Induced Pluripotent Stem Cells in Non-Model Species: Applications and Challenges

1. Technical Challenges and Advances in iPSC Reprogramming for Non-Model Species

1.1. Background

The generation of iPSCs from non-model species (NMS) presents challenges that are qualitatively distinct from routine derivation in humans or mice. Unlike in these well-characterised systems - where optimised factor combinations, validated culture conditions, and robust quality control frameworks are firmly established - reprogramming in NMS must contend with fundamental biological unknowns, infrastructural deficits, and species-specific barriers that compound at every stage of the workflow. Understanding these challenges is a prerequisite for the rational design of solutions, and the field has begun to respond with an expanding toolkit of adapted methodologies. Before examining these challenges in detail, it is necessary to establish two areas of background knowledge essential for interpreting what follows: the nature of pluripotency states across vertebrates, and the distinction between integrating and non-integrative reprogramming approaches.

Pluripotency, the capacity of a single cell to differentiate into all cell types of an adult organism, is not a single static condition but a continuum of cellular states that transition through distinct phases during early mammalian development and in vitro culture. The naïve pluripotent state, exemplified by mouse ESCs with dome-shaped colonies, two active X chromosomes in females, a Klf4/Nanog/Esrrb transcriptional network, and a globally hypomethylated genome, corresponds to the pre-implantation epiblast [13,14].

The primed pluripotent state, resembling the post-implantation epiblast as seen in mouse EpiSCs and conventional human ESCs, is characterized by flat colonies, X-chromosome inactivation, a more methylated epigenome, and reduced chimera competence [13,15]. Bridging these extremes is the formative pluripotent state - a transient but experimentally capturable intermediate characterized by loss of naïve-specific features, acquisition of competence to respond to germ-layer inductive cues, and a poised epigenetic landscape [14,16]. Metastable states, which arise under certain culture conditions as a heterogeneous mixture of properties, have also been described [17], and recent comparative single-cell embryogenesis atlases have further resolved these transitions across mammals [18].

A key distinction amongst these states is their signaling requirements: naïve cells require LIF/STAT3 and dual GSK3/MEK inhibition (2i/LIF), whereas primed cells depend on FGF and Activin [13]. During reprogramming, somatic cells undergo extensive epigenetic resetting, including DNA demethylation, histone remodeling, and X-chromosome reactivation in females. These species differences are particularly consequential for NMS, where the optimal pluripotency state is often unknown and conventional culture conditions fail to stabilize cells in any defined state.

Reprogramming approaches are distinguished by whether transgenes become permanently incorporated into the host genome - a distinction developed in detail in Section 1.3.2, and introduced here only to orient the reader. Integrating methods offer high transduction efficiency but carry risks of insertional mutagenesis and genomic instability that render them unsuitable for translational, reproductive, or germline applications. Non-integrating approaches generate genomically intact, footprint-free iPSCs and are strongly preferred wherever downstream fidelity is required, though generally at lower efficiency and greater technical demand. The choice between approaches is therefore not merely technical but consequential for the downstream utility of any line established.

1.2. Challenges

The following three subsections identify the primary biological, molecular, and conceptual obstacles to iPSC generation in NMS. Together they define the problem space that the methodological advances in section 1.3 are designed to address.

1.2.1. Pluripotency State Instability and Culture Incompatibility

A primary obstacle in NMS reprogramming is pluripotency state instability. Reprogrammed cells from NMS frequently default to intermediate or poorly defined states that resist long-term maintenance under standard conditions. This is not simply a culture optimisation problem: it reflects genuine biological uncertainty about which pluripotency state - naïve, formative, or primed - is the natural ground state for a given species, and whether the states characterised in mouse and human have direct equivalents across vertebrate diversity [13,14]. The two-inhibitor (2i/LIF) conditions that stabilise naïve murine pluripotency, and the FGF/Activin conditions that support primed human ESCs, routinely fail to sustain NMS iPSCs, producing either spontaneous differentiation or growth arrest rather than stable self-renewal [19]. Commercial maintenance media such as mTeSR - optimised for human iPSCs and incorporating undefined animal-derived components - introduce a further layer of uncertainty when applied to novel species, compounding the difficulty of interpreting failed cultures.

The problem is sharpened by the relative scarcity of single-cell embryogenesis data for most NMS. In mouse and human, comparative transcriptomic atlases of preimplantation development now provide a high-resolution map of the molecular transitions through which naïve, formative, and primed states are traversed in vivo [18]. For the vast majority of NMS, no such reference exists, meaning that the target pluripotency state cannot be empirically defined and reprogramming outcomes cannot be benchmarked against an in vivo standard.

Even where the target state is provisionally assumed, reprogramming and maintenance depend on a set of empirical culture variables that are optimized for human and murine cells and have no principled starting point in a new species. Chief among these is the culture matrix. Feeder-cell choice - mouse embryonic fibroblasts (MEFs), immortalized MEFs (STO cells), species-matched feeders, or feeder-free systems [20,21]- and coating substrate - Matrigel, vitronectin, or specific laminin isoforms such as laminin-521 [22,23]- each influence attachment, survival, and the balance between self-renewal and differentiation, yet the combination that supports a given NMS cannot be predicted in advance. Feeder choice carries an additional molecular dimension beyond matrix mechanics: because feeders act in part by secreting paracrine factors such as LIF, FGF2, and TGF-β1 [24,25], a species-mismatched feeder layer may fail for cross-species signaling reasons - a distinct problem from matrix incompatibility, and one addressed separately in Section 1.2.4. Additional variables compound the search space: basal medium formulation and serum or serum-replacement source; oxygen tension [26]; passaging method (enzymatic single-cell versus mechanical or clump passaging) and its interaction with dissociation-induced apoptosis [27]; seeding density; and the concentration and timing of small-molecule pathway modulators [28]. Any of these, individually or in combination, can convert an otherwise competent reprogramming attempt into apparent failure, and none can be assumed to transfer from a model species.

The practical consequence is that, for a novel NMS with no established protocol, sequential single-variable optimization is prohibitively slow and consumes scarce and often irreplaceable primary material. A more tractable strategy is to screen multiple conditions in parallel from the outset - varying, for example, pluripotency-state regime (2i/LIF versus FGF/Activin versus intermediate formulations), feeder and matrix, and oxygen tension across a factorial or matrix design - so that the permissive region of the condition space is identified empirically rather than assumed. Where primary cell numbers are severely limiting, miniaturized or microwell-format screens allow a broader condition set to be surveyed from a fixed starting population. This exploratory posture - treating the correct conditions as unknown and to be discovered rather than inherited from model-species protocols - is, at present, the most realistic route to establishing stable lines in species for which no reference protocol exists.

1.2.2. Genomic and Molecular Limitations

Incomplete genome assemblies constitute a second major constraint. For species lacking high-quality reference genomes, the identification of pluripotency gene orthologs is imprecise, promoter analyses to guide transgene design are unreliable, and cross-species transcriptomic comparisons - essential for evaluating reprogramming fidelity - are compromised [29,30]. The problem extends to reagent availability: validated antibodies against pluripotency markers such as OCT4, SOX2, NANOG, and SSEA antigens have been developed and quality-controlled for human and mouse epitopes, and cross-reactivity in NMS is inconsistent and frequently unreported. Immunofluorescence and flow cytometric characterization of putative iPSC colonies - standard in human and murine systems - therefore cannot be assumed to be reliable in novel species without independent antibody validation.

