Highly sensitive and non-disruptive detection of residual undifferentiated cells by measuring miRNAs in culture supernatant | Scientific Reports -…

Extraction method for miRNA from culture supernatants

We hypothesised that measuring miRNAs in the culture supernatant would enable the development of a highly sensitive and non-disruptive quality testing method (Fig.1). First, we optimised an extraction method for miRNAs specifically from the culture supernatants. Using phosphate-buffered saline (PBS) spiked with a known amount of synthesised cel-miR-54 as a sample, a commercially available small RNA extraction column could be used to extract approximately 20% of spiked miRNA (Fig.2a). On the contrary, we developed an miRNA extraction method specialised for culture supernatants that could recover approximately 60% of the spiked miRNA (Fig.2a).

Schematic representation of the concept of our novel method for detecting residual undifferentiated cells using miRNAs in culture supernatants.

Highly efficient method to extract and detect miRNAs in culture supernatants was developed. (a) The recovery rate of spiked cel-miR-54 in PBS using our method and a commercially available kit, which is used to extract miRNAs using columns. (b) The recovery rate of spiked cel-miR-54 in various culture media using our method. (cg) Calibration curves for each miRNA, plotted using the measured Ct values from 100 copies or 50 copies to 107 copies, and the theoretical copy number (log). The slope and R2 value for each calibration curve are presented within the graph. (h) Level of each miRNA detected in the culture supernatant of iPSCs and RPE cells. UD: no miRNA detected; error bar=+3SD.

We confirmed the miRNA extraction efficiency in different types of media: STEM-Fit and mTeSR, which are generally used as an iPSC culture medium; Dulbeccos modified Eagle medium (DMEM), Roswell Park Memorial Institute medium (RPMI), and Leibovits medium (L-15 medium), which are commonly used for other cell cultures; Yssel's serum-free T cell medium, which is a medium for T cells; and RtEBM, which is a medium for retinal pigment epithelium (RPE) cells. Each medium was spiked with synthetic cel-miR-54 to determine miRNA recovery rate. The results confirmed that our method extracted miRNA with a high recovery rate of more than approximately 70% from these seven media (Fig.2b).

The miR-302 family members and miR-367 (miR-302/367), which have been reported to be expressed abundantly in undifferentiated cells, such as iPSCs and ESCs15, were examined as potential markers for detecting undifferentiated cells. miR-371, miR-372, and miR-373 are also known as iPSC and ESC markers; however, their expression was lower than miR-302/367 expression (Supplementary Fig.1). The sequences of miR-302/367 are shown in Table 1. We constructed quantitative reverse-transcription PCR (qRT-PCR) assay systems for these miRNAs and confirmed by following the MIQE guidelines19. The linear dynamic ranges were from 50 to 107 copies for miR-302b and from 100 to 107 copies for the other four miRNAs (Fig.2cg). The PCR efficiency ranged from 97.5 to 112.7%. These trials were performed three times for each miRNA, and we confirmed that the lowest concentrations were detected at 100%. By comparing the expression of miR-302/367 in culture supernatants of iPSCs and RPE cells, miR-302b, miR-302c, and miR-367 were detected specifically in the iPSC supernatant (Fig.2h). miR-302b was not detected in RPE cells but was the most abundant in iPSCs. As miR-302b was the best marker for undifferentiated cells, we decided to carry out all further experiments using it.

Next, we confirmed the extent of release of the undifferentiated cell-specific miRNAs in the culture supernatant. Using iPSCs and RPE cells, we compared the expression levels of LIN28 and Oct4 (also known as Pou5f1), which are well-known undifferentiated cell markers, and miR-302b in cells and the culture supernatant. We compared the amount of nucleic acids in 105 iPSCs or RPE cells, and in 2mL of their culture supernatants. Approximately 1.4108 copies of miR-302b were expressed in 105 iPSCs (Fig.3a). The LIN28 and Oct4 levels in iPSCs were approximately 8,000 times higher than those in RPE cells (Fig.3b,c). In iPSCs, the LIN28 and Oct4 levels detected in the supernatant were 0.3% and 5.7% of intracellular expression, respectively. In contrast, approximately 6.0107 copies of miR-302b were released into the iPSC supernatant (Fig.3a), which was half of the intracellular expression level. LIN28 and Oct4 were also detected at negligible levels in RPE cells and their culture supernatant, but miR-302b was not detected in them (Fig.3ac).

miRNA was abundantly secreted into the culture supernatant. The expression level of (a) miR-302b, (b) LIN28, and (c) Oct4 in 105 iPSCs and 105 RPE cells and in 2mL of their culture supernatants. For LIN28 and Oct4, the vertical axis is the relative expression level as the expression level in RPE cells was 1. The level of miRNAs in the supernatant was corrected by multiplying the detected value with 20, because only 100L in 2mL of the culture supernatants was used as a sample. As the cells were extracted from whole cells, such a correction was not performed. UD: no miRNA detected; error bar=+3SD.

To use miRNAs in culture supernatants as targets for quality testing, properties such as stability during storage are important. We quantified the change in miRNA levels in the culture supernatant after changing the medium to determine the optimal timing for collecting the culture supernatant. The level of miR-302b in the culture supernatant reached saturation between 7 and 24h after changing the medium (Fig.4a). Next, to determine the storage conditions for the culture supernatant, the collected media were stored at various temperatures, and the changes in the level of miRNA in the medium were noted. At 25C or higher, the levels of miRNAs decreased to less than 10% of the original level within 3days (Fig.4b). At 4C, 70% of miRNAs were detected after 1week and 30% after 2weeks (Fig.4c). At 30C, the miRNA levels of more than 80% of the level on the day of collection could be detected after 2weeks (Fig.4b). Furthermore, at 80C, miRNA in the medium was stably detected even after 3months (Fig.4c). Therefore, we decided to collect the culture media 24h after the medium change and store them at 80C.

miRNAs in the culture supernatant were found to be stable. (a) Changes in the level of miR-302b in the culture supernatant observed up to 24h after changing the medium. The graph shows data from three independent experiments. (b) The collected culture medium was stored at various temperatures for 2weeks, and the change in the level of miR-302b was noted. Day 0 is the day of collection of the culture medium. (c) When the collected medium was stored at 80C, the level of miR-302b detected before and that after 3months of storage were comparable. Hence, 80C was used in all further experiments. Error bar=+3SD.

