Induced Pluripotent Stem Cells from Pigs and Other Ungulate Species: An Alternative to Embryonic Stem Cells?

Authors


Author’s address (for correspondence):
RM Roberts, 240b Bond Life Sciences Center, 1201 E. Rollins Street, Columbia, MO 65211-7310, USA. E-mail: robertsrm@missouri.edu

Contents

Robust embryonic stem cell (ESC) lines from livestock species have been difficult to derive and maintain, and unlike mouse ESC, have not contributed to our ability to understand directed differentiation in vitro. Nor have such cells yet provided a simpler means than pronuclear injection to manipulate the genomes of agriculturally important species, such as cattle, sheep and pigs. Induced pluripotent stem cells (iPSC) generated by reprogramming somatic cells, such as fibroblasts, with a set of stemness genes, most usually but not exclusively POU5F1, SOX2, KLF4 and c-MYC, offer an alternative to ESC in these regards, as they exhibit a pluripotent phenotype resembling that of ESC, yet are readily generated in the laboratory. Accordingly, such cells, in association with cloning technologies, may be useful for introducing complex genetic changes into livestock, although this potential has yet to be demonstrated. Porcine iPSC may be especially valuable because the pig is a prime biomedical model for tissue transplantation. In general, iPSC from livestock, like those from humans, are of the epiblast type and depend upon FGF2 and activin/nodal signalling systems to maintain their pluripotency and growth. Recent experiments, in which newly reprogrammed porcine and bovine cells were selected on a LIF-based medium in presence of specific protein kinase inhibitors, have allowed iPSC cells of the naïve type, resembling the more amenable blastocyst-derived mouse ESC and iPSC to be isolated. However, hurdles still remain if such cells are to achieve their biotechnological promise.

Introduction

The discovery that it was possible to reprogram mouse embryonic fibroblasts to a pliable, undifferentiated state resembling that of embryonic stem cells (ESC) by ectopic expression of a combination of ‘stemness’ genes, most famously Oct4 (Pou5f1), Sox2, Klf4 and cMyc (Takahashi and Yamanaka 2006), caught the public’s imagination. Suddenly there appeared to be a means of readily generating human pluripotent cells matched to a patient that could be used clinically to replace dead, damaged and even genetically faulty cells without the concern that such donor cells would be rejected by the patient’s immune system. Such cells also appeared to bypass the ethical concerns surrounding the use of stem cell lines that originated from human embryos, as well as possible future scenarios where patient-matched cell lines would be fashioned by somatic cell nuclear transfer (SCNT), a technique that would require the use of donated eggs. The realization that induced pluripotent stem cells (iPSC) could be readily created also held out the promise of permitting pluripotent cells to be derived from animal models important in biomedical research, such as the dog and pig, where production of ‘classical’ ESCs had hitherto proved challenging, as well as from endangered species and disappearing breeds (Roberts et al. 2009; Telugu et al. 2010b). Not only had agriculturally relevant species proved to be among such ‘difficult’ subjects, also included in this group were some important mouse strains. The purpose of this review, however, is not to cover the entire field of cellular reprogramming and assess whether iPSC are living up to their promise for medical cures. Instead, we focus on a much narrower topic, namely iPSC from ungulate farm species, and begin to address questions regarding their scientific value and practical usefulness and, particularly, whether such cells can be a substitute for ESC.

Why Create iPSC from Farm Animals?

Mouse ESC revolutionized mouse genetics and made the mouse the premier mammalian model in biomedical research. Like embryonal carcinoma cells, ESC from some strains of mice could readily colonize the inner cell mass (ICM) of blastocysts and, after embryo transfer to surrogate mothers, contribute to tissues of the foetus and most importantly to the germ line (Evans and Kaufman 1981; Martin 1981). Moreover, by the mid 1980s, it was shown that such ESC could be manipulated genetically by homologous recombination (for a review of early work and the development of the technology (Capecchi 1989). Genes could be either removed, replaced with mutated copies or extra copies added, and, because ESC continue to divide more or less indefinitely, selection of mutants could be carried out before the cells entered senescence, a concern when employing primary cultures from somatic tissues, such as fibroblasts from skin. Derivation of ESC from livestock, therefore, became a priority for the biotechnology industry because it offered an attractive alternative to pro-nuclear injection as a means for creating transgenic animals, but, despite some promise (Notarianni et al. 1990, 1991; Piedrahita et al. 1990; Chen et al. 1999; Li et al. 2003), this area of research has not progressed sufficiently well to influence the course of genetic modification of ungulate livestock in any meaningful way.