Karyotypic instability represents an additional genomic concern. Even when stable iPSC lines are established, chromosomal abnormalities arise at elevated frequency during prolonged culture, and transgene silencing efficiency - critical for confirming that reprogramming factor expression has been extinguished and that the self-renewal program is genuinely endogenous - remains unpredictable across species [31]. Without reliable silencing, the risk persists that apparent iPSC maintenance reflects ongoing transgene activity rather than stable pluripotency.

Beyond confounding pluripotency assessment, unsilenced reprogramming transgenes actively oppose lineage commitment during directed differentiation - constitutive expression of OCT4, SOX2, KLF4, or c-MYC resists the transcriptional remodeling required for germ layer specification, meaning that transgene-active lines frequently fail at the differentiation stage regardless of their apparent pluripotency characteristics.

1.2.3. Functional Validation Challenges

A deeper conceptual problem concerns the benchmarks used to validate iPSC identity. In mice, the gold standard for genuine pluripotency is germline-competent chimera formation: injection of iPSCs into blastocysts, with subsequent contribution to all somatic tissues and the germline of the resulting animal. This benchmark has been achieved for only a handful of NMS - most notably rabbit [32] - and remains undemonstrated for the overwhelming majority of species for which iPSC generation has been claimed [7]. In its absence, published NMS iPSC lines are typically validated by surrogate criteria: alkaline phosphatase activity, pluripotency marker expression, embryoid body formation, and teratoma assays. While these are meaningful indicators, they are insufficient to exclude the possibility that cells represent partially reprogrammed intermediates rather than bona fide pluripotent stem cells. The field would benefit from agreed minimum validation standards for NMS iPSCs that are both scientifically rigorous and practically achievable given the logistical constraints of working with protected or rare species.

1.2.4. Cross-Species Incompatibility of Reprogramming Reagents

A final challenge specific to NMS, and one that connects several of the barriers above, is that the reprogramming and maintenance toolkit was optimized almost entirely in human and murine systems and cannot be assumed to transfer seamlessly to divergent taxa. Incompatibility can arise at three mechanistically distinct levels: the secreted ligand–receptor interface, the intracellular reprogramming-factor interface, and the innate sensing of the delivery vehicle. Distinguishing these matters, because each fails differently and each is addressed by a different intervention.

At the level of secreted factors, incompatibility is less pervasive than first principles would predict. The core self-renewal signaling modules - FGF, TGF-β/Activin/Nodal, and the gp130–JAK–STAT3 axis - are deeply conserved across vertebrates, and recombinant human FGF2 and Activin A support pluripotency across a broad taxonomic range, including avian and non-primate mammalian systems in which the ligand–receptor interface remains functional despite considerable evolutionary distance [33,34,35,36]. The principal exception is leukemia inhibitory factor (LIF), whose failure to sustain NMS iPSCs was noted in Section 1.2.1. LIF signaling illustrates why sequence conservation is a poor predictor of cross-species function: despite approximately 78% ligand and 76% receptor identity between mouse and human, mouse LIF binds only its cognate receptor whereas human LIF binds both, the species-specificity residing in the receptor's immunoglobulin-like domain [37,38]. Critically, this promiscuity does not generalize across taxa. In the domestic cat, only feline LIF - not the murine protein conventionally used in rodent systems - maintains iPSC pluripotency [39], a direct demonstration that a reagent validated in model species can fail outright in a non-model one and require a species-specific ortholog. Even where a species-specific factor is not strictly required, apparent LIF failure in a non-model line more often reflects pluripotency-state preference than a recognition defect: most non-rodent species examined to date are more readily stabilized in primed-type, FGF2/Activin-dependent conditions in which LIF–STAT3 signaling does not sustain self-renewal irrespective of whether the ligand is recognized. This distinction between recognition and functional response is frequently collapsed in the literature and should be resolved empirically, using the diagnostic readouts described at the end of this subsection, rather than inferred from colony outcome.

At the intracellular level, the reprogramming factors themselves display the opposite of what phylogenetic distance might suggest: they are strikingly tolerant of cross-species transfer. In the avian species examined to date, pluripotency-associated factors show functional divergence from the mammalian network, including differences involving POU5F3/POUV, SOX-family factors, and KLF-family factors [40,41]. Yet avian iPSCs have been derived using human or murine reprogramming cocktails despite more than 300 million years of divergence from mammals, indicating that the exogenous mammalian factors engage sufficiently conserved transcriptional targets to initiate reprogramming [42,43]. Incompatibility at this level is therefore quantitative rather than absolute. The canonical mammalian four factors activate part of the endogenous avian network - chicken and duck fibroblasts upregulate endogenous POU5F3 and KLF4 - but stall short of full reprogramming, failing to upregulate NANOG and LIN28A and retaining a tumorigenic propensity, until NANOG and LIN28 are supplied exogenously [44]. The appropriate response is therefore not a different class of protein but species-tailored factor identity, stoichiometry, and supplementation, discussed as an advance in Section 1.3.

The third level, innate sensing, concerns the delivery vehicle rather than any maintenance factor or reprogramming protein. Commercial Sendai virus vectors used for reprogramming are specifically engineered - through mutations in the P and L proteins - to minimise innate immune responses, and there is no strong evidence that type I interferon induction is a primary cause of poor transduction efficiency in NMS cells [45,46]. Where Sendai virus performs poorly in a novel NMS cell type, the more likely explanation is the absence of host factors required for viral entry or replication - a tropism constraint rather than an immune barrier. A second and underappreciated limitation of standard commercial Sendai virus vectors is their temperature-sensitive design: the mutations introduced to reduce immunogenicity and facilitate vector clearance also reduce viral persistence and limit the duration and level of transgene expression at physiological temperature, which can be insufficient to drive full reprogramming in cell types that reprogram slowly or require prolonged factor exposure [45]. Taken together, the likely bottlenecks for Sendai virus-based reprogramming in NMS cells are tropism incompatibility, inefficient viral replication in divergent host cell environments, and transient vector activity caused by temperature-sensitive mutations - rather than innate immune sensing per se. An alternative non-integrative platform with broader species tropism is Venezuelan equine encephalitis virus (VEEV)-based replicons, which can deliver reprogramming factors across a wider range of cell types, though here innate immune sensing is more important and requires active suppression using B18R, a vaccinia virus-derived decoy receptor for type I interferons [47]. Moving forward, a more effective strategy for NMS reprogramming may be to use persistent Sendai virus vectors lacking temperature-sensitive mutations - maintaining sustained transgene expression throughout the reprogramming window - combined with siRNA targeting of viral genes for controlled vector elimination once reprogramming is complete, and species-specific reprogramming factor combinations as discussed in Section 1.3.1 [46]. A cell line that responds poorly to standard temperature-sensitive Sendai virus should therefore prompt consideration of persistent vector formats or alternative delivery platforms rather than immediate attribution of failure to the pluripotency network or culture conditions.

