Mice
We generated CD206CreERT2; Tgf-β1f/f mice by crossing CD206 CreERT2 mice and Tgf-β1 flox/flox (Tgf-β1f/f) as previously described by Nawaz et al 46. All animals were housed, 6 mice in one cage, in a room with the temperature (24\(\:\pm\:\)2 degrees and humidity (55\(\:\pm\:\)5 percentage) were controlled automatically, and the cycle of light/dark was maintained at 12:12 hours. The mice had free access to water ad libitum and food (a normal chow diet: CE-2 CLEA, Japan).
Genotyping
Whole genomic DNA was obtained by lysis tail tissue with Direct PCR (Tail) lysis solution (Viagen Biotech) and proteinase K (Roche Diagnostics), following the manufacturer's instructions. We performed PCR by using a Tks Gflex DNA polymerase kit from TAKARA (Shiga, Japan) with this crude DNA.
PCR conditions for CD206 CreERT2 included segment 1: 1 cycle of 94 degrees for 1 minute, segment 2: 30 cycle includes 98 degrees for 10 seconds; 58 degrees for 30 seconds; and 68 degrees for 30 seconds. Then PCR productions were kept at 4 degrees. The expected DNA fragment size is 299 bp. The primers used for PCR had the sequence GGTCGATGCAACGAGTGATGAG
(primer 1) and GTGAAACAGCATTGCTGTCACTTGG (primer 2)
The PCR condition for Tgf-β1f/f included segment 1: 1 cycle of 94 degrees for 1 minute, segment 2: 40 cycle including 98 degrees for 10 seconds; 54 degrees for 30 seconds; and 68 degrees for 30 seconds. Then PCR productions were kept at 4 degrees. The expected DNA fragment sizes of WT and f/f mice were 210 bp and 338 bp, respectively. The primers' sequences were AAGACCTGGGTTGGAAGTG (primer 1) and CTTCTCCGTTTCTCTGTCACCCTAT (primer 2). Both primers for PCR were purchased from Invitrogen™ Life Technology (Tokyo, Japan). Then PCR products were separated using 1.5% Agarose gel (Nippon gel) electrophoresis for 30 minutes. Ethidium bromide (1:1000) was added to visualize DNA on the gel.
Tamoxifen administration
We used sunflower oil (WAKO) to dissolve tamoxifen (TAM: sigma-Aldrich) incubated at 55 degrees and vortexed every 5 minutes until dissolved. After dissolved, TAM was administered to both Tgf-β1f/f and Tgf-β1 KO at the dose of 225mg/kg body weight for five consecutive days, as previously described 10 at the 18-week-old following schematic protocol in Fig. 1B.
Glucose tolerance and Insulin tolerance test
For the intraperitoneal glucose tolerance test (ip-GTT), the mice were fasted for 5 hours. Glucose was injected into both Tgf-β1 KO and Tgf- β1f/f at a dose of 1mg/g body weight. The blood glucose level was measured at 0, 15, 30, 60, 90 and 120 minutes. For the intraperitoneal insulin tolerance test (ip- ITT), mice were fasting for 4 hours. Both Tgf-β1 KO and Tgf- β1f/f mice were injected with human insulin (Humalin R) with a dose of 0.8 units/g. The blood glucose level was measured at 0, 15, 30, 45, 60, 90, and 120 minutes. In both ip-GTT and ip-ITT, the blood glucose level was taken from the tail vein using the STAT STRIP Express 900 (Nova Biomedical, Waltham, MA).
Real-time polymerase chain reaction (RT-PCR)
eWAT whole tissue was collected and extracted using the Qiagen RNeasy kit following the manufacturer's instructions. The TaKaRa PrimerScript RNA Kit was used following the company's guidance for reverse transcription. The quantitative PCR amplification reaction was performed using gene-specific primers (provided in Supplementary Table S2) and TB Green Fast Premix (Takara, Shiga Japan), followed by the manufacturer's instructions. The relative mRNA expression levels were calculated by \(\:\varDelta\:\varDelta\:\)Ct value and normalized by internal control TF2B or RPL13a.
Flow cytometry analysis
To isolate and prepare stromal vascular fraction (SVF) of eWAT 47,48. Tissue was collected and digested in collagenase (Sigma) for 45 minutes at 37 degrees before filtering through a 100- µm strainer to harvest a single cell. The 7AAD− population was gated to analyze lineage-negative (CD31−CD45−) populations, followed by Sca1+, then separated into Dpp4+, Icam1+, and Dpp4+ Icam1+ populations. For justification of the gating strategy, unstained and fluorescence minus one (FMO) were used. All this experiment and cell sorting were performed using BD FACS Aria™ SORP II machine and the FlowJo offline software (v10) to analyze data.
Magnetic-activated cell sorting study
SVF was dissociated from eWAT tissue as previously described 8,47. The SVF was processed for magnetic cell sorting with anti-Pdgfr\(\:\alpha\:\) microbeads, then we collected a positive population, extracted RNA, and performed qPCR analysis of adipocyte progenitors and cell cycle. All incubation and procedure were performed at 4 degrees for 10 to 15 minutes following the manufacturer’s instructions. Microbead Kit was purchased from Miltenyi Biotech.
Histology
After collection, tissue was fixed in 4%PFA, and paraffin sections were prepared with 5–10 µm thickness and then mounted on the slide.
For Hematoxylin and Eosin (H/E staining), the slide was stained with hematoxylin and eosin. Hematoxylin eosin was captured using Keyence BZ-X800 with a 20x lens (scale bar 200 µm).
Immunohistochemistry
After collection, tissue was fixed in 4%PFA, and paraffin sections were prepared with 5–10 µm thickness and then mounted on the slide. As described previously, paraffin-embedded tissue sections were used in immunohistochemical staining. The primary and secondary antibodies are used following the manufacturer's instructions, with the ratio for primary antibody being 1:100, the secondary antibody being 1:250, and DAPI being 1:400. Primary antibodies included CD206, Tgf-β1, PDGFRα, and mKi-67. Secondary antibodies included anti-rabbit, anti-mouse, and anti-goat. All primer sources were provided in Supplementary Table S1. All images were taken by an LSM 900 with an Arycan confocal microscope.
Quantification of adipocyte size
The number of adipocytes was counted at 3.9x105 µm2 (area). The multi-point tool in ImageJ 1.53a (National Institute of Health, USA) was used for adipocyte counting. The “ Set Scale” function in ImageJ adipocyte size was used to analyze adipocyte size manually. We measured 4 random fields/specimens, with 4 specimens in each group.
Statistical Analysis
Statistical significance between the Tgf-β1 KO and Tgf-β1f/f group was performed using two-way ANOVA followed by the Sidak multiple comparison test for GTT and ITT. Other data used two-tail unpaired Student’s t-test, *p < 0.05, ** p < 0.01, ***p < 0.001, ****p < 0.0001. Data are expressed as mean ± SEM.