We obtained Mir 221 KO cond (Mir221 ^tm2; EMMA ID, EM:05507) from EMMA (the European Mouse Mutant Archive) and Helmholz Zentrum München (Neuherberg, Germany). The targeting vector cassette composed of Neo flanked by FRT and distal loxP sites was inserted outside the genomic region encoding Mir221 and Mir222 precursors. Neo was removed by Flp in the conditional KO allele of Mir 221 KO cond. The conditional KO allele was confirmed by primer pair; 1544_29 (5’- GCT CTG TTT TCC TAA GTG ATG G-3’) and 1544_30 (5’-CTG ACA GGA AGT AAA TCA TCT TAG C-3’). The expected fragments were 265 bp in wild and 384 bp in conditional KO allele. We also used B6.FVB-Tg (Adipoq-cre)1Evdr/J to produce Mir221/222AdipoKO mice by crossing to Mir 221 KO cond. Adipoq-cre transgene was screened by primers, Forward (5'-AGC GAT GGA TTT CCG TCT CT-3') and Reverse (5'-CAC CAG CTT GCA TGA TCT CC-3'). The primers, oIMR7338 (5’-CTA GGC CAC AGA ATT GAA AGA TCT-3’) and oIMR7339 (5’-GTA GGT GGA AAT TCT AGC ATC ATC C-3’), were used for internal positive control. The expected sizes for transgene and internal positive control were 200 bp and 324 bp, respectively. By crossing male Tg (Adipoq-cre) and female Mir221 ^tm2 /y C57BL/6JJcl mice, we generated male Mir221/222 ^flox/y and male Mir221/222AdipoKO littermates.

Five-week-old mice were randomly assigned to standard diet (STD) group (MF, Oriental Yeast, Japan) or high fat high sucrose diet (HFHS) group (D12331, Research Diets, New Brunswick, NJ). The detailed formulation of STD (10) (https://www.oyc.co.jp/bio/LAD-equipment/LAD/ingredient.html and https://www.oyc.co.jp/bio/LAD-equipment/LAD/rodents.html) and HFHS (https://www.eptrading.co.jp/service/researchdiets/pdf/D12328-D12331.pdf) diets are shown in Supplementary Table 1 . At 22 weeks of age, we obtained various organs and they were subjected to following experiments.

3T3-L1 pre-adipocytes were cultured in Dulbecco’s modified eagle’s medium (DMEM, 2124951, Gibco). On day 0, the cells were treated with the differentiation media; DMEM supplemented with 10% FBS, 10 µg/ml insulin (I1882, Sigma), 1 µM DEX (D2915, Sigma) and 0.5 mM IBMX (I5879, Sigma). Then, the media were changed to DMEM supplemented with 10 µg/ml insulin and 10% FBS on day 2 and cultured for 10 or 20 days. Undifferentiated 3T3-L1 cells were subjected to Lentiviral miRNA expression, Lentiviral miRNA inhibition studies, and luciferase reporter assays.

Human serum samples were collected from 69 patients with type 2 diabetes in Okayama University Hospital and 45 subjects with normal fasting glucose (NGT).

The 13-week-old mice were fasted for 16 hours in GTT and for 3 hours in ITT. They were then intraperitoneally injected with glucose solution (1 mg/g body weight) and human insulin (1 unit/kg in HFHS groups and 0.75 unit/kg in STD groups) for GTT and ITT, respectively.

At 18 weeks of age, O₂/CO₂ metabolism measuring system (MK-5000, Muromachi Kikai, Tokyo, Japan) were used to quantify oxygen consumption rate and carbon dioxide production for the estimation of V̇O2 and respiratory quotient (RQ). The locomotor activity was recorded for 24 hours by the frequency of interrupting an infrared sensor (ACTIMO-100, Shinfactory, Fukuoka, Japan). Daily food intake was measured and calculated; daily food intake [g/day/body weight (BW)] = [initial food weight (g) – leftover food weight (g)]/measurement period (days)/BW (g). The 6-10 mice in each experimental group were examined.

