Generation of IL-11 knockout mice
IL-11 conventional knockout (IL-11-/-) mice (Accession No. CDB0614K, http://www2.clst.riken.jp/arg/mutant%20mice%20list.html) were generated by using TT2 ES cells 24 (Supplementary Fig. 10a). IL-11-/- mice appeared normal with similar fertilizing ability, body length and growth curve compared to their wild-type (WT) littermates (Supplementary Fig. 6a,6b). Genomic PCR confirmed that murine IL-11 gene was present only in WT mice but not in IL-11-/- mice (Supplementary Fig. 10b). Female mice were used for all experiments.
All animal experiments were performed according to the guidelines of the Animal Research Committee, the University of Tokushima Graduate School of Health Biosciences and the Institutional Animal Care and Use Committee of RIKEN Kobe Branch.
All the mice were housed in standard conditions, 12 hours light/dark cycle in 22-25°C. Most of the analyses in this study were performed in female mice. To determine the effect of high-fat diet (HFD)-induced obesity, mice of each genotype were allocated into two groups with either a regular diet (RD) or a HFD after weaning at 4 weeks with free access to water. Micro CT analysis of adipose tissue areas was performed at 12 and 24 weeks. The time of each measurement is described in the Figure legends.
Compositions of RD (MFG chow, Oriental Yeast Co. Ltd, Japan) and HFD (F2HFHSD diet, Oriental Yeast Co. Ltd, Japan) were as follows:
For HFD, the source of fat was tallow (14 wt%), lard (14 wt%) and soybean oil (2 wt%), the source of carbohydrate was sucrose (20 wt%), cornstarch (14.87 wt%), and the source of protein was casein (25 wt%). Other constituents included cellulose powder (5 wt%), AIN-93 vitamin mixture (1 wt%) and AIN-93G mineral mixture (3.5 wt%).
Generation of TOPGAL;IL-11-/- mice
Wnt indicator TOPGAL mice were obtained from The Jackson Laboratory (Bar Harbor, ME, USA). TOPGAL mice were bred with IL-11-/- and WT mice to generate TOPGAL;IL-11-/- and TOPGAL;WT mice.
Generation of conditional IL-11 knockout mice
To establish IL-11 floxed mice (Accession No. CDB1231K: http://www2.clst.riken.jp/arg/mutant%20mice%20list.html) in which exon 4 of IL-11 gene was flanked by loxP sequences, IL-11-targeted mice carrying a flippase (FLP) recognition target (FRT)-flanked neomycin resistance gene cassette were generated by using TT2 ES cells 24 (Supplementary Fig. 8a). Because these homozygous IL-11-targeted female mice were infertile, we obtained embryo from these mice, and deleted neomycin cassette using FLP-FRT recombination by electroporation of FLP mRNA into the embryo. Genomic PCR confirmed that the size of murine IL-11 gene was larger in floxed mice compared with WT mice (Supplementary Fig. 8b). Transgenic mice expressing Cre recombinase under the control of osteocalcin promoter (Ocn-Cre) 25 (courtesy of Prof. Hiroshi Takayanagi and Dr. Kazuo Okamoto, University of Tokyo) and adiponectin promoter (Apn-Cre) 26 (courtesy of Prof. Wataru Ogawa and Dr. Tetsuya Hosooka, Kobe University) were mated with IL-11 floxed (IL-11fl/fl) mice to obtain conditional knockout mice. Genomic nested PCR of Ocn-Cre;IL-11fl/fl, Apn-Cre;IL-11fl/fl and IL-11fl/fl confirmed that exon 4 of IL-11 gene was deleted in the bone only in Ocn-Cre;IL-11fl/fl mice, and was deleted in the AT only in Apn-Cre;IL-11fl/fl mice (Fig. 6a).
Genomic DNA extracted from mice tails was analyzed by PCR. Genotyping was performed according to the genomic PCR protocol. Briefly, 12.7μl ddH2O, 2μl 10x buffer, 1.6μl dNTP, 0.8μl 12.5pmol/l forward and reverse primers, 0.1μl Ex Taq DNA polymerase (TaKaRa Bio Inc., Japan) and 3μl DNA sample were used. The PCR program used was as follows: 95°C for 5 minutes, followed by 35 cycles of 95°C for 1 minute, 60°C for 1 minute, 72°C for 1 minute, and the final step of 75°C for 5 minutes.
