To verify the existence of miRNAs in human breast milk, we extracted total RNA from human breast milk. miRNAs were detected in each individual at concentrations ranging from 9.7 ng/ml to 228.2 ng/ml. The samples contained a substantial amount of RNA, but no or only very low amounts of ribosomal RNA (18S and 28S) (Figure 1a). Furthermore, large amounts of small RNA (<300 nucleotides) were detected in the milk.

Next we performed a miRNA microarray to show the existence of miRNAs in the milk. Using microarray analysis, 281 of 723 known human miRNAs were detected. Interestingly, several immune-related miRNAs were abundant in the milk (Figure 1b): miR-155, a regulator of T- and B-cell maturation and the innate immune response; miR-181a and miR-181b, regulators of B-cell differentiation and CD4+ T-cell selection; miR-17 and miR-92 cluster: a ubiquitous regulator of B-cell, T-cell and monocyte development, miR-125b, a negative regulator of tumor necrosis factor-α production, activation and sensitivity; miR-146b, a negative regulator of the innate immune response; miR-223, a regulator of neutrophil proliferation and activation; and let-7i, a regulator of Toll-like receptor 4 expression in human cholangiocytes. By contrast, T- and B-cell regulatory miR-150 was not detected, and several tissue-specific miRNAs, such as miR-122 (liver), miR-216, miR-217 (pancreas) and miR-142-5p and miR-142-3p (hematopoietic cells), were barely detectable (Figure 1b). Furthermore, the expression pattern of miRNA within different breast milk samples from the same mother did not differ greatly with time after birth (Figure 1c, d). Between individuals, however, there was more variation. Notably, we detected higher expression of several immune system-related miRNAs from human breast milk in the first 6 months, which is the period before infants receive complementary food (Figure 2).

The existence of RNase in body fluids is already known 13 , and thus there should be no intact RNA present. The presence of miRNAs suggests that these miRNA species are resistant to RNase digestion. To prove this possibility, human breast milk was treated with RNase A/T.

Strikingly, treatment with RNase had hardly any effect on human breast milk miRNAs (Figure 3a). By contrast, the profile of total RNA from that milk and exogenously added synthetic miRNAs were degraded by the same treatment (data not shown). These data clearly demonstrate that human breast milk miRNAs are resistant to RNase digestion.

Their stability was further studied after treatment under harsh conditions including freeze-thaw cycles and low pH. The samples were analyzed by quantitative reverse transcription (qRT)-PCR analysis.

Results of qRT-PCR analysis of miRNAs in human breast milk samples subjected to harsh conditions were not significantly different from samples not subjected to these conditions. As shown in Figure 3b, human breast milk miRNAs were resistant to several freeze-thaw cycles. Moreover, human breast milk miRNAs were stable when the milk was treated for 1 h in an acidic (pH 1) solution (Figure 3c).

Observation of the stability of human breast milk miRNAs in vitro suggests the possibility that they are contained within microvesicles or exosomes 14 . To verify this, we isolated the CD63-positive exosome fraction and investigated miRNA expression.

As shown in Figure 3d, miR-181a and miR-17, which we detected by microarray analysis, were also detected in the CD63-positive exosome fraction. However, it is still possible that other miRNA protection mechanisms, such as an apoptotic body, may protect miRNA from harsh conditions.

It was previously reported that miRNAs are also detected in serum samples 11 12 13 . To investigate differences in miRNA expression in different body fluids, we analyzed expression of several immune-regulated miRNAs in human serum and human breast milk.

Interestingly, the granulocyte-regulated miRNA, miR-223, is the most strongly expressed miRNA in normal human plasma, but its expression in breast milk is low compared with that in serum (Figure 4) 14 . On the other hand, breast milk is rich in miR-181 and miR-155, and these were detected at similar levels in serum. Furthermore, miRNA microarray analysis revealed that expression of miR-146a, which is high in plasma, was not found in breast milk, whereas miR-146b, which is not abundant in plasma was abundantly expressed in breast milk (Figure 1b).

All the women gave their signed informed consent to participate. The study was approved by Dr. Tadao Ishii, Manager of Morinaga Co., Ltd and the company's ethics committee (the Ethical Committee of Functional Food Creation).

Human breast milk samples were collected from eight women enrolled in a breastfeeding study at Morinaga Milk Industry Co., Ltd. Human breast milk samples were collected when the infant were aged between 4 days and 11 months (see Table 1). The milk was collected and then the collected samples totaling 50-100 ml were put into in storage bags. All samples were stored at - 80°C until analyzed.

We chose the samples given <6 months after birth and samples given between 6 and 12 months after birth. Cells and large debris within the breast milk were removed by centrifugation at 2,000 × g for 10 min twice; the supernatant was then centrifuged at 12,000 × g for 30 min to remove cellular debris. The clear supernatants were used for the analysis.

Total RNA from breast milk was extracted using a mirVana miRNA isolation kit (Ambion, Austin, TX, USA). Breast milk was thawed on ice, diluted with two volumes of mirVana Lysis/Binding Solution, mixed thoroughly by vortex for 30 s and incubated for 5 min. Then 1/10 volumes of miRNA homogenate additive was added, mixed thoroughly by vortex for 30 s and incubated on ice for 10 min. An equal volume of acid/phenol/chloroform (Ambion) was then added to each aliquot. The resulting solutions were mixed by vortex for 1 min and spun for 10 min at 10,000 × g. The resulting aqueous volume was mixed thoroughly with 1.25 volumes of 100% molecular-grade ethanol and passed through a mirVana column in sequential 700 μl aliquots. The column was washed according to the manufacturer's protocol, and RNA was eluted in nuclease-free water at 95°C. RNA extraction from the serum was performed using the same method.

To detect the expression of miRNAs in human breast milk, 70 ng of total RNA was labeled and hybridized using a Human microRNA Microarray Kit (Agilent Technologies) according to the manufacturer's protocol (Protocol for Use with Agilent Microrna Microarrays Version 1.5). Hybridization signals were detected by a DNA microarray scanner (Agilent Technologies), and the scanned images were analyzed using Agilent Feature Extraction software.

To confirm that the miRNA is resistant to RNase digestion, human breast milk was treated with 10 U/ml RNase A and 400 U/ml RNase T1 (Ambion) for 60 min at 37°C. After these treatments, the RNA was extracted from the milk as described above.

To investigate the stability of miRNAs in human breast milk, the milk was subjected to three freeze-thaw cycles of at -20°C or treated for 3 h in a low pH solution (pH 1). miRNA levels were assessed by TaqMan qRT-PCR.

Exosomes in human breast milk were specifically isolated by magnetic beads, using anti-CD63 antibody (BD, Erembodgem, Belgium). Human breast milk (0.3 ml) was incubated with anti-CD63 antibody (BD) or mouse IgG1 (Sigma-Aldrich, St Louis, MO, USA) coupled to magnetic microbeads (50 μl). These were mixed and incubated for 16 h at room temperature. The magnetic immune complexes were washed four times with 500 μl of PBS, then the RNA was extracted as described above.

Clear supernatants from human breast milk were centrifuged at 100,000 × g for two hours, then washed in phosphate-buffered saline (PBS) and pelleted by ultracentrifugation (100,000 × g). The pellet was diluted in PBS. Resuspended exosomes were fixed in 1% glutaraldehyde in PBS (pH 7.4). The samples were stained for 10 min with 1% uranyl acetate. Excess fluid was removed with a piece of Whatman filter paper. All transmission electron micrographs were obtained using JEM1220 electron microscopy at 120 kv.