Two practical points follow for non-model derivation. First, reagent incompatibility should not be assumed to be the dominant cause of reprogramming failure without pathway-level evidence to support it. Premature transgene silencing, incomplete epigenetic resetting, karyotypic instability, and culture adaptation each produce the same terminal phenotype - absence of stable colonies - and are readily misattributed to reagent non-recognition when the underlying mechanism is not interrogated. The exceptions are instructive rather than refuting: feline LIF represents a case where species-specific substitution was strictly required, and in avian species, NANOG and LIN28 supplementation identifies a genuine reagent-level bottleneck that mammalian OSKM alone cannot resolve [44]. Critically, both failures were identified because pathway-level readouts were applied - without them, the outcome would have been indistinguishable from culture failure or epigenetic stalling. Second, and following directly from this: cleanly attributing a reprogramming failure to reagent incompatibility requires a receptor-engagement or pathway-activation readout - ligand-induced phosphorylation of STAT3, SMAD2/3, or ERK, for example - rather than inference from colony morphology or count alone. In the absence of such evidence, the conclusion that a given human factor or standard maintenance medium has failed in an NMS context should be treated as a working hypothesis requiring mechanistic resolution - not as an established finding that justifies protocol switching. Substituting a species-specific ortholog, changing the growth medium, or adding supplementary factors without first identifying the level at which the current approach fails risks compounding unknown variables and forfeiting the diagnostic clarity that pathway-level readouts would otherwise provide. The primary challenges identified across Section 1.2, and the methodological responses developed in Section 1.3, are summarised in Table 1.

1.3. Advances & Responses

The following three subsections describe the methodological toolkit that has emerged in response to the challenges above - spanning factor composition, delivery method, and quality standards - and assess how far each advance has moved the field toward reliable, broadly applicable NMS reprogramming (Table 1).

1.3.1. Expanded and Species-Specific Factor Combinations

The first methodological response to NMS reprogramming challenges concerns factor composition. Canonical OSKM combinations, optimised through iterative screening in mouse and human, frequently perform poorly in NMS - not because the pluripotency network is absent but because the transcriptional wiring that OSKM engages is species-specifically configured. Three complementary strategies have emerged in response: substitution of canonical factors with species-appropriate orthologues, supplementation with auxiliary factors that address species-specific network gaps, and engineering of enhanced reprogramming factors with amplified transcriptional potency.

The relationship between factor origin and reprogramming outcome is species-specific rather than following a simple rule of homology, and the evidence base now spans multiple taxa with instructively varied outcomes. The porcine case is the most systematically documented in livestock: early piPSC lines generated with human or murine OSKM consistently failed to silence exogenous transgenes and showed insufficient activation of endogenous pluripotency markers, whereas porcine-origin OSKM activated Hippo, PI3K-Akt, and MAPK signalling pathways to levels resembling the porcine blastocyst inner cell mass [51]. In felids, cat-derived factors produced more efficient reprogramming than their human counterparts, and feline LIF - rather than the murine or human protein - is strictly required for iPSC maintenance, a case of absolute species-specificity rather than a quantitative efficiency difference [39,52]. The divergence from the mammalian pluripotency blueprint is the most structurally profound of any vertebrate group examined to date: avian ESC biology indicates that birds encode POU5F3 rather than POU5F1 as the functional OCT4 equivalent, express SOX3 rather than SOX2 as the effective SoxB1 factor, and deploy KLF3, KLF5, or KLF6 rather than KLF4 - substitutions that reflect genuine evolutionary divergence in the transcriptional architecture of the pluripotency network rather than simple sequence drift [40,41]. Whether systematically replacing mammalian OSKM components with these avian-appropriate orthologues improves reprogramming efficiency beyond what has been achieved with mammalian factors plus NANOG and LIN28 supplementation remains an open and important experimental question - one that is now technically tractable given the availability of annotated avian genomes and established avian iPSC lines as validation benchmarks.

1.3.2. Auxiliary Factor Supplementation

Where canonical OSKM engages an incomplete version of the endogenous pluripotency network, supplementary factors can bridge the gap. In avian cells, NANOG and LIN28 supplementation is required to advance reprogramming beyond the partially reprogrammed state reached by mammalian OSKM alone - chicken and duck fibroblasts upregulate endogenous POU5F3 and KLF4 under mammalian OSKM but stall without NANOG and LIN28, retaining tumorigenic propensity until these are supplied [44]. In felids and bovidae, NANOG, LIN28, and the histone demethylase KDM4A have been found to enhance or complement OSKM components, with factor requirements varying substantially across taxa [53,54]. Notably, two recent studies, currently under review, demonstrated that human pluripotency factors can reprogram porcine cells into bona fide iPSCs when combined with appropriate culture optimisation [55,56], suggesting that factor identity and protocol optimisation interact - and that species-specific factors are not always strictly required when culture conditions adequately support the endogenous network.

1.3.3. Engineered Super-Factors

Beyond substitution and supplementation, a parallel strategy involves engineering reprogramming factors themselves to be more potent transcriptional activators. The most significant advance in this direction is the development of enhanced SOX factors - principally SOX17 fusions and the so-called Super-SOX constructs - which carry amplified transactivation domains that dramatically increase reprogramming efficiency in human cells and, critically, in species where endogenous SOX engagement is partial [57,58,59]. SOX constructs, generated by fusing the SOX2 HMG domain with a potent transactivation domain derived from VPR or similar synthetic activators, achieves reprogramming in human cells with reduced factor combinations - in some formulations replacing the entire OSKM cocktail - and shows enhanced performance in cell types that are refractory to conventional OSKM, including aged cells and cells from species with divergent SOX network wiring [60,61]. In the NMS context, the principle is directly applicable: species in which endogenous SOX2 or its ortholog is poorly engaged by the exogenous human factor are strong candidates for Super-SOX-based approaches. The MyoD transactivation domain fusion strategy - in which reprogramming factors are fused to the MyoD transactivation domain to increase chromatin accessibility at pluripotency loci - represents a related engineering approach that has improved reprogramming efficiency in porcine and bovine cells [62,63]. Chemical enhancement of transcription factor activity through small molecules targeting chromatin remodelling - VPA, sodium butyrate, BIX-01294 - provides a pharmacological parallel to factor engineering that is particularly relevant for NMS where vector payload constraints limit the number of transgenes deliverable simultaneously.

1.3.4. Key Success Stories

Several milestones illustrate the cumulative impact of these approaches. First, the derivation of avian iPSCs from four endangered Japanese species - Okinawa rail, ptarmigan, fish owl, and golden eagle - using a PiggyBac system delivering seven mouse factors including NANOG and LIN28 represented the first demonstration of iPSC derivation from birds, achieved only after the factor combination was extended beyond canonical OSKM [64]. Second, the derivation of giant panda iPSCs using a fully chemically defined approach combining OCT4, SOX2, and KLF4 with the iCD3 small molecule cocktail achieved reprogramming without animal-derived components — a significant advance for a critically endangered species where reagent control and genomic fidelity are paramount [65]. Third, the generation of iPSCs from nine individuals representing the last northern white rhinoceros individuals using Sendai virus, enabling PGCLC induction and the establishment of an allelic diversity resource from a functionally extinct species, represents the clearest demonstration that non-integrative reprogramming with appropriately chosen factors can achieve conservation-grade cellular rescue from the brink of extinction [8] (Figure 2). It is also worth noting the preprint showing the derivation of Bornean orangutan iPSCs using Sendai virus reprogramming from fibroblasts, with resulting cells exhibiting primed state pluripotency characteristics which can be upgraded to naïve state with PGCLC potential by aggregation, demonstrated that delivery method and culture condition optimisation together can achieve a pluripotency state in a great ape that recapitulates the most developmentally useful form of pluripotency known in any species [66]. Taken together, these success stories share a common feature: each required deliberate species-specific adaptation of the factor combination, delivery method, or culture condition rather than direct application of human protocols - a consistent finding that supports the principle that wherever annotated genomes permit the cloning of native orthologs or the characterisation of species-specific network topology, this information should be used to guide factor design from the outset rather than treated as a secondary optimisation step.

