Reprogramming Roadblocks Are System Dependent

Abstract

Since the first generation of induced pluripotent stem cells (iPSCs), several reprogramming systems have been used to study its molecular mechanisms. However, the system of choice largely affects the reprogramming efficiency, influencing our view on the mechanisms. Here, we demonstrate that reprogramming triggered by less efficient polycistronic reprogramming cassettes not only highlights mesenchymal-to-epithelial transition (MET) as a roadblock but also faces more severe difficulties to attain a pluripotent state even post-MET. In contrast, more efficient cassettes can reprogram both wild-type and Nanog(-/-) fibroblasts with comparable efficiencies, routes, and kinetics, unlike the less efficient reprogramming systems. Moreover, we attribute a previously reported variation in the N terminus of KLF4 as a dominant factor underlying these critical differences. Our data establish that some reprogramming roadblocks are system dependent, highlighting the need to pursue mechanistic studies with close attention to the systems to better understand reprogramming.

Graphical abstract

Figure 1

PB Reprogramming with Different Polycistronic Cassettes (A) Structures of PB transposons carrying different polycistronic cassettes used in this study. (B) Numbers of Nanog-GFP^+/− colonies 15 days after reprogramming using the PB MKOS, OSKM, STEMCCA, or OKMS transposons. Error bars represent SD; n = 3 independent experiments. (C) CD44, ICAM1, and Nanog-GFP expression changes in PB reprogramming with the MKOS or OKMS cassette. Red indicates Nanog-GFP⁻, and green indicates Nanog-GFP⁺. Gates 1, 2, and 3 indicate ICAM1^low/CD44^high, ICAM1^low/CD44^low, and ICAM1^high/CD44^low, respectively, as indicated in the right panel.

Figure 2

TNG MKOS/OKMS MEF Reprogramming System (A) The Sp3 locus targeting scheme and Southern blot analysis. The green boxes represent Sp3 exons 1–4 from the right side. The red triangle in the third intron indicates the PB transposon integration site identified in the D6s4B5 iPSC line. SacI/SphI double-genome digestion yielded WT 15 kb and targeted 10-kb fragments detected by the probe indicated as a red bar. pA, poly(A) signal. (B) Number of Nanog-GFP⁺/GFP⁻ colonies on day 15 of TNG MKOS/OKMS reprogramming. Error bars represent SD; n = 3 independent experiments. (C) Whole-well merged images of mOrange (red) and Nanog-GFP (green) on day 15 of TNG MKOS/OKMS reprogramming. Scale bar, 7 mm. (D) Tracking images of a typical TNG MKOS or OKMS reprogramming colony from days 5 to 15. Scale bar, 500 μm. (E) Tg cell numbers during TNG MKOS/OKMS reprogramming. Error bars represent SD; n = 3 independent experiments. (F) Chimeric mice generated with TNG MKOS or OKMS iPSC lines.

Figure 3

Inefficient Reprogramming Progression of OKMS Reprogramming Intermediates (A) E-CAD and Nanog-GFP expression changes during TNG MKOS/OKMS reprogramming. Red indicates E-CAD⁻Nanog-GFP⁻, white indicates E-CAD⁺Nanog-GFP⁻, and green indicates E-CAD⁺Nanog-GFP⁺. (B) CD44 and ICAM1 expression changes during TNG MKOS/OKMS reprogramming with E-CAD, Nanog-GFP expression color codes in (A). (C) Flow cytometry analysis of sorted day-10 E-CAD^−/+ 2NG− (Nanog-GFP⁻ CD44⁻ ICAM1⁻), 3NG⁻ (Nanog-GFP⁻ CD44⁻ ICAM1⁺), and 3NG⁺ (Nanog-GFP⁺ CD44⁻ ICAM1⁺) cells after a 24-hr culture. dox, doxycycline. (D) E-CAD^−/+ 2NG⁻, 3NG⁻, and 3NG⁺ (Nanog-GFP⁺ CD44⁻ ICAM1⁺) cells on day 10 were seeded at clonal density, and Nanog-GFP⁺ iPSC colonies were counted after 10 days of further culture. The graph depicts the relative Nanog-GFP⁺ CFA compared to that of MKOS 3NG⁺ cells. Error bars represent SD; n = 3 independent experiments.

