Animals

All animal experiments were approved by the government of Middle Franconia. Mice were housed in the Franz-Penzold-Zentrum animal facility of the University of Erlangen-Nuremberg.

For in vitro osteoclast differentiation and assay, C57BL/6JRj (Stock: 000664) mice were bought from Charles River Laboratories.

Casp8fl/fl mice bearing the caspase 8 floxed allele, (B6.129-Casp8tm1Hed/J) were purchased from Jackson Laboratory (JAX stock #027002). The exon 3 of the caspase 8 gene was flanked with loxP sites by homologous recombination.

Cx3cr1cre mice (Stock: Tg(Cx3cr1cre)MW126Gsat/Mmucd, identification number 036395-UCD) were obtained from the Mutant Mouse Regional Resource Center (MMRC), a National Institutes of Health (NIH)-funded strain repository, and were donated to the MMRRC by the National Institute of Neurological Disorders and Stroke (NINDS)-funded GENSAT BAC transgenic project.

We crossed B6.129-Casp8tm1Hed/J and Tg(Cx3cr1cre)MW126Gsat/Mmucd mice to obtain Casp8fl/flCx3Cr1cre mice.

Bone marrow cell isolation, culture and differentiationMouse bone marrow monocyte-derived osteoclasts

Bone marrow was isolated from the hind long bones of control C57/BL6 mice and differentiated into osteoclasts. Shortly, femur and tibia were aseptically collected from animals at approximately 12 weeks of age. After cutting the metaphysis, the bone marrow was flushed out with sterile PBS on a cell strainer to obtain a single cell suspension. After centrifugation, erythrocyte lysis was performed using ACK Buffer and the solution was neutralized with PBS. The cell suspension was then filtrated, centrifuged again and resuspended in medium supplemented with M-CSF-conditioned media. The next day, non-adherent cells were harvested and centrifuged. Afterwards, cells were resuspended in medium containing RANKL (10 or 50 ng/mL) and M-CSF (L929 conditioned medium) and plated in 96- or 6-well plates. Control macrophages were cultured with M-CSF alone. After 2 to 5 days, cells were used for TRAP staining, RNA isolation for real-time PCR, or protein isolation for Western blot analysis.

Macrophage-derived osteoclasts

For Casp8fl/flCx3Cr1cre mice, bone marrow was isolated according to the previously described bone marrow isolation protocol and non-adherent cells were cultured in M-CSF containing medium for 5 days in order to increase CX3CR1 expression and thus, optimize caspase 8 deletion. After the “long M-CSF treatment”, cells were plated as described above in the presence of M-CSF and RANKL.

Human peripheral blood monocyte-derived osteoclasts

Peripheral Blood Mononuclear Cells (PBMC) were isolated from healthy donors and differentiated in vitro into osteoclasts. Briefly, 45 mL of blood were drawn, diluted 1:1 with PBS and layered onto 10 mL of Lymphoflot for density gradient centrifugation. The buffy coat was aspirated, transferred into a fresh tube and washed three times with ice cold 1 mmol/L EDTA-PBS. Cells were plated in the presence of hTGF-β (1 ng/mL), hM-CSF (30 ng/mL) and hRANKL (3 ng/mL) in a 96 well plate for TRAP staining or in 6-wells plates for Western blot protein sampling.

Cell lines and culture

MØP cells: Macrophage precursor cell line (MØP). Briefly, MØP cells were generated by the Taylor Lab using a MMLV-derived retroviral vector which expresses an estrogen-dependent Hoxb transcription factor (Rosas et al.13) from a mouse constitutively expressing Cas9. Cells were maintained in culture in RPMI supplemented with 10% FCS and 1% Penicillin/Streptomycin in the presence of GM-CSF (10 ng/mL) and β-oestradiol (1 μmol/L). For osteoclast differentiation, MØP cells were pelleted, washed and resuspended in medium containing RANKL (50 ng/mL) and M-CSF (L929 conditioned media), and plated in 96- or 6-well plates. Control macrophages were cultured with M-CSF only. After 5 to 7 days, cells were used for either TRAP staining or apoptosis assay.