RNAs were extracted from frozen tissues and cultured 3T3-L1 cells with RNeasy Mini kit (74106, Qiagen). The QIAamp Circulating Nucleic Acid Kit (Qiagen) were used for the isolation of total RNAs from serum. For gene expression analyses, cDNAs were prepared with High-Capacity RNA-to-cDNA Kit (Thermo Fisher Scientific). TaqMan gene expression primers, Ddit4 (Mm00512504_g1), Rplp0 (Mm00725448_s1), Rn18s (Mm03928990_g1), Adipoq (Mm00456425_m1), Lep (Mm00434759_m1), Lpl (Mm01345523_m1), Srebf1 (Mm00550338_m1), Cebpa (Mm00514283_s1), Fabp4 (Mm00445878_m1), Pparg (Mm01184322_m1), Lipe (Mm00495359_m1), Pnpla2 (Mm00503040_m1), Ucp1 (Mm01244861_m1), Cox8b (Mm00432648_m1), Prdm16 (Mm00712556_m1), Cidea (Mm00432554_m1), Ppargc1a (Mm01208835_m1), G6pc (Mm00839363_m1), Gck (Mm00439129_m1), Fasn (Mm00662319_m1), Ppara (Mm00440939_m1), Il6 (Mm00446190_m1), Ifng (Mm01168134_m1), Tnf (Mm00443258_m1), Il1b (Mm01336189_m1) were used (Thermo Fisher Scientific). For miRNA expression studies, cDNAs were prepared from total RNAs by TaqMan MicroRNA Reverse Transcription Kit (Life Technologies). MicroRNA primers, mmu-miR-222-3p (CTAAJ3), hsa-miR-221-3p (000524), hsa-miR-222-3p (002276), snoRNA202 (001232), snoRNA234 (001234), and cel-miR-39 (000200) were used (Thermo Fisher Scientific). Rplp0, Rn18s, snoRNA202, snoRNA234, and cel-miR-39 were served as the invariant controls. The RT-qPCR was performed using TaqMan Universal PCR Master mix II (no UNG) at a StepOne Plus Real-Time PCR system. The quantification was performed by the 2^−ΔΔCT analysis method.

Long non-coding RNA, Mir221hg, was cloned using SMARTer RACE 5’/3’ Kit (634858, Clontech). 5’- and 3’-RACE-Ready cDNAs from epididymal adipose tissues poly A⁺ RNA were prepared by SMARTScribe Reverse Transcriptase. 5’- and 3’-RACE-Ready cDNAs were amplified by PCR using 5’ GSP (5’-GAT TAC GCC AAG CTT CCA GCA GAC AAT GTA GC TGT TGC-3’) and 3’ GSP (5’-GAT TAC GCC AAG CTT TCC AGG TCT GGG GCA TGA ACC TG-3’), respectively. Furthermore, nested PCR was performed using 5’ NGSP (5’-GAT TAC GCC AAG CTT GTA TGC CAG GTT CAT GCC CCA GAC-3’) and 3’ NGSP (5’-GAT TAC GCC AAG CTT GCA ACA GCT ACA TTG TCT GCT GG-3’). PCR products were analyzed by agarose/EtBr gel and purified by NucleoSpin Gel and PCR Clean-Up Kit. The purified RACE products were subcloned into pRACE vector by In-Fusion Cloning (Clontech). Independent 4 clones of both 5’- and 3’-RACE were sequenced.