1.3.5. Non-Integrative Delivery Methods

Integrating methods - principally retroviral and lentiviral vectors - dominated early NMS reprogramming and retain practical relevance for contexts where efficiency is the primary constraint and genomic fidelity is not. Retroviruses transduce dividing cells with high efficiency and undergo transcriptional silencing upon successful reprogramming; when starting material is limited, lentiviruses extend this to non-dividing or slowly dividing primary cells, which is advantageous when starting material is [1,67]. The liabilities are well-documented: stable genomic integration carries risk of insertional mutagenesis, incomplete silencing may sustain residual transgene expression that confounds pluripotency assessment, and integrated cassettes render cells unsuitable for translational, reproductive, or germline applications [53]. For basic research where the goal is simply to demonstrate reprogramming feasibility in a novel species, integrating vectors remain a pragmatic first-pass approach - but any line intended for downstream conservation, clinical, or agricultural use should be derived by non-integrative means from the outset.

For such applications, the shift to non-integrative delivery is not merely preferable but essential: cells intended for assisted reproduction, embryo generation, or germline use must be genomically intact, and insertional mutagenesis renders them unsuitable regardless of pluripotency characteristics [53]. Three non-integrative platforms have emerged as leading options for NMS, alongside the intermediate excisable transposon approach.

Sendai virus - a negative-sense RNA virus that replicates exclusively in the cytoplasm without nuclear entry - has become the most widely adopted non-integrative system across mammalian NMS, combining high transduction efficiency with a clean genomic footprint and commercial availability [50]. Its adoption has driven the recent expansion of integration-free iPSC generation across rhinoceros, primates, pigs, and bat species. Its principal limitation is that it was developed and optimised for human cells, and transduction efficiency in morphologically or physiologically distinct cell types from divergent species can be substantially lower.

Episomal vectors - self-replicating DNA plasmids progressively lost as cells divide - offer an alternative with a smaller technical footprint, requiring only standard transfection equipment rather than biosafety level 2 viral work. Episomal approaches have successfully generated iPSCs from giant panda and several Southeast Asian primate species [65,68], and their practical accessibility makes them particularly relevant for institutions without viral vector infrastructure. Their principal limitation is lower reprogramming efficiency relative to Sendai virus, which may be prohibitive when starting material is scarce.

Self-replicating RNA (saRNA) vectors represent an emerging non-integrative alternative that addresses one of the principal limitations of conventional synthetic mRNA delivery: the requirement for high input doses to achieve sustained transgene expression [69]. saRNA constructs encode both the reprogramming factors and a viral RNA-dependent RNA polymerase (typically derived from alphaviruses such as Venezuelan equine encephalitis virus) that amplifies the RNA intracellularly [70], producing high and prolonged transgene expression from a substantially lower initial dose without nuclear entry or genomic integration [47]. This amplification property is particularly relevant for NMS reprogramming, where transfection efficiency into primary somatic cells is often poor [71], starting material is scarce, and repeated transfection - the standard workaround for conventional mRNA delivery [72] - may not be feasible. saRNA-based reprogramming has been demonstrated in human cells and selected NMS [47], and while the approach remains less widely adopted than Sendai virus or episomal vectors, its low-dose, footprint-free profile positions it as a strong candidate for resource-limited conservation settings in which biosafety level 2 viral infrastructure is unavailable and starting cell numbers are low.

The excisable piggyBac transposon system occupies an intermediate position: it integrates into the genome during reprogramming but can theoretically be removed by transposase-mediated excision, generating transgene-free iPSCs at efficiencies comparable to viral methods [73]. It has been employed successfully in Ryukyu spiny rat and avian NMS [64,74]. Excision is not always precise, however, and residual mutations at excision sites have been documented [64,75,76], so the system should not be treated as equivalent to fully non-integrative methods in applications where absolute genomic fidelity is required.

Fully chemically defined reprogramming - avoiding both viral vectors and animal-derived components entirely - represents the most stringent approach and has been achieved in a small number of NMS, including giant panda iPSCs generated using OCT4, SOX2, and KLF4 alongside miRNA plasmids and the iCD3 small molecule cocktail [65], and porcine iPSCs under a defined medium incorporating CHIR99021, IWR-1, WH-4-023, LIF, Activin A, and FGF2 [77]. These iCD formulations represent a growing toolkit that, while currently limited to a small number of species, may be iteratively adapted as the chemical logic of pluripotency induction becomes better understood across taxa. The trajectory of NMS iPSC derivation across species and methods is summarised in Figure 2.

1.3.6. Minimum Standards and Prioritisation

Collectively, the technical challenges described above make clear that iPSC generation in NMS cannot yet be treated as a routine procedure. Each new species encountered is, in a meaningful sense, a new problem requiring bespoke solutions at the level of factor choice, delivery method, culture conditions, and validation strategy. Two practical responses to this reality are worth emphasising.

First, the field requires agreed minimum reporting and validation standards for NMS iPSC claims - specifying which pluripotency markers, functional assays, and genomic quality control steps constitute an acceptable evidence base for a published iPSC line. The ISSCR guidelines for stem cell research and clinical translation provide a framework applicable to human and model species iPSCs [78], but no equivalent standard has been formally adopted for NMS, and the heterogeneous validation depth evident across the published literature reflects this gap directly. Second, given the resource intensity of species-specific protocol development, strategic prioritisation is essential: investment should be concentrated on species that are simultaneously conservation-critical, technically tractable, and capable of generating broadly transferable methodological insights. As genomic resources improve, organoid technology matures, and non-integrative reprogramming efficiency increases across taxa, the per-species cost of iPSC derivation will decline - but in the near term, selectivity is a scientific and practical necessity.

1.3.7. Applications of iPSCs in Human and Model Species: The Baseline from Which NMS Work Departs

The trajectory toward clinical deployment has accelerated markedly: as of December 2024, 115 regulatory-approved trials were underway testing 83 hPSC-derived products in over 1,200 patients, with no generalised safety concerns reported [79]. In February 2026, Japan granted conditional marketing authorisation to the world's first two iPSC-derived therapies - ReHeart (iPSC-derived cardiomyocyte sheets for heart failure; Cuorips) and AMCHEPRY (raguneprocel, iPSC-derived dopaminergic progenitors for Parkinson's disease; Sumitomo Pharma/CiRA Foundation) - while BlueRock Therapeutics received FDA IND clearance for OpCT-001, an iPSC-derived photoreceptor therapy entering Phase I/IIa evaluation (Sumitomo Pharma; BlueRock Therapeutics). Also, autologous or HLA-matched allogeneic iPSC-derived cell products have entered early-phase trials for conditions including age-related macular degeneration, Parkinson's disease, heart failure, and cancer therapy, with demonstrated feasibility and initial safety profiles [80,81,82,83,84]. Complementing human studies, iPSCs derived from model species - including mouse, rat, pig, and non-human primates - serve as indispensable preclinical platforms enabling evaluation of graft safety, biodistribution, teratoma formation, immune compatibility, and long-term functional integration in immunocompetent, physiologically relevant whole-organism systems prior to human application [85,86].