MØP CRISPR/Cas 9 knockout: MØP Cas 9 cells were transduced with a gRNA pair targeting mXkr8 via a lentiviral vector in order to generate MØP Xkr8 knockout cell line. After selection, cells were differentiated into osteoclasts according to the description above.

Bone analysesEx vivo bone density assessment and imaging

For each mouse, one tibia and the vertebral were prepared from the mice and fixed overnight at room temperature in 4% paraformaldehyde solution in PBS. The next day, samples were transferred to a 70% ethanol solution and analysed by µCT.

Bone density was measured with a SCANCO Medical µCT 40 scanner to produce the images and analysed with SCANCO evaluation software for segmentation, 3D morphometric analysis, density and distance parameters.

Immunofluorescence staining

For staining of coverslip-cultured osteoclasts, samples were permeabilized with 0.1% Triton X-100 in PBS for 30 min at room temperature and blocked with 0.2% BSA in PBS for 1 h. For immunofluorescence staining, the antibodies listed in the antibody table were used. Stainings were performed for 1 h at room temperature using the indicated antibodies diluted 1:100 in blocking solution. Active Caspases stainings were performed with antibodies from the Apoptosis Kit. Unbound primary antibodies were washed off with blocking solution and unlabeled primary antibodies were counterstained with donkey anti-Rabbit IgG AF594 or AF647 antibody in blocking solution for 1 h at room temperature and washed with 0.05% Tween-20 in PBS. Samples were stained with DAPI (1:2 000 dilution), washed two times with PBS, and embedded onto a glass slide with Fluorescence Mounting Medium.

Confocal laser scanning microscopy

For high-magnification imaging of osteoclasts, a Leica TCS SP 5 II confocal microscope with acousto-optic tunable filter and acousto-optical beam splitter was used. Imaging of coverslip-cultured samples was performed using an HCX PL APO 63× glycerol objective. Fluorescence signals were generated via sequential scans, exciting Vybrant Dye Red or Alexa Fluor 594 using a diode-pumped solid-state laser (DPSS) at 561 nm, Alexa Fluor 488 or FITC-labeled staining using an argon laser at 488 nm for excitation. A third imaging sequence involved a simultaneous excitation of DAPI with a 405-nm argon laser and of Alexa Fluor 647 with a 633-nm helium-neon laser. Generated images were projected on the z-axis with ImageJ software.

Cytochrome C quantification

The generated images were visualized and quantified with Imaris software. The amount of cytochrome C localized in cells was determined by volume rendering of the fluorescence signal for the corresponding channel and the phalloidin channel and overlapping volumes were calculated for each condition.

Spinning disk confocal microscopy

For spinning disk confocal microscopy of histological joint sections, an inverted Zeiss Spinning Disc Axio Observer.Z1 with a Yokogawa CSU-X1M 5000 spinning disk unit, a LD C-Apochromat 63× water immersion objective (NA 1.15) and an Evolve 512 EMCCD camera was used. Fluorescence signal of pSIVA (AF488) was excited and detected at λex: 488 nm DPSS laser and λem: 525/50 nm BP filter. Acquired images were processed via Zen Blue 2.3 image acquisition software.

Enzyme-linked immunosorbent assay

For serum analysis, blood was harvested from mice by cardiac or submandibular vein puncture and serum was separated with serum separation tubes. For cell culture supernatant analysis, culture supernatants were harvested and cleared by centrifugation before analysis. All ELISAs were performed according to the instructions of the respective manufacturer’s protocol.

Plasma membrane extracts – protein extraction

The extraction and purification of plasma membrane and cytoplasmic protein fraction was performed using the Plasma Membrane ProteoExtract Kit, according to the manufacturer’s protocol.

Western blotting

For murine osteoclast differentiation, cells were plated at 3 × 106 cells per well in a 6-well plate. After 72 h of osteoclastogenic differentiation, cells were washed with ice-cold PBS, harvested for sampling by lysis in SDS sample buffer containing β-mercaptoethanol and denaturated at 95 °C for 10 min.

