The study utilized a transgenic (Tg) mouse (DMP-1 DTR Tg) with targeted expression of diphtheria toxin receptor (DTR) under the promoter of DMP-1 (17). The DMP-1 DTR Tg mouse was backcrossed for more than nine generations onto a C57BL/6 background prior to use. The WT and Tg littermate mice (age, 15 weeks, body weight, ~30 g) were injected with a single dose of diphtheria toxin (DT; 20 µg/kg, intraperitoneally; Sigma-Aldrich, St. Louis, MO, USA), and then plasma was harvested 3 weeks later (at 18 weeks). Mice were housed under a 12-h light/dark cycle at a temperature of 23–25°C, humidity between 50–70% and were fed a normal diet. All animal experiments were approved by the Animal Care and Use Committee of Kobe University (Kobe, Japan), and were conducted according to the Kobe University Animal Experimentation Regulations.

Exosomes were isolated from the plasma using the Total Exosome Isolation kit (from plasma) (Invitrogen; Thermo Fisher Scientific, Inc., Waltham, MA, USA) according to the manufacturer's instructions. Briefly, 0.2 volume of the total exosome isolation reagent was added to the plasma and vortexed to mix. Samples were incubated overnight at 4°C, and subsequently centrifuged at 15,000 × g for 1 h at 4°C. The supernatants were discarded and the exosome pellets were resuspended in 100 µl phosphate-buffered saline (PBS).

Total RNA containing the small RNA fraction was extracted using a mirVana miRNA isolation kit (Ambion; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. This protocol effectively recovers both mRNA and miRNA. Subsequently, the concentration of the total RNA samples was evaluated using a NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific, Inc.). The integrity of the total RNA samples was also analyzed using the Agilent 2100 Bioanalyzer Lab-on-a-Chip instrument system (Agilent Technologies, Santa Clara, CA, USA) with the RNA 6000 Nano Chip (Applied Biosystems; Thermo Fisher Scientific, Inc.).

Microarray analyses were performed using the 3D-Gene miRNA microarray platform (Toray Industries, Inc., Kamakura, Japan). Briefly, total RNA containing the small RNA fraction was labeled with a 3D-Gene miRNA labeling kit (Toray Industries, Inc.), and hybridized at 32°C for 16 h on the 3D-Gene chip. The microarray was then scanned, and the obtained images were numerated using a 3D-Gene scanner 3000 (Toray Industries, Inc.). The microarray images obtained were analyzed using Genepix Pro 4.0 software (Molecular Devices, Sunnyvale, CA, USA). Differences in the total fluorescence intensity between arrays were adjusted by global normalization. The mean values for duplicate microarrays were calculated and used for comparison between the groups. When the difference in relative miRNA expression between the two groups was >2.0-fold, this was defined as a change in the expression. Subsequently, the expression level of each miRNA was globally normalized using the background-subtracted signal intensity of the entire miRNAs in each microarray. The mean values from duplicate microarrays were calculated and used for comparison between the groups.

The immortalized mouse osteocytic MLO-Y4 cell line was provided by Dr Lynda Bonewald (University of Missouri-Kansas City, School of Dentistry, Kansas City, MO, USA) (18). MLO-Y4 cells were cultured in α-modified minimum essential medium (Sigma-Aldrich) containing 10% fetal bovine serum (SAFC Bioscience, Inc., Lenexa, KS, USA) with 100 µg/ml kanamycin (Meiji, Ltd., Tokyo, Japan) and maintained in plates or flasks coated with 0.15 mg/ml rat tail collagen type I (BD Biosciences, Bedford, MA, USA) at 37°C in a humidified atmosphere of 5% CO₂ in air. In addition, the mouse stromal ST2 cell line was obtained from the RIKEN Cell Bank (Tsukuba, Japan), and cells were cultured as described previously (19).

At 3 days after reaching 100% confluence, culture medium was harvested and exosomes were isolated using the Total Exosome Isolation Reagent (from cell culture media) (Invitrogen; Thermo Fisher Scientific, Inc.) according to the manufacturer's instructions. Briefly, in order to remove cellular debris, the cell culture medium was centrifuged at 2,000 × g for 20 min, and the supernatant was transferred to a new microfuge tube. Subsequently, 0.5 volumes of reagent were added to the supernatant and vortexed to mix. Samples were incubated overnight at 4°C, and then centrifuged at 15,000 × g for 1 h at 4°C. The supernatants were discarded and each pellet was resuspended in 100 µl PBS.