The epididymal fat tissues from 22-week-old mice and cultured 3T3-L1 cells were homogenized in RIPA lysis buffer (radioimmunoprecipitation buffer) plus protease inhibitors. The samples were boiled in SDS-PAGE loading buffer, separated on 12% Mini-PROTEAN TGX Precast Protein Gels (Bio-Rad), and transferred to a PVDF Blotting Membrane (cytiva). After blocking with 5% nonfat milk for 1 hour at room temperature (RT), the blots were incubated with REDD-1/DDIT4 Antibody, rabbit polyclonal (ab106356, RRID: AB_10864294), C/EBPα Antibody, rabbit polyclonal (2295, RRID: AB_10692506), PPARγ (D69) Antibody, rabbit polyclonal (2430, RRID: AB_823599), mTOR (7C10) Rabbit mAb (2983, RRID: AB_2105622), Phospho-mTOR (Ser2448) (D9C2) XP Rabbit mAb (5536, RRID: AB_10691552), Akt Antibody, rabbit polyclonal (9272, RRID: AB_329827), Phospho-Akt (Thr308) Antibody, rabbit polyclonal (9275, RRID: AB_329828), p70 S6 Kinase (49D7) Rabbit mAb (2708, RRID: AB_390722), Phospho-p70 S6 Kinase (Thr389) Antibody, rabbit polyclonal (9205, RRID: AB_330944), Tuberin/TSC2 Antibody, rabbit polyclonal (3612, RRID: AB_2207804), Phospho-Tuberin/TSC2 (Thr1462) Antibody, rabbit polyclonal (3611, RRID: AB_329855, Cell Signaling Technology) overnight at 4°C. GAPDH (D16H11) XP Rabbit mAb (HRP Conjugate) (8884, RRID: AB_11129865) was used as a loading control (Cell Signaling Technology). After washing three times with Tris-buffered saline (TBS), the blots were incubated with ECL Donkey Anti-Rabbit IgG, HRP-Conjugated Antibodies (NA934V, GE healthcare Life science, 1:10000) at RT for 1 hour. The blots were developed with Pierce ECL Western Blotting Substrate (TE261327, Thermo Fisher Scientific). The chemiluminescence was analyzed using ImageQuant LAS-4000 mini (FUJIFILM).

Epidydimal adipose tissues were fixed by 10% formalin, embedded with paraffin. The 5-μm paraffin sections were prepared and stained with PAS. The images were captured using an Olympus BX51 microscope. The size of the adipocytes was analyzed by Keyence Hybrid cell count software. Epidydimal adipose tissues were taken from 3-5 individual animals from each experimental group.

SVF was isolated from epididymal adipose tissue of 24-week-old mice. Briefly, fresh mouse epididymal fat pads were minced and digested with collagenase type 1 (CLSS1, Worthington) in HBSS containing 10% FBS for 45 minutes at 37°C. The mixture was filtered through a nylon mesh (100 μm), then centrifuged at 400 g for 1 minute. The adipocyte fraction was obtained from the supernatant, they were again centrifuged at 800 g for 10 minutes, and the SVF was obtained from the pellet.

The mRNA microarray was performed by GeneChip Mouse Gene 2.0 array using total RNA of epidydimal fat obtained from 16-week-old mice (1 animal from each group) and analyzed by Filgen (Nagoya, Japan). The raw data are available in Gene Expression Omnibus (GEO; https://www.ncbi.nlm.nih.gov/geo/) (GSE163921). TargetScan (http://www.targetscan.org/vert_72/), miRDB (http://www.mirdb.org/), and DIANA-microT v5.0 (https://bio.tools/DIANA-microT) and were used to identify potential target genes for Mir221/222.