For human osteoclast differentiation, cells were plated at 6 × 106 cells per well in a 6-well plate. After 7 days of differentiation, cells were washed once with ice-cold PBS and harvested for sampling in PBS by scraping the bottom of the well using a plastic policeman. Protein fractions were extracted using the ProteoExtract Kit. To adjust protein concentration, extracts were quantified with the Pierce BCA Protein-Assay Kit according to the manufacturer’s instructions and denatured in SDS sample buffer containing β-mercaptoethanol.

Protein extracts were separated on 12% SDS–polyacrylamide gels and transferred to nitrocellulose membranes. Membranes were blocked with 5% milk powder in Tris-buffered saline solution with 0.05% Tween-20 for 1 h. Blots were probed overnight with antibodies against specific full-length caspases, cleaved caspases, Xkr8, GAPDH or β-actin. Horseradish peroxidase–conjugated immunoglobulin G was used as a secondary antibody. HRP signals were revealed with Pierce ECL Western blotting substrate on radiography films. For sequential detections, blots were stripped with ReBlot Plus Strong Antibody Stripping Solution, blocked, and probed as described earlier.

Statistical analyses

Datasets are shown as means ± SEM with sample sizes indicated in each legend. Outliers within datasets were excluded according to Grubb’s test for variation from a normal distribution. Values below the detection limit were defined as zero. All statistical analyses were performed using GraphPad Prism 8 either by two-tailed Student’s t-test or two-tailed Mann-Whitney U test, unless stated otherwise. Group differences were considered statistically significant when P-value ≤ 0.05.

Single-cell RNA sequencing

Single cell RNA sequencing was performed on in vitro differentiated osteoclasts from a C57BL/6JRj wild-type mouse, at day 3 of culture with 10 ng/mL RANKL and 30 ng/mL. Single cells were captured with the 10X Genomics Chromium system. Sequencing library was generated using the 10X Genomics Single Cell 3’ Solution kit. Sequencing was performed with an Illumina sequencing system (HiSeq 4000) according to manufacturer’s protocol. Alignment and quantification of sample count matrices were performed using the 10X Genomics Cell Ranger pipeline.

Computational analysis of single-cell RNA sequencing data

Computational analysis was performed with R GNU (version 4.0.3) using Seurat R package (3.2.3). During quality control, cells expressing more than 10% mitochondrial gene reads and less than 800 features were excluded. Additionally, cells containing between 300 and 25 000 RNA transcript counts were selected for analysis. 4 497 cells remained for analysis. Cell cycle phase scoring and regression were performed in order to mitigate the effects of cell cycle heterogeneity. Read counts per cell were normalized and scaled using the regularized negative binomial regression and variable features identified via the SCTransform function. The first 18 principal components were retained for clustering using the ElbowPlot function. Cells were clustered using a graph-based shared nearest neighbor (SNN) approach, dimensionally reduced and visualized with a Uniform Manifold Approximation and Projection (UMAP) at a resolution of 0.5.

The significantly differentially regulated marker genes for each cluster present in at least 25% of all cells were identified by Wilcoxon rank-sum test, with an adjusted P value < 0.05 by Bonferroni correction via the FindAllMarkers function. Gene expression was visualized for single cell on UMAP plot using the FeaturePlot function.

Enrichment analysis utilizing active subnetwork was performed using the PathfindR package (1.6.0), KEGG protein interaction network (Kyoto Encyclopedia of Genes and Genomes) and murine mmuKEGG database reference.

Construction of single-cell trajectories, identification of genes changing as a function of pseudotime and clustering of genes by pseudotemporal expression patterns were performed using the Monocle3 package (0.2.3.0). Seurat object was transferred into the Monocle3 analysis using the SeuratWrappers package (0.3.0) for the pseudotime analysis. Pseudotime calculations were performed on the top 1 000 differentially expressed genes between clusters.