You are hereDecember 13, 2011 | Pluripotent Stem Cells
A microRNA-Based System for Selecting and Maintaining the Pluripotent State in Human Induced Pluripotent Stem Cells
From Stem Cells
By Stuart P. Atkinson
The ability to monitor the homogeneity of pluripotent cell cultures and their differentiated derivatives is potentially useful if such cells are to be made applicable to cellular therapy. Another use for such monitoring systems is in the generation of pure induced pluripotent stem cultures (iPSCs) containing only fully reprogrammed pluripotent cell types. At present, identification of “true” iPSCs relies on monitoring morphological criteria and is time consuming, even for well-trained operators. A potentially effective system to allow the purification of fully pluripotent cells has been posited by the group of Vania Broccoli at the San Raffaele Scientific Institute, Milan, Italy, which takes advantage of the current knowledge of microRNAs (miRNAs) and their role in pluripotency and differentiation. Building on previous work (Brown et al and Kelly et al), the group describe a reporter system based on Let7a and mir-292 miRNA-mediated regulation for selecting and maintaining the pluripotent state of human iPSCs (hiPSCs) (Di Stefano et al).
Let7a is a member of the Let7 family, which are expressed in human embryonic stem cells (hESCs) but not processed into their mature form due to LIN28-mediated binding and repression. Upon differentiation of hESCs and the associated downregulation of LIN28, Let7 family members are processed into their mature form and can then mediate target repression/degradation. Therefore, the Let7a target sequence (let7aT) fused to green fluorescent protein (GFP) should specifically mark hESCs, while Let7a should mediate the degradation of the target sequence and extinguish GFP in differentiating cells. miR-292 is a member of the miR290-295 cluster, highly expressed in pluripotent cells and largely undetectable in differentiated cells and so miR-292 target sequences (292T) should allow expression of an associated reporter in somatic cells but not in pluripotent stem cells. To test this hypothesis, mouse embryonic stem cells (mESCs) and mouse embryonic fibroblasts (MEFs) were transduced with lentiviral vectors containing GFP cDNA with four tandem repeat sequences perfectly complementary to the two miRNAs (let7aT or 292T) in the 3´-untranslated region (3´-UTR). Accordingly, GFP expression was suppressed in mESC-292T-GFP cells and MEF-Let7aT-GFP cells and upon differentiation of mESC-Let7aT-GFP cells into cardiomyocytes and neurons. Analysis in human cells demonstrated that GFP was absent in let7aT-GFP-transduced IMR90 fetal fibroblast cells and neural stem cells, while Let7aT-GFP-transduced hESCs robustly expressed GFP. The Let7aT-GFP reporter was then assayed in iPSC lines and it was shown that in an hESC-like iPSC culture, 90% of cells were SSEA1+ and 50% of these were GFP+, while another iPSC line that was less hESC-like and mainly constituted of proliferating single cells contained only 15% SSEA1+ cells with only 3% of these being GFP+, suggesting that reporter expression is associated with pluripotent cells. Further proof of this was shown upon the transduction of an Oct4-GFP mouse iPSC (miPSC) line with a Let7aT-red fluorescence protein (RFP) construct and the complete co-expression of GFP and RFP, and the co-localisation of GFP with OCT4 and NANOG positive cells in Let7aT-GFP transduced hiPSCs.
The relative usefulness of the miRNA vectors in tracking hiPSC formation during reprogramming was assayed using a bidirectional vector expressing Leta7T-GFP and NGFR (L7G construct). L7G was then transduced into IMR90 fibroblasts and primary fibroblasts derived from skin biopsies of Parkinson’s disease and Rett syndrome patients carrying LRRK2, CDKL5 and MeCP2 mutations. These were sorted for high nerve growth factor receptor (NGFR) expression to obtain a homogenous population of cells which were then reprogrammed using OSKM (OCT4, SOX2, KLF4 and MYC). During reprogramming of patient-derived adult fibroblasts, GFP was expressed from day 14 onwards in a small fraction of the emerging colonies, concomitant with TRA-1-60, but preceded by SSEA3 and alkaline phosphatase (AP) expression 4 and 6 days earlier, respectively. Over days 16-24, 78% of the TRA-1-60+ and 66% of the SEEA3+ colonies were GFP+ suggesting that the transgene expression displayed the most selective expression pattern. At the final stages of reprogramming, NANOG and GFP displayed almost complete co-expression and, interestingly, a large fraction of AP+ colonies were GFP-, even when displaying an ESC-like morphology. Such colonies lacked TRA-1-60 and SSEA4 expression and failed to expand into stable cell lines. GFP+ clones however co-expressed pluripotency markers and were able to generate stable iPSC lines with an overall efficiency of 88%.
To further understand the relationship between Let7aT-GFP expression and pluripotency, SSEA4+GFPhi and SSEA4+GFPlo populations were sorted from hiPSCs. GFPhi cells had elevated LIN28 (with lower pre-let7a levels and similar pri-let7a levels), NANOG, REX1 (ZFP42), OCT4, and DPPA4 expression compared to the GFPlo fraction. This suggests that the GFPlo fraction could represent a partially reprogrammed population or a fraction of hiPSCs prone to spontaneous differentiation, although as the cells were sorted at a relatively late passage (>35), this heterogeneity is a feature of well-established, mature iPSC cultures. Embryoid bodies (EBs) and teratomas were then generated from hiPSCs to analyse Let7aT-GFP expression during multilineage differentiation and, as predicted, differentiated iPSCs and teratoma cells completely lacked any distinguishable GFP expression while multilineage differentiation was permitted. Further analysis of the hiPSC-EBs demonstrated that at day 3, GFP was reduced but detectable, similar to NANOG, SOX2 and OCT4, while at day 18 no cell expressed NANOG or OCT4, with only a few cells at the outermost layer of the EB expressing GFP. Analysis also showed a more rapid silencing of REX1 and NANOG than for LIN28 at early stages, while miRNA analysis demonstrated that mature levels of let7a increased upon EB differentiation, while the pluripotency-associated miR-302a decreased. To allow the potential use of Let7 in the maintenance of homogenous pluripotent cultures, a vector containing GFP and a neomycin resistance gene (L7GN) was infected into miPSCs. G418 treatment allowed successful purification of the SSEA1+GFP+ population with selective loss of the GFP- fraction from a self-renewing population and, importantly, a partially differentiated population.
Overall, the Let7a-mediated reporter system appears to be efficient, effective and sensitive enough to lend itself as a valuable tool with which to select for cells acquiring a full pluripotent state during reprogramming, by facilitating and accelerating patient-specific iPSC generation and expansion, and to maintain homogenous cultures of pluripotent cells by getting rid of contaminating differentiated cells. Further evolution of this technology could allow the production of a miRNA-suicide gene construct to rid pluripotent colonies of differentiating cells or to remove potentially tumourigenic cells from differentiated cultures destined for therapeutic use.
Di Stefano B, Maffioletti SM, Gentner B et al.
A microRNA-Based System for Selecting and Maintaining the Pluripotent State in Human Induced Pluripotent Stem Cells.
Stem Cells. 2011 Nov;29(11):1684-95.
Brown BD, Venneri MA, Zingale A et al.
Endogenous microRNA regulation suppresses transgene expression in hematopoietic lineages and enables stable gene transfer.
Nat Med. 2006 May;12(5):585-91.
Kelly EJ, Hadac EM, Greiner S, Russell SJ
Engineering microRNA responsiveness to decrease virus pathogenicity.
Nat Med. 2008 Nov;14(11):1278-83.