Animals. C57BL/6 WT and Grk3-/- age-matched (8-12 week-old) mice were used under standard IACUC-approved protocols in the AAALAC-accredited vivarium of UNC, and care of animals was in accordance with institutional guidelines. The Grk3-/- mouse strain was provided by Dr. Robert J. Lefkowitz (Duke University) and backcrossed >12 generations on the C57BL/6 background. The line is re-derived every 1-2 years to prevent genetic drift from the C57/BL/6 strain.
Bone marrow-derived mesenchymal stem cell isolations. BmMSC were isolated from RPMI-flushed femurs and tibias of two male mice and cultured in complete isolation media containing RPMI (Corning, 10040CV) with 10% fetal bovine serum (FBS, Atlanta Biologicals, S12450), 10% horse serum (HS, HyClone, SH30074.03), 1% 100 IU/mL penicillin G/100 µg/mL streptomycin (P/S, Corning, 30-002-Cl), and 12 µM final concentration of L-Glutamine (Corning, 25-005-Cl), as similar methodologies have been previously described (13-15). BmMSC expansion was in complete expansion media (CEM) containing IMDM (Gibco, 12440-053), 10% FBS, 10% HS, 1% P/S, and 12 µM final concentration of L-Glutamine (14, 15), which was followed by hematopoietic CD45 (Stem Cell Technologies, 19771) and CD11b (Miltenyi, 130-049-601 or 130-093-634) depletion, as recommended by (13), via magnetic negative selection at early passages 1-3. BmMSCs were passaged at 70-80% confluency and used for experiments at passages 4-15.
Chondrogenic, adipogenic, and osteogenic differentiation. Chondrogenic differentiation. BmMSCs were suspended in CEM at 1.6x107 viable cells/mL. Micromasses were made by adding 5 µL droplets of the cell suspension onto a 6-well plate and given 3-4 hours to attach. Chondrogenic media (Gibco StemPro® Chondrogenesis Differentiation Kit, A10071-01) supplemented with penicillin streptomycin was added to each well and incubated for 21 days. Cells were fixed with 10% formalin and stained using Alcian Blue. Adipogenic differentiation. BmMSCs were plated at 1x105 cells/well of a 6-well plate in CEM supplemented with 50 µM indomethacin, 5 µg/mL insulin, and 0.1 µM dexamethasone. Cells were fixed with 10% formalin and stained using Oil Red O. Osteogenic differentiation. BmMSCs were plated at 1x105 cells/well of a 6-well plate in CEM supplemented with 50 µg/mL ascorbic acid and 20 mM ß-glycerophosphate. Cells were fixed with 10% formalin and stained using Alizarin Red. For SKI-treated osteogenic differentiation, BmMSCs were plated at 2x104 cells/well in a 24-well plate in CEM, and osteogenic differentiation was induced after an overnight rest. Cells were treated with sphingosine kinase inhibitor 2 (SKI, Cayman Chemical, 10009222) at a final concentration of 5 µM or vehicle (DMSO). Fresh media changes occurred every third day using CEM plus SKI or vehicle. Cells were fixed with 10% formalin and stained using Alizarin Red stain for analysis. Images were captured using the Olympus 1X-81 microscope and MetaMorph software.
Bone marker mRNA expression (qRT-PCR). Total RNA from BmMSCs undergoing osteogenic differentiation was prepared using the RNeasy Mini/ Plus kit (Qiagen) according to manufacturer’s instructions. Reverse transcriptase cDNA synthesis was performed using iScript cDNA synthesis kit (Bio-Rad, 170-8891). qRT-PCR was performed in duplicate (SYBR® Green, Bio-Rad, 172-5261) and normalized to housekeeping gene IDUA. Mean fold change of alkaline phosphatase was determined by -2ΔΔCt with WT day 0 as control. Primers utilized for qRT-PCR were Alkaline Phosphatase forward: AAG GCT TCT TCT TGC TGG TG, Alkaline Phosphatase reverse: GCC TTA CCC TCA TGA TGT CC; IDUA forward: GCA TCC AAG TGG GTG AAG TT and IDUA reverse: CAT TGA GCA GGT CCG GAT AC.