Lentiviral transduction using pLV-miRNA and pLV-miR-Locker system to 3T3-L1 cells were performed according to the manufacturer’s manual (Biosettia). After transformation to E. coli JM109 cells, pLV-[mmu-mir-221] plasmid (mir-p177m, Biosettia), pLV-[mmu-mir-222] plasmid (mir-p178m, Biosettia), pLV-[mmu-mir-221-3p] locker plasmid (mir-mp0337, Biosettia), pLV-[mmu-mir-222-3p] locker plasmid (mir-mp0339, Biosettia), pLV-[mir-control] plasmid (mir-p000, Biosettia), pLV-miR-locker control plasmid (mir-locker-ctrl, Biosettia), pMDLg/pRRE (12251, Addgene), pRSV/Rev (12253, Addgene), and pMD2.G (12259, Addgene) were isolated with EndoFree Plasmid Maxi Kit (12362, Qiagen). 293T cells (10 × 10⁷/5 mL) were transfected with each pLV plasmid, pMDLg/pRRE, pRSV/Rev, pMD2.G, Lipofectamine LTX, and 1.5 mL Opti-MEM. The supernatants were collected after 48 hours after transfection. 3T3-L1 cells were transduced with lentivirus stock in complete media containing 10 μg/mL polybrene for 12 hours, replaced with fresh complete medium

To quantitatively evaluate miRNA activity on cloned miRNA target sequence from 3’-untranslated region (3’-UTR) of Ddit4, pmirGLO dual luciferase miRNA Target expression vector (E1330, Promega) was used. Firstly, the pmirGLO plasmid was linearized by double digestion with XhoI and SalI. The cDNA of Ddit4 wild type (WT) 3′-UTR was amplified by PCR and ligated with CIP treated pmirGLO Vector. The primers are Forward-XhoI-3’UTR-Ddit4: 5’-GGG GGG CTC GAG CAG CTG CTC ATT GAA GAG TG-3′, and Reverse-SalI-3’UTR-Ddit4: 5’-GGG GGG GTCGAC CAA ACC AAC AGA GGA GAC AG-3′. pmirGLO-Ddit4 MT 3’-UTR was prepared by site directed mutagenesis by PCR using primers; Forward-MT-Seed-Ddit4: 5’-CTG GAT GTG TAT CTG CAT GTA C-3’ and Reverse-MT-Seed-Ddit4: 5’-GTA CAT GCA GAT ACA CAT CCA G-3’. The seed sequence “CGATGTA” was mutated to “CGTCTA”. After transformation to E. coli JM109 cells, pmirGLO-Ddit4 WT 3’-UTR, pmirGLO-Ddit4 MT 3’-UTR, and pmirGLO no-insert control plasmids were isolated with EndoFree Plasmid Maxi Kit (12362, Qiagen). 3T3-L1 cells were seeded at a density of 120,000 cells/mL, then co-transfected with Syn-mmu-miR-221-3p (MIMAT0000669, Qiagen), Syn-mmu-miR-222-3p (MIMAT0000670, Qiagen), negative control siRNA (1027280, Qiagen), inhibitor negative control (1027271, Qiagen), pmirGLO-Ddit4 WT 3’-UTR, pmirGLO-Ddit4 MT 3’-UTR, and pmirGLO no-insert control plasmids. Twenty-four hours after transfection, the cells were analyzed to measure luciferase activities using the Dual-Glo Luciferase Assay System and a GloMax 20/20 luminometer (Promega).

All values were represented as the mean ± standard deviation (SD). Statistical analyses were conducted using IBM SPSS Statistics 23 and GraphPad Prism (version 8.0). Independent t-test, Mann-Whitney’s U test, and one-way ANOVA with Tukey test was used to determine the differences. For correlation, non-parametric Spearman r coefficient was used. p<0.05 was considered statistically significant.

To identify miRNAs critically involved in the disease process of obesity and type 2 diabetes (T2D), we performed miRNA profiling of serum, liver and epididymal fat tissues in C57BL/6JJcl mice fed with standard (STD) and high fat-high sucrose (HFHS) chow. The Illumina RNA sequencing data (Gene Expression Omnibus number GSE61959) demonstrated that the read numbers of Mir221 and Mir222 were 5.7 and 8.2-fold up-regulated in epididymal adipose tissues in HFHS group compared with STD group ( Supplementary Table 2 ). Mir221 and Mir222 are paralog genes located in proximity on X chromosome and they share identical seed sequence. To further investigate the role of Mir221/222 in obesity and diabetes, we crossed male Tg (Adipoq-cre) mice and female Mir221^tm2/y C57BL/6JJcl mice and generated male Mir221/222^flox/y and male Mir221/222AdipoKO littermates.