Different primer sets were used to distinguish WT and conventional IL-11-/- mice: for WT forward (intron 2) 5’-agattggagggacagggaat-3’, reverse (intron 4) 5’- atttgggggacacaaaacaa-3’, and for IL-11-/- forward (exon 2/intron 2) 5’-agctgcacagatggtaggagattg-3’, reverse (within NeoR cassette) 5’-tatgatcggaattcgatagcggcc-3’. The sizes of the PCR product of WT allele was 737 bp, and that of KO was 401 bp.
In order to distinguish IL-11fl/fl from WT and conditional knockout mice, we first amplified IL-11 gene with the following primers: forward (intron 2) 5’-ttgggcacttgacgaagggg-3’, reverse (intron 4) 5’-ggcatcttaagacctaggcctc-3’. To distinguish IL-11fl/fl, Ocn-Cre;IL-11fl/fl, and Apn-Cre;IL-11fl/fl, the following primer sets were used for nested PCR: forward (exon 3) 5’-atgagcgctgggacattggg-3’, reverse (intron 4) 5’-tcatgggctgcgatttgggg-3’. The sizes of the PCR products of IL-11fl/fl and the conditional KO mice, Ocn-Cre;IL-11fl/fl and Apn-Cre;IL-11fl/fl, were 890 bp and 254 bp, respectively.
Establishment of Il11ra gene knock-out cell lines by VIKING method
CRSPR/Cas9-mediated Il11ra gene editing was conducted following the VIKING method described previously 27. For genome editing of the mouse Il11ra gene locus (Ensemble ID: ENSMUST00000098132.10), annealed oligonucleotides comprising the sequences of Il11ra gene (5'-CACCGATTCCACCCGCAGTCCTTG-3', 5'-AAACCAAGGACTGCGGGTGGAATC-3') were cloned into pX330 (Addgene; #42230) as a locus-specific cleaving vector. For the VIKING method, a donor vector pVKG1-Puro (Addgene; #108670) and a donor cleaving vector VKG1-gRNA-pX330 (Addgene; #108671) were prepared as VIKING modules.
MC3T3-E1 cells were suspended in Opti-MEM (11058-021; Life Technologies, USA) and transfected with 15 µg DNA in the VIKING modules and Il11ra locus-specific cleaving vector using a Lipofectamine LTX reagent (15338100; Life Technologies) according to the manufacturer’s protocol. Transfected cells were seeded into 100-mm dishes and pre-cultured in MEMa without antibiotics for 24 h. Transduced MC3T3-E1 cells were selected following puromycin treatment (0.1 µg/mL) for 14 days to isolate clonal colonies. For genome sequencing, PCR amplification from genomic DNA of each isolated cell line was conducted using primers “5’-TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGACACACACTGTGGGAAGGAAT-3” and “5’-GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACCGAGACACACTGCAGAC-3” with GoTaq master mix (M7132; Promega, USA) or KOD FX Neo (KFX-201; Toyobo, Japan) according to the manufacturer’s protocol. Each PCR product was directly sequenced using next-generation sequencing (MiSeq; Illumina, USA) (Supplementary Fig. 1a-1c).
Immunoblotting of phosphorylated STAT1 and 3
Cells were collected and lysed in lysis buffer (Thermo Fisher Scientific) after stimulation of mIL-11 (R&D systems), electrophoresed on a 10% SDS-PAGE, and blotted onto PVDF membranes (Biorad). After blocking, the membranes were incubated with primary antibodies overnight at 4 °C, and then with HRP-conjugated secondary antibodies for 1h. Protein bands were visualized with a SuperSignal West Pico Chemiluminescent Substrate (Thermo Fisher). The sources for the antibodies were as follows: Rabbit monoclonal antibodies to STAT1, STAT3, p-STAT1, p-STAT3, β-actin and goat anti-rabbit IgG antibody (Cell Signaling)
Tail suspension model
Tail suspension was performed as previously described4. Female 8-week-old WT and IL-11-/- mice were used. In brief, tails of the mice were suspended to lift off the hind limbs from ground for two weeks. During the period of tail suspension, mice were able to access food and water freely. After the tail suspension, mice were allowed to move freely in the ground for three weeks (Reloading group) or sacrificed (Tail suspension group) to obtain serum and bone samples.