Taken together, these milestones confirm that human iPSC-derived cell therapies have crossed from experimental biology into regulated clinical and commercial practice, establishing the safety, manufacturing, and regulatory precedent that NMS applications will eventually need to meet.

2. Conservation Applications

2.1. iPSCs as a Versatile Tool for Wildlife Genetic Rescue

The accelerating extinction and biodiversity crisis has pushed traditional conservation methods - habitat protection, anti-poaching patrols, and captive breeding - to their limits [7]. In response, a suite of biotechnologies, collectively termed genetic rescue, has emerged as a potentially transformative complement to these established approaches.

Biobanking has become a vital pillar of genetic rescue. Facilities such as the Frozen Zoo at the San Diego Zoo Wildlife Alliance store tissues, somatic cells, gametes, and embryos from thousands of individuals across hundreds of species, providing the raw biological material for downstream reproductive and biotechnological applications including Somatic Cell Nuclear Transfer (SCNT), iPSC derivation, and Assisted Reproductive Technologies (ART). The logic is straightforward: even where a species has suffered a severe genetic bottleneck, its preserved living cells and genetic material remain available for future population supplementation and diversity recovery. ART - including artificial insemination, in vitro fertilisation (IVF), and intracytoplasmic sperm injection (ICSI) - represents the most established application of such biobanked material, and in several species has already contributed to population recovery efforts.

Interspecies SCNT (iSCNT) circumvents the reliance on oocytes from threatened species by using those from closely related, more accessible surrogates. However, the approach is beset by compounding limitations. Efficiency is low even between closely matched species and has been confirmed in only a small number of species combinations - among them gaur cloned using domestic cattle oocytes and African wildcats using domestic cat oocytes - with failures far outnumbering successes across the broader range of attempts [87]. The use of heterologous oocytes introduces additional complications, including mitochondrial DNA incompatibility and failure of embryonic genome activation, rendering the process an unreliable pathway for most threatened species [88]. Beyond these biological constraints, iSCNT is logistically demanding and costly: it requires access to recipient oocytes and surrogate mothers that are themselves non-threatened, reproductively tractable, and available in sufficient numbers - conditions that are difficult to meet for most wildlife species, which lack a domesticated or otherwise accessible close relative that can serve in this role. Domestic cattle, pigs, and rabbits have been deployed as universal surrogate oocyte donors in some studies, but phylogenetic distance compounds already poor nuclear-cytoplasmic reprogramming efficiency, and the procedural costs - spanning hormone-stimulated oocyte retrieval, nuclear transfer, embryo culture, and surgical embryo transfer - collectively render iSCNT an impractical route for broad application in wildlife conservation.

iPSC technology reframes rather than eliminates these challenges, offering a complementary route that bypasses the most constraining dependencies of the cloning pipeline. Somatic cells from a threatened species can be reprogrammed into iPSCs using non-integrating methods, expanded indefinitely in culture, cryopreserved, and later differentiated into functional gametes or even blastoids - synthetic embryo-like structures capable of mimicking embryo development [7]. Critically, because iPSC lines can be established from many individuals, including those that died without reproducing, they are inherently more genetically representative than SCNT-based approaches, which are constrained by the availability of oocytes and reproductive surrogates. The derivation of iPSCs from nine genetically distinct Northern white rhinoceros individuals - several of whom never reproduced in life - illustrates this advantage directly, capturing allelic diversity that would otherwise have been permanently lost [8]. It is important to note that iPSC-based gamete derivation and blastoid generation remain largely proof-of-concept in most NMS, and the path from banked somatic cells to a live animal involves substantial unresolved biological and technical steps. Nevertheless, iPSCs establish a uniquely flexible starting point: a living, expandable, genomically intact cellular resource that can be revisited as the reproductive technologies required to deploy it continue to mature.

2.2. In Vitro Gametogenesis and Germline Applications

One of the most consequential long-term applications of iPSC technology in conservation is in vitro gametogenesis (IVG): the derivation of functional sperm and eggs from pluripotent stem cells [89,90,91]. IVG would enable gamete generation from deceased, sterile, or otherwise non-reproductive individuals, with the resulting gametes deployed through ART to restore genetic diversity to otherwise unrecoverable populations. The stakes are highest for species such as the Northern white rhinoceros, where only two non-reproductive females survive, and the window for conventional reproductive rescue has effectively closed [8,92].

The foundational proof-of-concept for iPSC-based IVG was established in the mouse. Hayashi and colleagues demonstrated that both spermatogonial stem cells and functional oocytes could be derived from mouse ESCs and iPSCs via an epiblast-like intermediate state, with the resulting gametes producing fertile offspring upon fertilisation and embryo transfer [89,93]. A fully in vitro pipeline - requiring no transplantation into a living animal at any stage - was subsequently achieved for oogenesis in mouse, in which PGCLCs were cocultured with reconstituted ovarian somatic cells to produce metaphase II oocytes capable of fertilisation [91]. These murine milestones established the conceptual and technical framework upon which conservation-focused IVG efforts are now being built. A critical insight from this work is that PGCLCs alone are insufficient: functional gametogenesis requires the support of gonadal somatic cell environments - either through transplantation or in vitro reconstitution - that provide the niche signals governing meiotic entry, epigenetic reprogramming, and sex-specific maturation [90]. Developing equivalent gonadal somatic cell systems for non-rodent species represents one of the field's most pressing unresolved challenges.

Progress in translating IVG to conservation-relevant NMS has been meaningful but incremental. The most significant advance to date was the robust induction of PGCLCs from both Southern and Northern white rhinoceros ESCs and iPSCs by the BioRescue consortium [94]. This was the first time PGCLCs had been generated from a large, threatened mammalian species, and required identification of SOX17 - rather than the murine PRDM14 - as the key transcriptional driver of rhinoceros PGC induction, underscoring that germline specification pathways are not fully conserved across mammals and must be characterised de novo in each taxon. Surface markers CD9 and ITGA6 were identified as conserved across both rhinoceros subspecies, providing tools for PGCLCs isolation applicable to future gamete maturation steps. These PGCLCs now represent the last remaining biological bridge between the nine cryobanked Northern white rhinoceros iPSC lines and the production of functional gametes - a pipeline that, if completed, would constitute the first recovery of a functionally extinct species through iPSC-derived reproduction.