RNA (1 µg) was converted into a cDNA library using a TruSeq Small RNA Library Preparation kit (cat. no. RS-200-0012; Illumina, Inc., San Diego, CA, USA) according the manufacturer's protocol. Briefly, 5′ and 3′ adaptors were ligated, followed by reverse transcription using SuperScript II reverse transcriptase (Invitrogen; cat. no. 18064014) and library enrichment by PCR amplification. The PCR conditions were as follows: Initial denaturation at 98°C for 30 sec, followed by 13 cycles of 98°C for 10 sec, 60°C for 30 sec, 72°C for 15 sec, and a final extension step at 72°C for 10 min. Following enrichment, the cDNA library was purified and run on a 6% polyacrylamide gel, and then the appropriate band (18–30 nt in size) was excised and eluted. The purified cDNA library was run on a HiSeq 2500 ultra-high-throughput sequencing system (Illumina, Inc.) with read length of 101 nt, and the resulting reads were analyzed.

In order to perform real-time image analysis and base calling, the HiSeq Control Software (Illumina, Inc.) were used on HiSeq 2500 (Illumina, Inc.). Raw readings, passed through a chastity filter, were initially extracted from the FASTQ files. and the Cutadapt application (https://cutadapt.readthedocs.io/en/stable/) was used to remove the adaptor sequences from each sequence read. Any sequences with <10 nt after trimming were not used. The prepared sequences were aligned using the BLAST+ tool (version 2.2.29+) downloaded via FTP from ftp://ftp.ncbi.nlm.nih.gov/blast/db/against mouse miRNA sequences miRBase (Release 21; http://www.mirbase.org/).

To elucidate whether exosomes from osteocytes circulate in the peripheral blood, a Tg mouse model with targeted expression of DTR under the control of the DMP-1 promoter was used (17). WT and Tg littermate mice (15-week-old) were injected with DT. After 3 weeks, the DT-injected mice showed a comparable lacuno-canalicular interstitial fluid space to that of WT mice; however, there was a marked reduction in the osteocyte network, and thuse these mice were denoted as OL (5,20). In the initial attempt to identify differentially expressed miRNAs between the plasma exosomes of OL and WT mice, exosomes were isolated from plasma samples, and then total RNA was extracted from the exosome preparations. The quality and quantity of these RNA isolates were determined using a bioanalyzer. In exosomal RNA derived from the OL and WT plasma samples, bioanalyzer profiles indicated that these RNAs lack detectable quantities of 18S and 28S ribosomal RNA, indicating that significant amounts of RNA are present (data not shown).

In order to investigate the miRNA content of these exosomes, the expression of miRNA was profiled using a 3D-Gene miRNA microarray platform that can detect ~1,300 known miRNAs. The normalized data were globally normalized per array, such that the median of the signal intensity was adjusted to 25. The number of expressed (detected) miRNAs present at significant levels (global normalization value of >100) was determined. Among the expressed miRNAs in the plasma exosomes, the expression level of 30 miRNAs was downregulated (log₂ value; fold-change <0.5) and that of another 30 miRNAs was upregulated (log₂ value >1.5) in OL mice compared with the expression in WT mice (Table I). Among the downregulated miRNAs in OL mouse plasma, the relative expression levels of miR-3473a, miR-3473b, miR-3473e, miR-5128, miR-6244, miR-6239, miR-5132-5p, miR-705, miR-208a-5p miR-3104-5p, miR-1224-5p and miR-5621-5p were −1.0 or lower (log₂ value) compared with the WT mice (Table I). These results indicated that the expression levels of certain miRNAs were altered in exosomes derived from the plasma of OL mice.