β-Galactosidase staining of bone tissues and adipocytes was conducted as previously described28. Briefly, tissues were collected and fixed in 2% (v/v) paraformaldehyde and 0.02% (v/v) glutaraldehyde for 1 hour, and bone tissues were decalcified in 15% EDTA for 10 days, and then incubated in 0.1% 4-chloro-5-bromo-3-indolyl β-D-galactosidase (X-Gal solution; Wako, Osaka, Japan) at 37°C overnight. Tissues were washed, then followed by post-fixation with 4% paraformaldehyde at 4°C for 16 hours, and rinsed in an ascending series of ethanol, embedded in paraffin, sectioned and counterstained with eosin.
Tibias were excised and subsequently fixed in 4% paraformaldehyde at 4°C overnight, then decalcified in 15% EDTA for 15 days. Samples were dehydrated in ethanol, embedded in paraffin blocks, and then cut into 10 μm sections. Before staining, the sections were deparaffinized in xylene, and pretreated in 3% (v/v) hydrogen peroxide/methanol (Wako, Osaka, Japan), followed by blocking using Protein Block-Serum Free (Dako, CA, USA). Then slides were incubated in 1.5% normal goat serum for 30 minutes at room temperature using Vectastain ABC Goat IgG kit (Vector Laboratories Inc., CA, USA). The primary antibody (Goat anti-sclerostin, Goat anti-Dkk1, Goat anti-Dkk2, R&D Systems, Minneapolis, MN, USA; and Goat anti-IL-11, Santa Cruz Biotechnology, Dallas, TX, USA) was 1:100 diluted and incubated at 4°C overnight. Slides were washed and incubated in secondary antibody (Vectastain kit) diluted 1:200 for 30 minutes at room temperature, followed by ABC reagent using Vectastain kit for 30 minutes in accordance to the manufacture’s protocol. After washing, slides were developed in a working solution of Imm PACTTM DAB Peroxidase Substrate kit (Vector Laboratories Inc., Burlingame, CA, USA), followed by counterstaining with Weak Methyl Green (Dako, CA, USA).
Real-Time PCR analysis
Total RNA was extracted from femoral bones or adipose tissues and isolated with TRIzol reagent (Invitrogen, CA, USA) according to the manufacture’s protocol. Complementary DNA was synthesized using PrimeScript RT Master Mix (Takara Bio Inc., Japan). Then samples were subjected to quantitative real-time PCR analysis using ABI 7300 Real-Time PCR system (Applied biosystems, Foster City, CA) with SYBR® Green Premix Ex TaqTMII kit (Takara, Shiga, Japan). The sequences of primers are listed in Supplementary Table 1.
To determine the serum bone metabolic parameters, Mouse Osteocalcin EIA Kit (Biomedical Technologies Inc., MA, USA), Mouse TRAP Assay (Immuno diagnostic system Ltd, UK) and Alkaline phosphatase (ALP) test kit (Wako, Osaka, Japan) were used for measuring serum samples in accordance with the manufactures’ protocols.
Serum leptin, adiponectin and insulin concentrations were measured with Mouse and Rat Leptin ELISA kit (BioVendor, Brno, Czech Republic), Mouse Adiponection ELISA kit (BioVendor, Brno, Czech Republic) and Ultra Sensitive Mouse Insulin ELISA kit (Morinaga Institute of Biological Science, Inc., Yokohama, Japan), respectively, according to the manufactures’ protocols.
Micro-computed tomography (μCT) analysis
Before and during μCT scan, mice were anesthetized by inhalation of isoflurane (Abbott, Tokyo, Japan). Mice were placed on abdominal position in 48 mm wide specimen holder with 96 mm pixel resolution. Hindlimbs were extended to keep the femur and spine into right angle, and then scanned from the proximal end of L1 vertebra to the distal end of L5 to measure vertebral BMD, and whole femoral bones were scanned using LaTheta LCT-200 (Hitachi-Aloka, Tokyo, Japan), as previously described29. Calvaria bone samples were extracted and stored. Bone mineral density (mg/cm3) and adipose tissue area were calculated by LaTheta software.