Beyond rhinoceros, proof-of-concept germline progress has been achieved across several taxa. Female Ryukyu spiny rat (Tokudaia osimensis) iPSCs produced spermatocytes and spermatids when injected into male mouse testes - a remarkable finding given the species' unusual XO/XO sex chromosome system, which lacks a Y chromosome entirely, and one that demonstrated the flexibility of the germline specification pathway across divergent reproductive architectures [64,74]. In rhesus macaque, iPSCs were differentiated into functional spermatogonial stem cells that, upon transplantation into sterilised testes, restored spermatogenesis - providing a non-human primate model with direct relevance to both conservation and reproductive medicine [95]. PGCLCs have been generated from cynomolgus monkey ESCs and iPSCs under WNT-inhibited defined conditions, with transcriptomic profiles closely resembling embryonic PGCs in vivo [96]. For the common marmoset, an mRNA-based induction protocol generated PGCLCs and subsequent gonocyte-like cells, advancing understanding of primate germline specification in a species increasingly used as a bridge model between rodents and humans [97].

Complementing iPSC-based approaches, spermatogonial stem cells xenotransplantation strategies - in which testicular tissue biobanked from deceased animals is grafted into reproductively active recipients of a related species to restore spermatogenesis - are under active development as part of integrated conservation germline pipelines [98]. These approaches are not mutually exclusive with IVG: testicular tissue biobanking provides an immediate bridge to germline preservation while longer-term iPSC-based gametogenesis protocols are refined.

Despite this progress, the distance between current achievements and the deployment of IVG-derived gametes in conservation breeding programmes remains substantial. In vitro spermatogenesis beyond the spermatid stage - the step required to generate motile, fertilisation-competent spermatozoa without testicular transplantation - has not been achieved in any species outside the mouse [99]. Full in vitro oogenesis beyond the PGCLCs stage has similarly not been demonstrated in any NMS. Both processes depend critically on the reconstitution of species-specific gonadal somatic environments whose cellular composition, signalling requirements, and developmental timing remain largely uncharacterised outside rodents and a small number of primate species.

The most recent advances in the human IVG baseline illustrate how the gonadal somatic niche bottleneck is beginning to be approached from both sides of gametogenesis. On the female side, Conception reported in June 2026 the first generation of fully iPSC-derived human follicles containing primary oocytes that have entered meiosis, assembled within a niche reconstructed entirely from iPSC-derived granulosa-like cells (The first early human eggs from stem cells - Conception | Advancing the Future of Fertility ). On the male side, human and rhesus iPSCs have been differentiated into undifferentiated and differentiated spermatogonia and preleptotene spermatocytes using xenogeneic reconstituted testes assembled from mouse fetal testicular cells — the furthest advance in primate male IVG yet reported [100]. Neither studies has achieved gametes capable of fertilisation; but together they establish that species-authentic gonadal somatic reconstitution - whether from iPSC-derived or xenogeneic somatic sources - can support primate germ cell progression beyond the PGCLC stage, and set the technical benchmarks against which NMS reproductive biology programmes will ultimately be measured.

2.3. De-Extinction and Genome Editing: iPSCs as a Platform for Species Resurrection

De-extinction is not a single, well-defined outcome but a spectrum of biological interventions that differ fundamentally in their ambition, feasibility, and scientific validity. At one end lies functional trait restoration: the introduction of specific adaptive traits from an extinct species into a living relative, producing an animal that can perform an ecological role previously filled by the extinct taxon without recapitulating its full genome or phenotype. In the middle sits the hybrid proxy: an organism whose genome has been edited at many loci with the explicit goal of systematically approximating the extinct species, sharing a substantial proportion of its distinctive biology while remaining genetically a modified version of its living relative. At the far end is true resurrection: the recovery of a reproductively isolated population that is genomically, phenotypically, and ecologically equivalent to the extinct species - an outcome that would require not only a complete and accurate reference genome but also resolved reproductive pipelines, species-authentic developmental environments, and ultimately a self-sustaining wild population. These categories are not merely semantic. They differ in what counts as success, in the scientific rigour with which claims can be evaluated, and in the ecological and ethical frameworks they invoke. iPSC technology plays a distinct role at each position on this spectrum, as one of several editing platforms available for functional trait restoration, as the preferred multiplex-editable substrate for hybrid proxy generation where the scale of editing exceeds what zygote-based approaches can efficiently achieve, and - in the specific case of species for which viable cells were biobanked before extinction - as the reprogrammed cellular vehicle for true resurrection, through which a genomically authentic nucleus can be maintained, expanded, and deployed in iSCNT or IVG pipelines. The distance between current iPSC-based achievements and each endpoint varies enormously and conflating them produces both inflated claims and misplaced scepticism.

A critical biological constraint governs the entire spectrum and is rarely stated explicitly: true resurrection is only achievable if the intact genome of the extinct species is physically accessible in a viable cellular form - which means biobanking is not merely a conservation tool but a prerequisite for keeping true resurrection within the realm of biological possibility. Two routes exist. The first is cryopreserved gametes - sperm amenable to ICSI, or oocytes amenable to IVF - which provide direct access to the intact haploid genome without reprogramming. The second is cryopreserved somatic cells, which can be reprogrammed into iPSCs and deployed either as nuclear donors in iSCNT or as the starting material for IVG pipelines in which differentiation toward PGCLCs reconstitutes the germline from a diploid somatic source. Both routes presuppose that biobanking preceded extinction - and both are closed permanently once it has not. Any strategy that begins instead from the genome of a living relative produces an animal that is fundamentally the living relative's genome with modifications - it cannot be the extinct species because the starting material is not the extinct species. Ancient DNA provides a parts list, not a working cell: sequence information alone cannot reconstitute the epigenetic architecture, chromatin organisation, and cellular machinery required to develop a viable organism, none of which a reconstructed sequence can supply. This means that for every fully extinct species currently targeted by de-extinction programmes - woolly mammoth, thylacine, dire wolf, dodo, moa - true resurrection is biologically precluded by the same cause: no viable cells were preserved before extinction. The realistic ceiling of any programme working from reconstructed sequence and a living relative's cells is, by definition, not true resurrection - because the starting material determines the biological category of the outcome, not the ambition of the programme, the number of loci edited, or the language used to describe it. The sole existing exception is the northern white rhinoceros: a functionally rather than fully extinct species whose somatic cells were biobanked from living individuals before the last breeding males died, enabling iPSC derivation from a genomically authentic source and keeping true resurrection biologically possible [8]. For every species currently in decline, the decision of whether to bank today is, in effect, a decision about whether true resurrection remains an option in the future.

Against this constraint, the five de-extinction programmes currently underway at Colossal Biosciences - targeting the woolly mammoth, Tasmanian tiger, dire wolf, dodo, and, announced in July 2025 in partnership with Peter Jackson and the Ngāi Tahu Research Centre, the South Island giant moa (Dinornis robustus) - are best understood not as exceptions to the rule but as illustrations of it. Each follows the same logic: ancient DNA is recovered from museum or fossil material, divergent loci are identified by comparison with the closest living relative, and CRISPR-based edits are introduced into that relative's genome via a species-specific cellular platform. Evaluated against the spectrum, each programme occupies a distinct position.

The dire wolf sits firmly at functional trait restoration: the three grey wolves (Canis lupus) born in April 2025 carry 20 genomic edits across 14 genes selected for morphological resemblance - coat colour, texture, and body size proxies - with no systematic attempt to approximate the full genomic divergence between Aenocyon dirus and its living relative across their 5.7 million years of independent evolution [101]. What matters is not the number of edits but their intent: traits selected for visual resemblance rather than systematic genomic approximation place this programme unambiguously at the functional trait restoration end, regardless of how it is publicly framed.