In the present study, the MLO-Y4 osteocyte-like cell line was used as the osteocyte model for several reasons: i) These cells are isolated from weight-bearing bone and are a well-characterized osteocyte model that has been intensively studied since 1997 (18); and ii) MLO-Y4 cells express early osteocyte markers in vivo, such as E11, CD44 and low levels of the mature osteocyte marker sclerostin (3). By contrast, the ST2 cell line is a bone marrow-derived stromal cell line (21). ST2 cells cultured with ascorbic acid exhibit characteristics typical of osteoblasts, including the formation of mineralized nodules, indicating that ST2 cells are pre-osteoblastic stromal cells (22). Therefore, ST2 cells were used as control non-osteocytic cells in the present study.

In order to identify differentially-expressed miRNAs in exosomes obtained from MLO-Y4 and ST2 cells, exosomes were isolated from culture supernatants, and the RNA was isolated. Bioanalyzer profiles indicated that these exosomal RNAs lacked detectable amounts of ribosomal RNA, whereas total cellular RNAs contained ribosomal 18s and 28s RNAs and mRNAs as major components, indicating that significant amounts of small RNAs were present (data not shown). Using miRNA microarray analysis, the number of miRNAs expressed at significant levels in the exosomes prepared from MLO-Y4 and ST2 cells was detected. Analysis of these miRNA profiles revealed that 53 miRNAs were enriched at significantly higher levels (log₂ value >2.5) in MLO-Y4 cells compared with their levels in ST2 cells (Table II). Furthermore, the miRNA expression profile between the exosomal and cellular fractions of MLO-Y4 cells was compared (Table II). Taken together, these data demonstrated that characterized miRNAs were present in the MLO-Y4 exosomes, and 71.7% (38/53) of these miRNAs were enriched in the exosomes compared with their expression in host cells.

To determine whether exosomes derived from osteocytes are circulating in the blood, the miRNA expression levels of exosomes were analyzed and compared between OL mouse plasma and MLO-Y4 cells. Downregulated and upregulated miRNAs in OL mice, and relative miRNAs expression levels (MLO-Y4 exosomes/ST2 exosomes and MLO-Y4 exosomes/MLO-Y4 cells) are shown in Tables III and IV, respectively. With the exception of miR-208a-5p, a total of 12 miRNAs downregulated (log₂ value <-1) in the OL mouse plasma were expressed at higher levels in the MLO-Y4 exosomes compared with the levels in ST2 exosomes (Table III). By contrast, 10 upregulated (log₂ value >2) enriched miRNAs were undetectable in both exosomal and cellular fractions of MLO-Y4 cells with the exception of miR-92a-3p (Table IV). These results suggest that these exosomal miRNAs downregulated in OL mouse plasma may be derived from exosomes secreted by osteocytes.

The microarray method, relying on sequence hybridization to appropriately-designed annealing probes, can only detect annotated miRNAs, as well as immature miRNAs, such as pre- or pri-miRNAs, which include the same sequences that would be detected in mature miRNAs. Sequencing-based technology reveals the entire repertoire of expressed small RNAs (23). In order to analyze the abundance of miRNAs and to identify the miRNA sequences that detect only mature miRNAs from the MLO-Y4 exosomes, a small RNA cDNA library was generated and deep sequencing was performed. A total of ~20,000,000 raw reads were obtained from the sample. Subsequent to removing adapter sequences and selecting by size differences (1–17, 18–25 and >26 nt), 9,954,039 clean reads with sizes within the ranges of 1–17, 18–25 and >26 nt were generated (Fig. 1). In the 18–25 nt size range of small RNA sequences, small RNA characterization analysis focused on searching release 21 of miRBase. Fig. 1 demonstrates a summary of the different types of small RNAs detected in the samples, according to current miRNA annotations. Of the total reads, 195,581 were annotated to mouse miRNA sequences (Fig. 1). In addition, 527 known mouse mature miRNAs were detected, as well as 400 stem-loop structures of full-length and truncated pre-miRNAs (Fig. 1). Among these detected miRNAs, 32 miRNAs were found that were enriched in exosomes derived from MLO-Y4 cells, as shown in Table II, and the read counts were up to 30 (Table V). Furthermore, among these 32 miRNAs, miR3473b and miR3473e were reduced in exosomes from OL mouse plasma (Table VI). These results indicated that MLO-Y4 exosomes contain varying expression levels of characterized miRNAs.