ELISA. BmMSCs were plated at 1x105 cells/well of a 6-well plate in CEM and rested overnight for attachment before supernatant collections began at baseline (day 1) and each subsequent collection. Supernatants of undifferentiated BmMSC monolayers were collected and analyzed for CXCL12 protein using the CXCL12 DuoSet ELISA kit (R&D Systems, DY460), as per instructions.
Micro-computed tomography (µCT). For 8-12 week-old mice, µCT imaging was used to analyze the trabecular bone morphology at the distal femur at 12 micron resolution. The metaphyseal region of the distal femur was scanned beginning 720 microns proximal to the growth plate and extending 1500 microns towards the diaphysis of the femur. An automatic script was used to analyze the region of interest to separate the trabecular and cortical regions of the bone and quantify bone morphology. Trabecular analysis includes quantification of BV/TV (bone volume/total volume). For 17-20 month-old aged mice, µCT imaging morphology (VivaCT80, Scanco Medical, Basserdorf, Switzerland) was used to analyze the trabecular bone. Briefly, metaphysis region was selected for 100 slices under the femur growth plate. Trabecular analysis includes quantification of BV/TV. Analyses were conducted at 12 μm slice increment with an integration time of 300 ms, a current of 145 mA, and an energy setting of 55 kV. The threshold was chosen using 2D evaluation of several slices in the transverse anatomic plane so that mineralized bone was identified but surrounding soft tissue was excluded.
Cellular proliferation. Proliferation of 5x104 BmMSCs/well of a 6-well plate was analyzed after exogenous CCK-8 (Dojindo Molecular Technologies) was added to each well at the indicated timepoints. Absorbance was measured after 3-hour incubation using the Promega Glomax® Multi + Detection System. Data were analyzed by deducting background (media and CCK-8) absorbance from raw absorbance reads. For S1P studies, BmMSCs were plated at 1x104 cells/well in a 24-well plate and treated with SKI at a final concentration of 5 µM.
Immunoblotting. BmMSCs were plated in CEM at 2.25x105 cell density in a 6-well plate and incubated overnight. DPBS rinsed cells were rinsed three times with IMDM containing 20% charcoal-stripped FBS (to remove serum S1P) and 1% P/S, and incubated at 37ºC for 15 minutes. Fresh media was added, and BmMSCs were stimulated with 1 µM S1P for indicated timepoints. Unstimulated BmMSCs served as 0 minute timepoint control. Following stimulation with S1P, BmMSC cells were rinsed with DPBS and lysed in cold HBSS + 1% TritonX100 lysis buffer containing protease inhibitors (1 mM PMSF, 1 µg/mL aprotinin, 1 µg/mL pepstatin, and 1 µg/mL leupeptin) and phosphatase inhibitors (5 mM NaF and 2 mM NaVO4). All WT and Grk3-/- BmMSC lysates were normalized via Bicinchoninic acid (BCA) assay, and 6 µg of total protein in laemmli sample buffer (non-reducing) was freshly loaded onto AnyKD Mini-PROTEAN®TGX precast protein gel (Bio-Rad, 4569036). Gels were run at 100V for 1.5 - 2 hours in 1XTris/Glycine SDS buffer. Proteins were transferred overnight at 4ºC onto nitrocellulose membrane in Tris base (25 mM)/Glycine (192 mM) transfer buffer containing 20% methanol. The membrane was blocked in 3% fatty-acid free BSA in TBS plus 0.1% Tween-20 (TBS/T) for 2 hours at 25ºC and incubated with primary antibody 1:2,000 phospho-p44/42 MAPK or 1:2,000-3,000 p44/42 MAPK (Cell Signaling Technologies, 4370/4695) overnight at 4ºC, or 1:10,000 GAPDH (Trevigen, 2275-PC-100) for 2 hours at 4ºC. The membrane was washed three times for 10 minutes in TBS/T, incubated with secondary antibody 1:5,000 anti-rabbit IgG HRP (Cell Signaling Technologies, 7074) for 1 hour at 25ºC, and washed twice for 10 minutes in TBS/T and once in TBS. Detection was performed via ECL Prime or ECL Select (GE Healthcare, RPN2232/ RPN2235) and imaged on GeneSys image acquisition software. Densitometry was obtained by measuring ratio of phospho-ERK (pERK) over total ERK using Image J software.