The woolly mammoth programme is best characterised as pursuing hybrid proxy as its medium-term goal, with functional trait restoration as its near-term deliverable. Asian elephant iPSCs serve as the expandable, multiplex-editable substrate for stacking cold-adaptive edits across haemoglobin oxygen affinity, subcutaneous fat deposition, fur density, and sebaceous gland activity [102], with proof-of-concept multiplex editing of seven genes in mouse zygotes validating the approach [103]. The intent is systematic genomic approximation of mammoth cold-adaptation biology - placing it in the hybrid proxy category even at its current early stage.

The Tasmanian tiger programme similarly targets hybrid proxy: the goal is broad genomic approximation of thylacine biology in the fat-tailed dunnart (Sminthopsis crassicaudata), underpinned by a chromosome-scale genome assembly estimated at more than 99.9% accuracy [104]. Dunnart iPSC derivation amenable to multiplex editing remains the critical and currently unresolved cellular bottleneck - without it, the programme cannot advance beyond genomic characterisation.

The dodo and moa programmes are at the earliest stages and are best categorised as functional trait restoration in their current form. The dodo programme has achieved Nicobar pigeon PGC culture as a key milestone toward a chimeric surrogate strategy; the moa programme has begun ancient genome sequencing across nine species using tinamou or emu PGCs as the editing platform. Neither programme has yet defined a comprehensive editing target set that would qualify as hybrid proxy ambition. The moa programme faces an additional reproductive engineering challenge with no equivalent in mammalian de-extinction: a South Island Giant Moa egg is estimated to have been approximately 80 times the volume of a chicken egg and entirely beyond the carrying capacity of any living avian surrogate, meaning that an artificial incubation system is a prerequisite for the reproductive pipeline rather than an optional supplement. A significant step toward resolving this was taken in May 2026 when Colossal successfully hatched 26 live chicks from a fully artificial incubation platform - a 3D-printed lattice shell incorporating a bioengineered silicone-based membrane that replicates natural eggshell oxygen transfer under normal atmospheric conditions, without supplemental oxygen (Colossal Biosciences, press release, May 2026). The platform is designed to be scaled to moa egg dimensions and represents the first demonstration that complete avian embryonic development outside a biological eggshell is achievable - though the path from chicken proof-of-concept to a genetically edited moa embryo developing to term remains formidable.

3. iPSC Applications in Veterinary and Comparative Medicine

This section examines iPSC applications across companion animal, livestock, and non-human primate systems - collectively framed within the one medicine paradigm, which recognises that animal and human health are mechanistically and epidemiologically interdependent. A defining feature of this research is its bidirectionality: human iPSC methodology accelerates veterinary platform development, while spontaneous animal diseases - arising in outbred, environmentally relevant contexts with authentic genetic complexity - generate mechanistic insights that feed back into human medicine. Four domains are reviewed: regenerative medicine in companion and livestock species, comparative disease modelling, infectious disease platforms for host-restricted pathogens, and zoonotic surveillance. The section closes with xenotransplantation - the application in which porcine iPSC platforms translate most directly into human clinical medicine - as the clearest expression of what the one medicine framework can deliver in practice. The human iPSC clinical translation baseline established in Section 1.3.4 - including the first regulatory approvals of iPSC-derived therapies in Japan and ongoing Phase I/II trials in the United States - sets the translational benchmark against which all applications reviewed here will ultimately be measured.

3.1. Regenerative Medicine in Companion and Livestock Species

iPSCs offer a compelling alternative to lineage-restricted mesenchymal stem cells for tissue repair, combining unlimited self-renewal with broader differentiation potential [1]. Regenerative applications have advanced furthest in equine, canine, and feline systems, though the gap between cellular proof-of-concept and clinical translation remains substantial across all three.

In horses - where orthopaedic injury is a major clinical and economic burden - equine iPSCs (E-iPSCs) derived from adipose stem cells promote muscle regeneration and form donor-derived myofibres when transplanted into injured mice [105]. Tenogenic differentiation has proven harder: E-iPSCs form inert aggregates rather than functional tendon unless rescued by cyclic mechanical loading or overexpression of the Mohawk transcription factor, and their derivatives are acutely sensitive to the IL-1β and TNF-α that dominate injured tendon, which induces apoptosis and threatens graft survival [106,107]. E-iPSCs therefore remain a research tool, with bone-marrow- and adipose-derived MSCs still the evidence-based standard for equine tendon repair translation [108].

Canine iPSC (ciPSC) work has concentrated on cardiac disease and translational infrastructure. Dogs spontaneously develop dilated cardiomyopathy and degenerative valve disease resembling human heart failure, and canine iPSCs have been differentiated into beating cardiomyocytes expressing cardiac troponin T and MYH6, most recently via a defined AR culture medium (ciPSC medium supplemented with activin and IWR1) yielding spontaneously contracting cells [109,110]. Canine models have also been pivotal to retinal gene therapy - including the preclinical evidence base for Luxturna®, the first FDA-approved gene therapy for inherited retinal dystrophy - and that established infrastructure is now being leveraged to adapt human iPSC-derived retinal pigment epithelium protocols for canine inherited retinal degenerations [111,112]. In cats, non-integrative Sendai-virus reprogramming has produced footprint-free feline iPSCs (fiPSCs) from somatic tissues, from which mesenchymal stromal cells and renal cell models have been generated for regenerative and therapeutic-screening applications [52,113].

Across all species, the primary bottlenecks are shared: allogeneic off-the-shelf products risk immune rejection while MHC matching is impractical given the breadth of veterinary MHC diversity; GMP-grade autologous lines are economically prohibitive at scale; and differentiation fidelity together with jurisdictionally fragmented regulatory frameworks further slows translation [114,115]. The regenerative medicine landscape for companion and livestock iPSCs is one of genuine but unevenly distributed progress - with canine cardiomyocyte differentiation, and feline stromal cell alternatives, advancing on the strength of well-characterised disease models and clear human medicine analogues, while equine tendon repair and broadly applicable allogeneic cell therapy remain constrained by differentiation fidelity and inflammatory microenvironment sensitivity. iPSC technology does not yet deliver finished therapeutic products in these species, but it is progressively building the cellular infrastructure - renewable, genetically defined, and editable at the stem cell stage - from which those products will eventually emerge.

3.2. Disease Modelling and Comparative Medicine

Companion-animal cardiology illustrates the bidirectionality principle most clearly. Canine dilated cardiomyopathy shares causal genetics with the human disease — dystrophin (DMD) and RBM20 variants drive the condition in specific breeds — so ciPSC-derived cardiomyocytes from affected dogs offer a genetically authentic large-animal platform for mechanistic study and therapeutic testing unavailable in rodents [108]. In cats, CRISPR knock-in of the HCM-causing MYBPC3/R820W mutation into fiPSC-derived cardiomyocytes recapitulates cellular hypertrophy and impaired relaxation, with findings directly applicable to human hypertrophic cardiomyopathy [11].