Several reviews (Brevini et al. 2007; Keefer et al. 2007; Vackova et al. 2007; Talbot and Blomberg le 2008) have explained why progress on creating robust ESC lines from cattle, sheep and pigs has been so frustrating, and so the topic will not be covered here. However, one obvious reason for pursuing iPSC cell technologies, in lieu of dependable ESC, has been to provide cell lines that can be readily manipulated through introduction of locus-specific, genetic modifications. Conversely, the advent of animal cloning technologies in the late 1990s, in which a differentiated, somatic cell nucleus rather than one from a stem cell is introduced directly into an oocyte for reprogramming, has allowed animal scientists to bypass the intermediary step of chimera production and the accompanying requirement for germ-line transmission. Moreover, ESC were not necessary for the technology to be successful. Instead, simple genetic modification could be accomplished in certain kinds of somatic cell by homologous recombination, and mutants then selected and used as nuclear donors before the cells finally senesced (Prather et al. 2007; Welsh et al. 2009). Additionally, the resulting cloned animals, although few in number and frequently exhibiting abnormal phenotypes, could often be bred to yield developmentally normal progeny carrying the genetic modification. It remains to be seen whether complex genetic changes involving more than one gene can be performed on such somatic cells, thereby making iPSC less relevant to the furthering of the technology.

Nonetheless, iPSC may provide other advantages in addition to their ability to proliferate more or less indefinitely and hence survive extensive genetic manipulation. Their undifferentiated, presumed ‘embryonic’ state may allow their chromatin to be more readily re-programmed within the oocyte cytoplasm than chromatin from differentiated somatic cells, hence providing enhanced clone survival, although this hypothesis has not been rigorously tested, even in mice. Such clones may also carry fewer harmful epigenetic modifications than ones derived from somatic cells and hence exhibit more ‘normal’ phenotypes. However, such optimistic outcomes are not assured and will require careful testing. Moreover, iPSC, despite their similarity to ESC, may carry an epigenetic memory themselves which reflects the somatic cell type of origin.

Even if iPSC prove to be less useful than anticipated in genetic modification of livestock, they could have considerable utility in regenerative medicine. The general public has long been promised the benefits of stem cells as cures, yet in the USA and western Europe, at least, there is only a handful of cases where the application of ESC to tissue regeneration has reached the clinic, in large part because the safety of the technology is far from assured and the means for delivering the cells to the site of tissue injury and establishing functional grafts not established. Large animal models, especially the pig because of its similarities in physiology and anatomy to humans (Prather et al. 2003; Piedrahita and Mir 2004; Brandl et al. 2007), could play a major role in establishing the safety of stem cell medicine and developing the needed surgical and related technologies that are necessary before human trials take place. In these respects, the mouse is no match for a pig. The approach envisaged would be to establish primary cultures from a readily available tissue acquired soon after birth, such as skin of umbilical cord, and generate iPSC cells by using a suitable set of reprogramming genes. Then, provided that protocols can be developed to drive the iPSC along particular pathways of differentiation, the cells could be used as grafts on the same animal from which they had been derived, thereby minimizing the chances of immunological rejection.

Generation of iPSC from Domesticated Ungulates

Lines of iPSC have been successfully generated from pigs, cattle, sheep and goats (Table 1). A more detailed discussion follows.

Table 1.   Phenotypes of iPSC from four domestic ungulate species Thumbnail image of

Pigs

The first iPSC generated from a livestock species were from the pig (Esteban et al. 2009; Ezashi et al. 2009; Wu et al. 2009). In each case, essentially comparable approaches were employed, namely reprogramming fibroblasts with the standard ‘Yamanaka factors’ in either lentivirus or retrovirus vectors and with either human or mouse genes. The colonies that formed resembled human ESC and iPSC in morphology (Fig. 1) rather than analogous stem cells from mouse and expressed the expected, endogenous marker genes of pluripotent stem cells. Like the human cells they resembled, they were dependent upon the growth factor FGF2 and also activin/nodal signalling (Alberio et al. 2010), rather than LIF to maintain their pluripotent state. These pig iPSC were therefore of the epiblast rather than the naïve type (see below). This property also meant that the cells possess one of the major shortcomings of human ESC and iPSC, namely that during routine passage they cannot be readily dissociated into single cells without causing major cell death. Accordingly, they are usually propagated as clumps, a feature that could complicate the creation of chimeras, although there have been reports of production of chimeric piglets from such epiblast-type lines (West et al. 2010, 2011).