Figure 4

Distinct Gene Expression Profiles of MKOS/OKMS Reprogramming Intermediates (A) Hierarchical clustering of replicate averages with all genes. (B) PCA. Red or blue dots represent cells with MKOS or OKMS cassettes, respectively. Black dots represent cells without the reprogramming cassettes. (C) Expression heatmaps of MKOS/OKMS reprogramming with hierarchical clustering using DEGs, which were grouped to five clusters with distinct expression dynamics. (D) A chord diagram demonstrating three cross-classified DEG groups between MKOS A and OKMS B (MA_OB), MKOS B and OKMS A (MB_OA), and MKOS D and OKMS E (MD_OE). (E) Whole-transcriptome scatterplots highlighting the pluripotency genes (upper panels) and other transcription regulators (lower panels) identified in the MD_OE DEGs. The gray diagonal lines represent 1.5-fold differences in the expression levels.

Figure 5

Klf4_L Restores Surface Marker and Nanog-GFP Expression Patterns during Reprogramming (A) E-CAD upregulation in PB reprogramming with the MKOS, OSKM, OKMS, OK⁺⁹MS, STEMCCA, and STEMCCA⁺⁹ cassettes. (B) CD44, ICAM1, and Nanog-GFP expression in MKOS, OSKM, OKMS, OK⁺⁹MS, STEMCCA, and STEMCCA⁺⁹ reprogramming. Red indicates Nanog-GFP⁻, and green indicates Nanog-GFP⁺.

Figure 6

Reprogramming Cassette-Dependent Efficient Nanog Null MEF Reprogramming (A) A strategy for Nanog null TNG MKOS MEF reprogramming. The WT Nanog locus in TNG MKOS ES cells was converted to a Frt floxed allele via gene targeting, resulting in Nanog^G/fl ES cells. The remaining Nanog coding sequence was excised by transient expression of FLP, resulting in Nanog null MKOS ESCs (Nanog^G/G). Blue boxes 1–4 indicate exons 1–4. FLP recombination target sites FRT (F) and F3 in the cell lines are indicated in orange and yellow, respectively. (B) CD44/ICAM1 expression changes during Nanog^G/+ (TNG) and Nanog^G/G (null) MKOS MEF reprogramming. Red indicates Nanog-GFP⁻, and green indicates Nanog-GFP⁺. (C) Nanog-GFP⁺ CFA of 3NG⁺ cells sorted on day 10 of Nanog^G/+ and Nanog^G/G MKOS MEF reprogramming. Cells were cultured in the presence of doxycycline for 10 days after the sorting. Error bars represent SD; n = 3 independent experiments. ns, not significant compared to ESCs, by Student’s t test. (D) The absence of Nanog protein in Nanog^G/G MKOS ESCs and Nanog^G/G iPSC lines was confirmed by western blotting. (E) qRT-PCR analysis of pluripotency genes in Nanog^G/G iPSC lines in comparison to a WT ESC line. (F) Chimeric mice generated with Nanog^G/G iPSC cell lines. (G) Nanog^G/+ and Nanog^G/G MKOS MEFs were reprogrammed in the presence or absence of VitC (+VitC and -VitC, respectively). Reprogramming efficiency on day 15 was calculated as shown in Figure S2B. Error bars represent SD; n = 4–6 independent experiments. (H) Whole well images of Nanog^G/+ and Nanog^G/G MKOS MEF reprogramming in the presence or absence of VitC (+VitC and -VitC, respectively). Scale bar, 7 mm. (I) A strategy for Nanog null MEF reprogramming with various reprogramming cassettes with Klf4_L (red) or Klf4_S (blue). WT and Nanog null mixed MEFs isolated from E12.5 chimeric embryos, generated with a Nanog null ESC line BT12 constitutively expressing GFP, were reprogrammed via PB transposons. dox, doxycycline. (J) Numbers of DPPA4⁺ colonies on day 15 were scored, and reprogramming efficiencies of Nanog null cells against WT cells were shown. Error bars represent SD; n = 3. ^∗∗p < 0.01; ^∗∗∗p < 0.001, ns, not significant compared to +VitC or -VitC MKOS reprogramming, respectively, by Student’s t test. (K) Increased reprogramming efficiency by Nanog overexpression during TNG MKOS or OKMS reprogramming. Error bars represent SD; n = 3 independent experiments. See also Figure S2B.