Body weight of Mir221/222AdipoKO mice fed with HFHS chow was significantly reduced compared with Mir221/222^flox/y mice ( Figure 1A ). The weight of epididymal, mesenteric, subdermal, and brown fat was also reduced in Mir221/222AdipoKO mice ( Figure 1B ). The Mir221/222^flox/y and Mir221/222AdipoKO mice fed with STD chow demonstrated no significant differences in their body and tissue weight ( Figures 1A , B ). The average size of adipocytes in epididymal adipose tissues derived from Mir221/222AdipoKO mice fed with HFHS chow was smaller compared with Mir221/222^flox/y mice ( Figure 1C ). To investigate glucose homeostasis, we performed insulin tolerance test (ITT) and glucose tolerance test (GTT). The blood glucose levels of Mir221/222AdipoKO mice fed with HFHS chow were reduced in GTT and ITT compared with Mir221/222^flox/y , but they did not reach significant differences ( Figure 1D ). The serum insulin concentrations of Mir221/222AdipoKO mice were significantly lower than Mir221/222^flox/y , suggesting the improvement of insulin sensitivity in Mir221/222AdipoKO mice ( Figure 1D ). To investigate whether reduced adiposity in Mir221/222AdipoKO mice was due to changes in energy expenditure or energy intake, we measured basal metabolic rates, locomotor activity and food intake. Basal metabolic rate such as respiratory quotient and oxygen consumption were not altered between Mir221/222^flox/y and Mir221/222AdipoKO mice fed with HFHS chow ( Supplementary Figures 1A, B ). The locomotor activity was recorded for 24 hours, most of the activities were observed during the dark phase. Increased activity was observed in Mir221/222AdipoKO mice under HFHS chow compared with Mir221/222^flox/y during light (0.944 ± 0.309 vs 1.60 ± 0.764 counts/min, p=0.036) and dark (6.67 ± 4.38 vs 7.94 ± 4.38 counts/min, p=0.497) periods ( Supplementary Figure 1C ). Food consumptions were not altered in Mir221/222AdipoKO and Mir221/222^flox/y fed with HFHS ( Supplementary Figure 1D ). Under HFHS chow, serum leptin concentrations were significantly reduced in Mir221/222AdipoKO compared with Mir221/222^flox/y (24.5 ± 2.35 vs 21.5 ± 1.85 ng/ml, p=0.043), while serum adiponectin concentrations were not altered ( Supplementary Figures 1E, F ). In quantitative RT-PCR in epididymal adipose tissues, mRNA expression of Lep was significantly reduced in Mir221/222AdipoKO. The genes related to adipogenesis, such as Pparg, Cebpa, Ppargc1 and Prdm16, were not altered in Mir221/222AdipoKO ( Supplementary Figure 2 ).