Figure 1.

 Porcine induced pluripotent stem cells (iPSC) analogous to primed/epiblast type (b) typical of human embryonic stem cell (ESC) and naïve-type cells resembling mouse ESC (c). The scheme summarizes the reprogramming of somatic cells to iPSC. By using a combination of four genes, OCT4 (POU5F1), SOX2, KLF4 and c-MYC (OSKM), plus two additional (optional) factors LIN28 and NANOG (LN), porcine foetal fibroblasts (a) could be reprogrammed successfully. Following the transduction of the cells, the resultant cells were selected either in a medium containing FGF2 (b) or one containing LIF in combination with two synthetic inhibitors, CHIR99021 and kenpaullone (2i) (right), to give rise to two distinct types of pluripotent stem cell colonies. Culture in the FGF medium produces flat, ‘primed’, epiblast-type colonies (c), culture in the LIF-based medium generates small, rounded, ‘naïve’-type iPSC (right). (d, e) show that each type of colony stains positively for OCT4, SOX2 and NANOG by immunocytochemistry (Scale, 100 μm).

The porcine iPSC isolated by all groups (Esteban et al. 2009; Ezashi et al. 2009; Wu et al. 2009; West et al. 2010; Montserrat et al. 2011) were clearly pluripotent, as evidenced by their ability to differentiate into tissue types reflective of the three germ layers, ectoderm, endoderm and mesoderm, within either embryoid bodies or teratomas. The porcine ID6 line (Ezashi et al. 2009) has been used to generate retinal rod precursor cells that had the capacity to populate the retina (Zhou et al. 2011), but less success has been achieved in driving ID6 cells differentially along the endoderm or mesoderm lineages by protocols used with human ESC (Ezashi and Roberts, unpublished data), although multiple laboratories, including ours, have shown that iPSC readily form such derivatives spontaneously within embryoid bodies. It remains unclear whether or not the continued expression of re-programming genes (which may not be a universal phenomenon for all porcine iPSC lines) complicates directed differentiation along specific lineages, or if the protocols and reagents used have not been optimized for pig.

Since the original reports in 2009, several additional porcine iPSC cell lines have been described, including ones created with plasmid vectors (Telugu et al. 2010a). Other than some discrepancies relating to whether or not the lines express the surface marker SSEA3 and the relative strengths of expression of SSEA1 and SSEA4, the general phenotypes of the epiblast-type lines appear to be quite similar (Table 1). Most of these lines have been of the epiblast type, but, by employing a LIF-based medium and protein kinase inhibitors (see below), it has become possible to select for lines that resemble blastocyst-derived, mouse ESC with a naïve phenotype (Telugu et al. 2010a, 2011) (Table 1). In one of these studies (Telugu et al. 2011), our laboratory produced naïve-type ESC from the ICM of porcine blastocysts with the rationale that such cells might carry less of an epigenetic memory than reprogrammed somatic cells, a hypothesis that remains to be demonstrated.

Sheep, goats and cattle

Induced pluripotent stem cells have been successfully generated from ovine (Bao et al. 2011; Li et al. 2011; Liu et al. 2012; Sartori et al. 2012), caprine (Ren et al. 2011) and bovine somatic cells (Han et al. 2011; Huang et al. 2011; Sumer et al. 2011). All the lines could differentiate into cell types representing the three germ layers. One ovine line met the ultimate standard for pluripotency, namely successful incorporation into chimeras as judged by a genomic PCR detection for two of the transgenes (Pou5f1 and c-Myc) in lambs born after embryo transfer (Sartori et al. 2012), while another contributed to the ICM after being injected into early embryos (Liu et al. 2012).

The standard combination of reprogramming genes (POU5F1, SOX2, KLF4, c-MYC) was used to re-program the ovine and caprine fibroblasts, although one group used mouse rather than human genes (Sartori et al. 2012). From the reports of reprogramming mouse somatic cells in particular, it is quite clear that various combinations of genes can be employed with varying efficiencies, although POU5F1 is usually but not invariably included (Montserrat et al. 2011). These data also illustrate that the origin of the re-programming genes, human vs mouse, does not seem critical, at least in the case of ovine cells.

