Efficient derivation of transgene-free porcine induced pluripotent stem cells enables in vitro modeling of species-specific developmental timing

Abstract

Sus scrofa domesticus (pig) has served as a superb large mammalian model for biomedical studies because of its comparable physiology and organ size to humans. The derivation of transgene-free porcine induced pluripotent stem cells (PiPSCs) will, therefore, benefit the development of porcine-specific models for regenerative biology and its medical applications. In the past, this effort has been hampered by a lack of understanding of the signaling milieu that stabilizes the porcine pluripotent state in vitro. Here, we report that transgene-free PiPSCs can be efficiently derived from porcine fibroblasts by episomal vectors along with microRNA-302/367 using optimized protocols tailored for this species. PiPSCs can be differentiated into derivatives representing the primary germ layers in vitro and can form teratomas in immunocompromised mice. Furthermore, the transgene-free PiPSCs preserve intrinsic species-specific developmental timing in culture, known as developmental allochrony. This is demonstrated by establishing a porcine in vitro segmentation clock model that, for the first time, displays a specific periodicity at ∼3.7 h, a timescale recapitulating in vivo porcine somitogenesis. We conclude that the transgene-free PiPSCs can serve as a powerful tool for modeling development and disease and developing transplantation strategies. We also anticipate that they will provide insights into conserved and unique features on the regulations of mammalian pluripotency and developmental timing mechanisms.

Keywords: HES7; Notch signaling; cellular reprogramming; developmental allochrony; segmentation clock; species-specific developmental timing; transgene-free porcine iPSC.

Figure 1

Establishment of PiPSCs free of transgenes (A) Schematics of an optimized reprogramming strategy to establish transgene-free PiPSCs including representative images of porcine fibroblasts and a phase and alkaline phosphatase (AP)-staining of established PiPSC colonies. All scale bars, 100 μm. (B) Representative images of porcine fibroblasts during ∼1 month of reprogramming to establish PiPSC colonies. All scale bars, 100 μm. (C) (Left) Representative AP-staining of primary PiPSC colonies at ∼4 weeks post-electroporation with or without miRNA-302/367 in a 6-well plate. (Right) Quantification of primary PiPSC colonies at ∼4 weeks post-electroporation with or without miRNA-302/367. ^∗∗∗∗ p < 0.0001, Student’s t-test. Data are presented as mean ± SD. Each data point represents an AP-staining count from one well of a six-well plate, collected from nine independent reprogramming experiments. n = 26 from both groups. (D) (Top and bottom) Genomic PCR and qPCR detecting episomal EBNA1 and Orip sequences across 7 PiPSC clonal samples. E, episomal plasmid only control; F, porcine fibroblast; P, positive control, a partially reprogrammed PiPSCs with detectable episomal plasmids. (E) RT-PCR assays detecting transgene-specific (T-) or endogenous gene expression across 7 PiPSC clonal samples. (F) qRT-PCR assays detecting endogenous pluripotency marker gene expressions across nine independently established PiPSC clonal lines. (G) Immunofluorescence assays for POU5F1, NANOG, and SOX2 of a representative undifferentiated PiPSCs colony. Co-staining with Hoechst 33342. All scale bars, 50 μm. (H) PCA of RNA-seq experiments by comparing porcine fibroblasts and three established PiPSC clonal lines. Replicate samples are indicated by the same color. Total of three independent clones in triplicates. (I) Heatmap of gene expression from RNA-seq experiment in (G), representing fibroblasts and three clonal PiPSC lines as indicated. Selected markers were chosen to represent the pluripotent and somatic cell states. Expression values (normalized expect counts [nECs]) are shown and scaled from minimum to maximum expression per gene row, indicated as a horizontal bar.