We examined whether the level of miR-302b in the culture supernatant reflects a change in the cell state. We checked the behaviour of miR-302b in the culture supernatant during the differentiation of iPSCs to NPCs. The neural marker TUJ1 (also known as TUBB3) was expressed on day 15 after the induction of differentiation, confirming that iPSCs were induced to differentiate into NPCs (Fig.5a,b). During this process, miR-302b in the culture supernatant continually decreased, reaching 1/100 of the pre-differentiation level on day 15 and 1/1000 on day 20 (Fig.5c).

miR-302b expression level in the culture supernatant decreased as iPSCs were induced to differentiate to NPCs. The images of cells (a) before inducing differentiation, that is, iPSCs, and (b) on day 15 of the iPSC-derived NPC-induction process. Blue channel represents the nuclei and green is TUJ1. Scale bar is 50m. (c) The level of miR-302b in the culture supernatant continually decreased during the induction of iPSCs to NPC differentiation. Error bar=3SD.

To quantify the efficiency of detecting undifferentiated cells by miR-302b in culture supernatants, we established a residual undifferentiated cell model in which 106 RPE cells were spiked with an arbitrary amount of iPSCs. To determine the accuracy of the model, we constructed a model using only iPSCs that maintain an undifferentiated state. Because iPSCs are prone to differentiate at the colony margin during culture20, it is possible that undifferentiated iPSCs are contaminated with iPSCs that have lost their undifferentiated state. Therefore, we confirmed the expression of Oct4 in each cell using an imaging flow cytometer, and found that the proportion of Oct4-positive cells was 98.2% (Fig.6a). We then constructed our accurate residual undifferentiated cell model by measuring miR-302b in the supernatant, 0.001%, that is, 10 iPSCs could be detected in the RPE cell background (Fig.6b). In contrast, the measurement of LIN28 in the cells could only detect iPSCs up to 0.01% of RPE cells (Fig.6c). A non-disruptive undifferentiated cell detection method for detecting H-type3 (Fuca1-2Gal1-3GaINAc), a mucin-like o-glycan on the surface of iPSCs, with rBC2LCN lectin, has been reported21. We further measured H-type3 in the culture supernatant using the same samples, and achieved a performance of 0.1% (Fig.6d). Furthermore, we checked the detection sensitivity of miR-302b in the supernatant in Clonetics human hepatocyte cell system (liver cell), human umbilical vein endothelial cell (HUVEC), and mesenchymal stem cell (MSC) backgrounds using mixed supernatants. Specifically, the culture supernatants of iPSCs and differentiated cells were mixed at an arbitrary ratio according to the number of each cells. In these mixed supernatants, the miR-302b level was below the detection limit in 0% and 0.001% iPSC samples, and the detection performance was 0.01% in all three backgrounds (Fig.6e).

Sensitivity of detecting undifferentiated cells by measuring miR-302b in culture supernatant was 0.001%. (a) Oct4-positive rate of iPSCs used to develop the residual undifferentiated cell model was found to be 98.2%. (b) The detection sensitivity of miR-302b in the culture supernatant, (c) LIN28 in the cells, and (d) Fuca1-2Gal1-3GaINAc in the culture supernatant was compared in the same residual undifferentiated cell model in which 106 RPE cells were spiked with the appropriate number of iPSCs. (e) Detection sensitivity of undifferentiated cells measured by detecting miR-302b in the culture supernatant of liver cells, HUVECs, and MSCs. Error bar=+3SD. UD: no miRNA detected. *p value<0.01, compared with 0%, Students t-test.

A large number of cells is required for the transplantation of heart or liver cells generated from iPSCs, which are currently under clinical research. Nucleic acid extraction and PCR are inhibited when a target gene is detected in many cells. Therefore, we investigated the possibility of detecting miRNAs in the culture supernatant under scaled-up conditions. HCT116 cells (107) spiked with 1% (105 cells) or 10% (106 cells) of iPSCs were seeded in a 10-cm dish, and miR-302b was extracted and measured from the cells and supernatant, respectively. All detached cells and 100L of the supernatant from 10mL of the medium were used as samples. The positive control sample (PC) consisted of 106 iPSCs, and miR-302b was detected in both cells and supernatant (Fig.7). In the sample consisting of 0% iPSC sample, that is, only 107 HCT116 cells, miR-302b was not detected in the cells or supernatant (Fig.7). However, in the samples with HCT116 cells spiked with iPSCs, miR-302b was detected in the culture supernatant but not in the cells (Fig.7).

Measuring miR-302b in the culture supernatant allows the detection of iPSCs in a large number of cells. (a) 107 HCT116 cells were spiked with iPSCs, and miR-302b was measured in the culture supernatant. (b) 107 HCT116 cells were spiked with iPSCs and miR-302b was measured in the cells. PC: positive control, which is the culture condition of only 106 iPS cells. UD: no miRNA detected. Error bar=+3SD.

Read the rest here:
Highly sensitive and non-disruptive detection of residual undifferentiated cells by measuring miRNAs in culture supernatant | Scientific Reports -...