We investigated the expression of Mir221 and Mir222 in various organs. Both miR-221-3p and miR-222-3p were abundantly expressed in brain, kidney, and lung in Mir221/222^flox/y and they were down-regulated or not altered by the HFHS feeding compared with STD chow. In contrast, both miR-221-3p and miR-222-3p were significantly upregulated in the epidydimal fat tissues, and such upregulation was canceled in the epidydimal fat tissues of Mir221/222AdipoKO fed with HFHS chow ( Figures 2A , B ). Next, we investigated the localization of Mir221 and Mir222 in the cell fractions of epididymal adipose tissues. Both miR-221-3p and miR-222-3p were significantly induced by HFHS chow in mature adipocytes of Mir221/222^flox/y , while the upregulation was significantly reversed in the mature adipocyte of Mir221/222AdipoKO fed with HFHS chow. In contrast, miR-221-3p and miR-222-3p were not significantly upregulated in the stromal vascular fraction (SVF) from Mir221/222^flox/y fed with HFHS chow ( Figure 2C ). We also assessed the expression of miR-221-3p and miR-222-3p during 3T3-L1 adipocyte differentiation. Both miR-221-3p and miR-222-3p continuously down-regulated after the induction of adipocyte differentiation ( Supplementary Figure 3A ). Finally, we assessed the serum concentrations of mmu-miR-221-3p and mmu-miR-222-3p by quantitative PCR and they were not altered in Mir221/222^flox/y and Mir221/222AdipoKO fed with STD and HFHS chow, suggesting serum mature forms of Mir221 and Mir222 were derived from various organs but not exclusively from adipose tissues ( Supplementary Figure 4A ). In human samples from the subjects with normal glucose tolerance (NGT, n=45) and impaired glucose tolerance (IGT, n=69), hsa-miR-221-3p and hsa-miR-222-3p demonstrated negative (R² = 0.1311, p<0.0001) and positive correlations (R2 = 0.03841, p=0.0441) with HbA1c levels, respectively ( Supplementary Figures 4B, C ).

The miRNA genes are classified as intergenic and intragenic miRNAs, and the intragenic miRNAs and related host genes share the common transcriptional regulation. The long non-coding RNA (lncRNA), namely mir-221 host gene (MIR221HG), has been reported in bovine, which inhibits the adipocyte differentiation in cultured cells (11). Thus, we performed 5’ and 3’ rapid amplification of cDNA ends (5’/3’RACE) using poly A⁺ RNA purified from epidydimal fat tissues of C57BL/6JJcl mice and cloned 1537 bp single exon lncRNA (Mir221hg, MW581002) which overlap Mir221 ( Supplementary Figure 5A ). Next, we designed primers for quantitative PCR of Mir221hg and evaluated gene expression in various tissue. Mir221hg was abundantly expressed in brain in Mir221/222^flox/y and they were upregulated in brain, heart, liver, intestine, spleen by the HFHS feeding compared with STD chow. The expression of Mir221hg was rather low in various adipose tissues and its lncRNA expression was not changed in Mir221/222AdipoKO fed with STD and HFHS chow ( Supplementary Figures 5B, C ). Although Mir221hg lncRNA may be functional as a reservoir for miR-221-3p, the role of Mir221hg lncRNA in adipose tissues was limited in our experiments.

To investigate the mechanism for the resistance to diet-induced obesity in Mir221/222AdipoKO, we attempted to identify the potential target genes for miR-221-3p and miR-222-3p. We performed mRNA profiling by GeneChip Mouse Gene 2.0 array using total RNA purified from epidydimal fat tissues of Mir221/222^flox/y fed with STD or HFHS, Mir221/222AdipoKO fed with STD or HFHS chow ( Supplementary Tables 3 – 5 ). Then, we made the list of 355 target genes commonly predicted by TargetScan, miRDB, and DIANA-microT, and their seed sequences were conserved in both human and mouse genome ( Supplementary Table 6 ). By searching upregulated genes in Mir221/222AdipoKO compared with Mir221/222^flox/y in both STD and HFHS fed groups, we identified that DNA-damage-inducible transcript 4 (Ddit4) demonstrated 1.9-fold and 2.58-fold upregulation in Mir221/222AdipoKO fed with HFHS and STD chow, respectively ( Supplementary Table 7 ). To evaluate miRNA activity of miR-221-3p and miR-222-3p by the binding to 3’-untranslated region (3’-UTR) of Ddit4, we performed luciferase reporter assay in 3T3-L1 cells. We cloned 3’-UTR regions of Ddit4 gene and prepared the wild-type reporter vector, pmirGLO-Ddit4 WT 3’-UTR, and the mutant vector, pmirGLO-Ddit4 MT 3’-UTR, in which mutagenesis was induced on the seed sequence binding site ( Figure 3A ). The luciferase reporter vectors were co-transfected with Syn-mmu-miR-221-3p or Syn-mmu-miR-222-3p or negative control siRNA in 3T3-L1 cells. As a result, reporter activity of pmirGLO-Ddit4 WT 3’-UTR were significantly inhibited in the presence of both Syn-mmu-miR-221-3p and Syn-mmu-miR-222-3p. In contrast, Syn-mmu-miR-221-3p or Syn-mmu-miR-222-3p had almost no effect on the reporter activity of pmirGLO-Ddit4 MT 3’-UTR ( Figure 3A ). Next, we checked the expression of Ddit4 in the epidydimal adipose tissues of Mir221/222^flox/y and Mir221/222AdipoKO fed with HFHS by quantitative PCR and Western blot. Although the mRNA expression of Ddit4 was increased in Mir221/222AdipoKO without statistical significance, DDIT4 protein expression was significantly upregulated in the Mir221/222AdipoKO compared with Mir221/222^flox/y ( Figure 3B ), suggesting the possible involvement of translational repression by miR-221-3p and miR-222-3p.

