters, and pseudotime trajectory analysis revealed an association between clusters expressing gene markers identified in endothelial, hematological, and myeloid cells; B cells; and osteoblasts. Each cluster category was fairly distinct, and trajectory findings did not reveal an association among the various independent clusters. The Hes1 inactivation resulted in modest changes on cluster allocation and trajectory finding. This is in agreement with a predominant effect of HES1 on bone resorption through an inhibition of RANKL.
Although there was a distinct cluster expressing genes associated with osteoblasts, the cluster was constituted by a modest number of cells and was not affected by the Hes1 inactivation. IPA did not reveal substantial changes in signaling pathways between control and Hes1-deleted osteoblasts, confirming the absence of an osteogenic effect in primary cultures from LepR-Cre;Hes1^Δ/Δ^ mice. The results are in agreement with recent work from our laboratory in induced pluripotent human NCRM5 stem cells, where the HES1 deletion had minimal effects on osteoblastogenesis but HES1 was required for osteoclastogenesis (57).
There are limitations to the present work. The deletion of Hes1 in LepR^+^ cells outside the skeleton was not excluded and could have influenced the phenotype observed. This is possible since the LepR is expressed in the hypothalamus and in other tissues, including the pancreas, liver, kidney, lung, and stomach (7, 58). It is of interest that the action of leptin through the hypothalamic LepR regulates bone remodeling (59). Another limitation is the fact that changes in osteoblast and adipocyte differentiation were determined at the level of mRNA expression and not verified by the formation of mineralized nodules or mature adipocytes. This was due to the limited cellular yield of the bone marrow stromal cell preparations, possibly due to the age (5-month-old) of the animals.
In conclusion, inactivation of Hes1 in LepR^+^ cells increases bone mass by inhibiting RANKL expression and bone resorption but has no direct effects on osteoblastogenesis in vitro or bone formation in vivo. These observations indicate that HES1 regulates bone resorption not only through direct effects in cells of the myeloid lineage but also indirectly by inducing RANKL in osteogenic cells.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Siebel C, Lendahl U. Notch signaling in development, tissue homeostasis, and disease. Physiol Rev. 2017;97(4):1235‐1294.28794168 10.1152/physrev.00005.2017 · doi ↗ · pubmed ↗
- 2Kovall RA . More complicated than it looks: assembly of Notch pathway transcription complexes. Oncogene. 2008;27(38):5099‐5109.18758478 10.1038/onc.2008.223 · doi ↗ · pubmed ↗
- 3Shah S, Lee SF, Tabuchi K, et al Nicastrin functions as a gamma-secretase-substrate receptor. Cell. 2005;122(3):435‐447.16096062 10.1016/j.cell.2005.05.022 · doi ↗ · pubmed ↗
- 4Nam Y, Sliz P, Song L, Aster JC, Blacklow SC. Structural basis for cooperativity in recruitment of MAML coactivators to Notch transcription complexes. Cell. 2006;124(5):973‐983.16530044 10.1016/j.cell.2005.12.037 · doi ↗ · pubmed ↗
- 5Schroeter EH, Kisslinger JA, Kopan R. Notch-1 signalling requires ligand-induced proteolytic release of intracellular domain. Nature. 1998;393(6683):382‐386.9620803 10.1038/30756 · doi ↗ · pubmed ↗
- 6Wilson JJ, Kovall RA. Crystal structure of the CSL-Notch-mastermind ternary complex bound to DNA. Cell. 2006;124(5):985‐996.16530045 10.1016/j.cell.2006.01.035 · doi ↗ · pubmed ↗
- 7Iso T, Kedes L, Hamamori Y. HES and HERP families: multiple effectors of the Notch signaling pathway. J Cell Physiol. 2003;194(3):237‐255.12548545 10.1002/jcp.10208 · doi ↗ · pubmed ↗
- 8Kobayashi T, Kageyama R. Expression dynamics and functions of Hes factors in development and diseases. Curr Top Dev Biol. 2014;110:263‐283.25248479 10.1016/B 978-0-12-405943-6.00007-5 · doi ↗ · pubmed ↗