Bone histomorphometric analysis
Mice were double-labeled with 16mg/kgBW calcein (Sigma, St. Louis, USA) at 6 and 2 days before sacrifice. Lumbar vertebra were removed and fixed in 4% paraformaldehyde (PFA) at 4°C overnight, followed by dehydration with a series of ethanol, then embedded in methyl methacrylate monomer (MMA, Wako, Japan). The plastic sections were cut by a standard microtome (LEICA) into 7 μm samples for von Kossa staining and 4 μm for TRAP and Toluidine blue staining. Histomorphometric analysis was performed by OsteoMeasure (OsteoMetrics, Inc.,GA, USA) according to the ASBMR guideline30.
Ex vivo cell culture
Bone marrow cells were extracted from 12-week-old female mice. For osteoblastic differentiation, bone marrow stromal cells (BMSC) were cultured in α-Modified Eagle’s minimal essential medium (α-MEM; Gibco, NY, USA) supplemented with 10% fetal bovine serum (FBS, Thermo, Utah, USA) and induced with 50μg/ml ascorbic acid (Wako) and 10mM β-glycerophosphate (Sigma-Aldrich, Tokyo, Japan) (osteoblast differentiation medium). For adipogenic differentiation, BMSC were incubated with 10-6M Troglitazone (Sigma-Aldrich, Tokyo, Japan).
Oil-Red O staining
For Oil-Red O staining, BMSC were cultured for 14 days, fixed and rinsed in 60% 2-propanol for 3 times, then stained with Oil-Red O staining solution (Sigma-Aldrich, Tokyo, Japan) for 30 minutes at room temperature. The cells were again washed and rinsed with 60% 2-propanol. The Oil-Red O positive cell number was counted under microscope.
Alkaline phosphatase (ALP) staining and Alizarin Red staining
For ALP staining, cells were cultured in osteoblast differentiation medium for 7 days and fixed in 3.7% formaldehyde for 10 minutes. Then cells were incubated at 37°C with freshly prepared 1mg/ml Naphthol-AS phosphatase (Wako), 6mg/ml Fast-Blue BB (Wako), 0.5% (v/v) N,N-dimetylformaide, 1mM MgCl2, 1M Tris-HCl (pH=8.8) and stained for 5 minutes. Cells were cultured in osteoblast differentiation medium for 15 days for Alizarin Red staining, fixed in 10% formaldehyde for 10 minutes, washed and stained with 0.2% Alizarin Red (Sigma)/1M Tris-HCl (pH=8.3) at 37°C for 20 minutes.
Determination of adipocyte size and number
Visceral fat pad was extracted to determine the adipocyte size and number as previously described 31. In brief, WAT was fixed with osmium tetroxide (Sigma-Aldrich, Tokyo, Japan) and suspended in isotonic saline. To remove fibrous elements and trap adipocytes, samples were passed through 250μm and 25μm nylon filters, respectively. Adipocyte size was analyzed using a Coulter counter equipped with a 560μm aperture tube and a multichannel particle analyzer (Multisizer II, Coulter Electronics, Fullerton, CA). Adipocyte relative number was determined by dividing the total WAT weight (mg) by the estimated mean adipocyte weight (mg), which was calculated by adipocyte density (0.948mg/ml) × mean adipocyte volume (the average value of adipocyte diameter).
Glucose tolerance test and Insulin tolerance test
Mice were fasted for 16 hours before oGTT and then administrated with 1 g/kgBW glucose orally. Blood samples were collected from mice tails at various time points, and blood glucose was measured by glucometer (Medisafe mini GR-102, Terumo, Tokyo, Japan). HOMA-IR (homeostatic model assessment insulin resistance) as an index of insulin sensitivity was calculated by the following formula, where 100 pg/mL insulin corresponds to 2.6 mU/mL:
Fasting serum insulin (mU/mL) x fasting blood glucose (mg/dL) / 405 = HOMA-IR
Data are presented as mean ± SD. The significance of difference between two groups was assessed by the Student’s t-test, and the difference among multiple groups was evaluated using ANOVA test followed by appropriate post-hoc tests as described in the figure legends with SPSS version 21.0 (Chicago, USA). P < 0.05 was considered as statistically significant.