In our hands, it has proved difficult to reprogram bovine cells (skin and foetal lung fibroblasts, cells from amniotic fluid) by the conventional approach described above (Ezashi, Telugu, and Roberts, unpublished data). However, success has been achieved in other laboratories by including additional genes in the re-programming mixture, for example, Lin28 and Nanog (Han et al. 2011), or adding just NANOG (Sumer et al. 2011). Two reports have also stressed the importance of employing bovine rather than heterologous genes, for example, human or mouse, for efficient reprogramming (Han et al. 2011; Huang et al. 2011), and one group successfully used a plasmid-based vectors rather ones based on a virus (Huang et al. 2011). The observation that bovine sequences might be necessary to achieve efficient re-programming of bovine cells is interesting, but should be viewed in the light of the fact that bird cells have been reprogrammed successfully with human factors (Lu et al. 2012) and frog cells with mouse factors (Vivien et al. 2012).

Naïve vs Epiblast-Type Stem Cells

Human ESC have generally been derived from outgrowths from the ICM of blastocysts, but their colonies are very different in morphology from mouse blastocyst-derived ESC. Human ESC colonies have a flattened rather than raised morphology and require a different set of growth factors (FGF2 and activin) than their mouse counterparts, which depend upon LIF/STAT3 signalling for maintenance of their pluripotency (Hall et al. 2009). More recently, a different kind of mouse ESC was produced from the epiblast of gastrulation-stage mouse embryos (Brons et al. 2007; Tesar et al. 2007). These cells closely resemble hESC in colony morphology, requirement for activin A and FGF2, and lack of dependence on LIF, and have been called ‘primed’ or epiblast stem cells (EpiSC) (Nichols and Smith 2009; Hanna et al. 2010b). Although mouse EpiSC and naïve ESC can be inter-converted (Bao et al. 2009; Greber et al. 2010; Hanna et al. 2010a; Xu et al. 2010), it is clear that the two stemness states are supported by different signalling networks and differentiate differently in response to directing stimuli.

Naïve cells may have certain advantages over EpiSC as laboratory models. They generally proliferate more rapidly, resist spontaneous differentiation, and can be dissociated into single cells without loss of viability. Moreover, naïve cells are readily cryopreserved without great loss of plating efficiency. They also appear to be far more competent for producing germ-line chimeras than EpiSC (Brons et al. 2007; Tesar et al. 2007). Hence, a recent focus of iPSC research with livestock species has been to produce naïve-type cells that possess these beneficial features, by making use of strategies successful for isolating naïve-type ESC from rat (Buehr et al. 2008; Li et al. 2008) and ‘difficult’ strains of mouse (Hanna et al. 2009). The tactic has been to select cells after transduction with reprogramming vectors on a LIF-based medium containing protein kinase inhibitors that inhibit some and activate other signalling pathways. For example, in our attempts to reprogram pig cells, we employed CHIR99021 (CH) to activate the WNT signalling pathway and to bypass MYC function, and kenpaullone (KP) to replace endogenous KLF4, which is poorly expressed in epiblast-type porcine iPSC. By contrast, Huang et al. (2011) incorporated CH and PD0325901 to inhibit ERK-mediated pathways in their selection for bovine iPSC. Both approaches were successful, but further work is needed to determine whether or not these naïve-type cells have the features that have made mouse ESC of the naïve type so useful.

Concluding Comments

Although iPSC lines have been successfully from a range of ungulate farm species, their promise for biotechnology and biomedicine remains to be demonstrated. In part, future direction will depend upon whether funding agencies and relevant industrial groups are willing to underwrite work on large animal models. In the shorter term, however, it will be essential to demonstrate whether or not iPSC cells have advantages over somatic cells in transgenic technologies. Are their genomes more readily manipulated and do such cells offer superior clonability than somatic cells such as fibroblasts? A second priority must be to demonstrate that iPSC can be induced to differentiate in vitro along prescribed pathways in a reproducible manner. Such research will likely require knowing more about the reprogramming process itself, the minimal means needed to establish pluripotency and whether the cells retain an epigenetic memory of their progenitors. Most importantly, the cell lines should not express the reprogramming genes used to create them as, in all probability, their expression can interfere with differentiation. Finally, we should not assume that the recipes used to maintain human and mouse ESC, especially the growth factors, are necessarily optimal for iPSC from a pig or cow.

Acknowledgements

This project was supported through grants from the National Institutes of Health 1R01HD067759 and by Agriculture and Food Research Initiative Competitive Grant no. 2011-67015-20070 from the USDA National Institute of Food and Agriculture.

Conflicts of interest

None of the authors have any conflicts of interest to declare.

Author contributions

All authors contributed to writing the review.

Ancillary