Figure 2

Comparative analysis and differentiation of transgene-free PiPSCs (A) PCA of RNA-seq experiments by comparing porcine, human, and mouse PSCs collected from various studies as indicated (also see Methods and RNA-seq data analysis). (B) MA-plot comparing RNA-seq data from pig ESC and iPSC. Key pluripotency markers are indicated on the MA plot. (C) (Top) Selected GO terms enriched from the top 1,204 upregulated genes in pig iPSC compared with pig ESCs. (Bottom) Selected GO terms from the top 1,912 genes down-regulated genes in pig iPSC compared with pig ESCs. Full enriched GO terms are listed in Table S1. (D) Immunofluorescence assay for in vitro differentiation from PiPSCs. (i) Primitive streak differentiation staining for T/Brachyury. (ii) DE differentiation staining for SOX17. (iii) Cardiomyocyte differentiation staining for TNNT2. (iv) NPC differentiation staining for PAX6. All scale bars, 50 μm. All images were counter-stained with Hoechst 33342 as indicated. (E) A schematic of RNA-seq experiments profiling lineage-specific progenitors differentiated from PiPSCs. Total of 14 samples from 5 cell states were profiled by RNA-seq. (F) PCA of RNA-seq experiments described in (B). Replicate samples are indicated by the same color. (G) Heatmap of gene expression from RNA-seq experiment in (B), cell types are indicated as: Cardio, cardiomyocytes; ECs, endothelial cells; fibs, fibroblasts. Selected markers were chosen to represent each of the cell states. Expression values (normalized expect counts, nECs) are shown and scaled from minimum to maximum expression per gene row, indicated as a horizontal bar. (H) Selected histological features collected from three teratomas derived from three clonal PiPSC lines indicated in Figure S2I. Identified cell types and tissues are indicated. Black arrows in (viii) indicate pigmented cells. Scale bars, 200 μm.

Figure 3

Comparative characterizations of porcine and human segmentation clock models in vitro (A) Immunofluorescence assay of undifferentiated and PSM cells derived from a clonal PiPSC line (top) or hESCs (H1 hESC, bottom), co-stained with NANOG, TBX6, and Hoechst 33342. All scale bars, 100 μm. (B) Heatmap of somite differentiation RNA-seq experiments comparing data from PiPSCs (left, from 4 cell states in triplicate samples, total 12 samples) and hESCs (H1, right, from 4 cell states in triplicate samples, total 12 samples). Selected markers were chosen to represent each of the cell states. PS, primitive streak. Expression values (normalized expect counts [nECs]) are shown and scaled from minimum to maximum expression per gene row, indicated as horizontal bars. (C and D) Representative daily FACS data on pHES7- and hHES7-reporter activation during differentiation to PSM state. (E) Representative oscillation profile of pHES7-reporter gene oscillation (blue). All oscillation data are presented as the mean ± SD (light blue). (Right) Porcine quantification of peak-to-peak time (blue) or valley-to-valley time (red) from three independent experiments. Number of wells of data collected for quantifications is indicated. (F) Representative oscillation profile of hHES7-reporter gene oscillation (black). All oscillation data are presented as the mean ± SD (gray). (Right) Quantification of human clock peak-to-peak time (blue) or valley-to-valley time (red) from three independent experiments. Number of wells of data collected for quantifications are indicated by at least three independent experiments. (G) pHES7-reporter oscillation time when co-cultured with either porcine (left, control, n = 46) or human (right, testing, n = 50) PSM cells without the reporter. N reports the number of wells of data collected for quantifications from three independent experiments. (H) hHES7-reporter oscillation time when co-cultured with either human (left, control, n = 50) or porcine (right, testing, n = 40), PSM cells without the reporter. N reports the number of wells of data collected for quantifications from three independent experiments. (I) Representative decay difference between porcine (blue) or human (gray) HES7 measured by the decay of HES7-Nanoluc reporter activity after applying cycloheximide. (Right) Quantification of the rate of decay comparing porcine (n = 55) with human (n = 50) reporter activities. N reports the number of wells of data collected for quantifications from eight independent experiments. ^∗ p < 0.05, Student’s t-test. (J) A summary of the results presented here. PiPSC free of transgenes can be efficiently derived from our optimized protocols. PiPSCs are capable of multilineage differentiation. Importantly, PiPSC-derived PSM cells display specific segmentation clock periodicity.