Figure 2.

miR-181a Sensor Enriches for TICs in Ovarian Cancer

(A and B) miR-181a sensor-based sorting of mCherry^high and mCherry^low cells from OCI-P5X cells and HEYA8 cells (A, left and B, left) (sensor-sorted cells were analyzed three passages after sorting for reporter levels, and we did not observe changes in the reporter fluorescence activity even after 20 passages). Real-time PCR (A, right and B, right) showing ∼4-fold difference in miR-181a expression in mCherry sorted cells.

(C) In vivo tumor initiation showing increased tumor formation by miR-181a^high primary HGSOC cells as compared with no tumors formed by miR-181a^low primary HGSOC cells at 100,000 cells (day 93) and 1,000 cells (day 121).

(D) In vivo tumor initiation showing increased tumor formation by miR-181a^high HEYA8 cells (10,000 cells) (day 35).

(E and F) In vivo LDA tumor-initiation assay (E) and ELDA analysis (F) showing increased tumor-initiating cell frequency (∼10-fold) in vivo in miR-181a^high HEYA8 cells (day 28).

(G and H) Asymmetric and symmetric division of miR-181a sensor-sorted cells: top 10% miR-181a^high and miR-181a^low HEYA8 cells were sorted into single-cell-capture microfluidic devices and their growth was monitored daily for 15 days. Representative photomicrographs (G) of miR-181a^low/mCherry^high cell divisions showing these cells were only observed to divide to yield two miR-181a^low/mCherry^high cells. In contrast, miR-181a^high/mCherry^low cells divided both symmetrically to yield other mCherry-negative cells and asymmetrically to yield mCherry dim cells. (H) Summary of all divisions observed after 4 days of growth.

^∗∗p < 0.005, ^∗∗∗p < 0.005.