β-arrestin recruitment assay. GRK recruitment of β-arrestin-2 to the S1PR1 carboxy-terminus was measured using agonist-stimulation in a modified-TANGO assay, as previously reported (16). HTLA cells were transfected with a S1PR1-TCS-tTA receptor construct after removing the V2 vasopressin sequence to prevent nonspecific β-arrestin recruitment to the wildtype S1PR1. GRK over-expression was achieved via plasmids: GRK2 pcMyc_LIC and GRK3 pcMyc_LIC, and utilized a separate expression vector encoding yellow-fluorescent protein (YFP) that was simultaneously transfected to serve as a transfection control. HTLA cells were transiently transfected with 6.5 μg of total plasmid DNA (3 μg of S1P-Tango, 0.5 μg of YFP, and either 3 μg of empty-vector control, GRK2, or GRK3) via calcium-phosphate precipitation. Transfection efficiency was determined by YFP epifluorescence to be consistently >70% at 24 hours post-transfection. Cells were serum starved and then stimulated with S1P ligand at varied concentrations up to 1 µM. BriteGlo reagent (Promega, Madison, WI, USA) was added for luminescence detection via Promega Glomax® Multi + Detection System (0.5 sec/ well). Raw data were normalized by subtracting background for each independent run and setting the lowest concentration of the control condition at 0% and highest concentration at 100%.
S1PR1 internalization assay. BmMSCs cultured in CEM were rinsed with DPBS and incubated in serum-free CEM (SFM) for 2.5 hours. BmMSCs were detached using Accutase® (Sigma, A6964), rinsed once with cold SFM containing 5% charcoal-stripped FBS (to remove serum S1P), and resuspended in cold DPBS. 1x105 BmMSCs were stimulated with 1 µM S1P ligand in 100 µL FACS buffer (DPBS1X + 0.2% fatty acid free BSA + 0.1% sodium azide) at specific timepoints. Unstimulated BmMSCs served as 0 minute timepoint control. S1PR1 internalization was halted with 2 mL of ice cold FACS buffer and sample tubes were placed on ice. BmMSCs were stained for Sca1 (eBioscience clone: D7, APC-conjugated) and S1PR1 (R&D Systems clone: 713412, PE-conjugated) for 30 minutes on ice in 100 µL FACS Buffer, rinsed, and analyzed by flow cytometry.
Statistical analyses. All data were graphed utilizing GraphPad Prism v.7 and statistically evaluated using GraphPad Prism v.7 or Microsoft Office Excel program. Taking into consideration time and strain (WT and Grk3-/-), the bone marker qRT-PCR, CCK-8 proliferation, and S1PR1 internalization was statistically analyzed using a RM two-way ANOVA with Sidak’s multiple comparison test, a method preferred over Bonferroni due to increased power (17, 18). Similarly, taking into consideration time and strain (WT and Grk3-/-) but with multiple treatment groups, the SKI-treated CCK-8 proliferation was analyzed using a RM two-way ANOVA with Tukey’s multiple comparison test for pairwise comparisons (17, 19). Student’s t-test compared two independent groups (WT and Grk3-/-) for the CXCL12-detection ELISA, microcomputed tomography data, and western blot densitometry. Taking into consideration three independent groups (empty vector, GRK2, and GRK3), the β-arrestin recruitment assay was analyzed by one-way ANOVA with Dunnett’s multiple comparison test, which compared each group to the control (empty vector) (17, 20).