DDIT4 is known to activate tuberous sclerosis complex 1 (TSC1)/TSC2 complex by the inhibition of AKT (protein kinase B) and the release of 14-3-3 from TSC2 (12). It further negatively regulates mammalian target of rapamycin complex 1 (mTORC1) (13). Increased activity of the mTORC1 signaling has been associated with obesity (14), and the knockout mice for the mTORC1 downstream ribosomal protein S6 kinase (S6K) are protected against diet-induced-obesity (15). Thus, we further checked the inactivated form of p-TSC2 (Thr-1462) and activated form of p-S6K (16) in the epidydimal adipose tissues of Mir221/222^flox/y and Mir221/222AdipoKO. After overnight fasting, the mice were injected intraperitoneally with insulin and the epidydimal fat tissues were harvested at 7 min after insulin injection. In Mir221/222AdipoKO mice, both p-AKT and p-TSC2 (Thr-1462) were significantly reduced, and accordingly p-S6K was also significantly reduced ( Figures 4A, B ). p-mTOR was reduced in Mir221/222AdipoKO mice, however, it did not reach the statistically significant differences ( Figures 4A, B ).

Since mTOR has been shown to positively regulate adipogenesis and lipogenesis while inhibiting lipolysis, fatty acid oxidation and ketogenesis (17), we performed adipogenesis and glycerol assays in 3T3-L1 adipocytes. We constructed lentiviral vector expressing miRNA inhibitors (pLV-locker 221, pLV-locker 222, pLV-locker control) and miRNA mimics (pLV 221, pLV 222, pLV control). The efficiency for inhibitors and mimics was confirmed in 3T3-L1 cells at 3-7 days after the transduction ( Supplementary Figures 3B, C ). Lentiviral vectors were transduced to 3T3-L1 cells at 2 days before the induction of differentiation and cultured for 7 days. The pLV-locker 221, pLV-locker 222, and pLV-locker 221/222 did not alter the accumulation of lipid droplets demonstrated by oil red O staining measured by absorbance at 490 nm ( Figure 5A ). Similarly, pLV 221, pLV 222, and pLV221/222 only show slight increase in the accumulation of lipid droplets in 3T3-L1 cells without statistical significance ( Figure 5B ). Western blot analysis demonstrated that the expression of C/EBPα and PPARγ were not changed by pLV-locker 221 and pLV-locker 222 ( Figure 5C ), while they were significantly up-regulated by treatment with pLV 222 ( Figure 5D ). To investigate lipolysis, lentiviral vectors were transduced to the fully differentiated 3T3-L1 adipocyte and we performed glycerol assay 7 days after lentiviral transduction. pLV-locker 221/222 and pLV 221/222 did not alter the glycerol release ( Figure 5E ).