The same logic extends across disease areas. Feline immunodeficiency virus is a naturally occurring lentivirus that mirrors HIV/AIDS in receptor-mediated entry, CD4⁺ T-cell depletion, and neuroinvasion; fiPSC-derived immune-cell populations and brain organoids would enable study of retroviral entry, latency, and neurodegeneration in a species-authentic system [116]. Spontaneous canine tumours share orthologous driver mutations and heterogeneous microenvironments that laboratory-induced rodent models fail to reproduce, making ciPSCs from dogs with naturally occurring cancers a distinctive oncology platform [117]. In livestock, bovine iPSCs carrying the naturally occurring NPC1 variant found in Angus cattle offer a large-animal model of Niemann-Pick type C disease that translates directly to the rare human paediatric condition [118]. Non-human primate platforms extend the logic to neurodegeneration: marmosets develop age-related Aβ and tau pathology and share human-like brain network organisation, and marmoset iPSC-derived neural organoids carrying PSEN1 mutations model Alzheimer's disease under genomic conditions rodents cannot approximate [119,120]. What unites these examples is not merely methodological convergence but a shared translational outcome: in each case, the iPSC-derived model from the affected species has either resolved a mechanistic question that rodent surrogates could not, or opened a therapeutic or vaccine development pathway that was previously inaccessible. The one medicine framework is not simply a conceptual aspiration here - it is being operationalised, disease by disease, in cell types that actually determine pathogen and disease outcome.

3.3. Infectious Disease

iPSC-derived cell platforms have proven particularly transformative for studying pathogens with extreme host and cell-type specificity - diseases for which primary cell access is logistically prohibitive, genetically intractable, or ethically constrained [119,120]. The unifying principle is that iPSC derivation converts a scarce, terminally differentiated, manipulation-resistant target cell type into a renewable, genetically defined, CRISPR-amenable platform - and this conversion is especially consequential for pathogens whose host range is narrow and whose biology cannot be adequately studied in heterologous systems.

The most extensively developed example is African Swine Fever Virus (ASFV), which replicates almost exclusively in porcine macrophages - terminally differentiated, difficult to culture in quantity, and resistant to stable genetic manipulation. iPSC-derived macrophages that support productive ASFV and PRRSV infection and are amenable to CRISPR editing at the stem cell stage circumvent this bottleneck entirely [10]. A genome-wide knockout screen in porcine iPSC-derived macrophages identified the non-classical MHC II molecule SLA-DM as an essential host factor for ASFV replication, with SLA-DMA and SLA-DMB knockout producing severe replication defects confirmed by transgenic reconstitution - establishing SLA-DM editing in porcine iPSCs as a candidate route to ASF-resistant animals [121]. That CD163 knockout confers no ASFV resistance, despite conferring complete PRRSV resistance, confirms that these viruses require distinct host factors and validates SLA-DM as an independently identified iPSC editing target [122]. Extending the ASFV platform across Suidae - domestic pig, wild boar, and warthog - has further enabled direct in vitro recapitulation of the species-differential tolerance phenotype observed in vivo: warthog iPSC-derived macrophages fail to induce interferon-α upon ASFV challenge, a finding directly relevant to vaccine design and spillover risk assessment [52].

Feline infectious peritonitis virus (FIPV) presents an analogous macrophage-tropism problem in companion animal medicine. FIPV arises through mutation of feline enteric coronavirus and causes a near-universally fatal systemic disease characterised by monocyte and macrophage infection, vasculitis, and immune-mediated tissue destruction [123,124]. Antibody-dependent enhancement further complicates vaccine development by facilitating viral uptake by macrophages via non-neutralising antibodies, making mechanistic dissection of macrophage-virus interactions a priority [125,126]. fiPSC-derived macrophages and monocytes would provide the first genetically manipulable, renewable platform for studying FIPV entry, latency, ADE mechanisms, and antiviral candidate screening in species-authentic cells - a direct analogue to the ASFV macrophage system, and one for which the fiPSC derivation infrastructure now exists [49].

Bat iPSC platforms have opened a promising but still nascent angle on infectious disease: interrogating why bats harbour viruses that are highly pathogenic to humans and other animals without themselves succumbing to disease. Bat iPSCs were first derived from Myotis lucifugus using a piggyBac vector delivering eight reprogramming factors, establishing that chiropteran somatic cells are amenable to reprogramming despite their divergent biology [127]. Subsequent work derived iPSCs from Rhinolophus ferrumequinum - a horseshoe bat species closely related to the likely SARS-CoV-2 reservoir - and revealed large virus-filled vesicles persisting in pluripotent cells without compromising proliferation, suggesting an unusual entanglement between bat stem cells and endogenous viral elements. A further study compared SARS-CoV-2 infection of human and bat iPSCs following ACE2 transduction, finding substantially lower infectious virus titres and cytotoxicity in bat iPSCs, with bat embryonic fibroblasts showing fully abortive infection despite productive viral RNA and protein synthesis [128]. These findings are intriguing, but their translational value is currently limited by a critical gap: bat iPSCs have not yet been successfully differentiated into the tissue-specific cell types - respiratory epithelium, macrophages, hepatocytes - that would be needed to model infection at ecologically and clinically relevant target sites. In this respect, bat organoid systems derived directly from adult tissue have thus far proven more tractable: horseshoe bat intestinal organoids support robust SARS-CoV-2 replication [129], and more recently, bat nasal, bronchial, alveolar, and small intestinal organoids have been used to model Marburg virus infection and characterise species-specific antiviral responses at mucosal surfaces [130]. The iPSC route retains its distinctive advantages - genetic manipulability at the stem cell stage, indefinite expandability, and applicability to species from which adult tissue cannot be obtained - but realising these advantages for bat infectious disease biology requires the development of directed differentiation protocols into relevant epithelial and immune lineages, which remains an unmet priority for the field.

Across all three examples - porcine macrophage-tropic viruses, feline coronavirus, and bat reservoir biology - the iPSC framework enables a mode of investigation that is simply not achievable through conventional primary cell or heterologous model approaches: genetically defined, species-authentic, renewable infection modelling in the cell types that actually determine pathogen outcome. The cross-species comparative approach illustrated by the Suidae ASFV platform - in which host-restricted pathogen biology is interrogated simultaneously across phylogenetically diverse species - points directly toward the broader One Health application addressed in the following subsection.

3.4. iPSC-Based Models for Zoonotic Disease and One Health

The WHO One Health framework, most recently formalised in the May 2025 Pandemic Agreement, codifies as a priority integrated surveillance at the human-animal-environment interface - a priority that iPSC technology is conceptually well-positioned to support (https://apps.who.int/gb/ebwha/pdf_files/WHA78/A78_R1-en.pdf). By generating species-specific iPSC-derived organoids from reservoir hosts, intermediate hosts, and humans, it is in principle possible to model discrete stages of viral spillover - reservoir tolerance, intermediate host adaptation, and human susceptibility - within a controlled, reproducible system without the ethical and logistical constraints of multi-species in vivo infection studies [131,132,133]. Multi-species respiratory organoid panels, for instance, could enable rapid assessment of emerging pathogen tropism and adaptation potential at the earliest stages of a spillover event.

This case is compelling in concept but must be stated honestly: no validated multi-species iPSC spillover model currently exists, and the pathway from proof-of-principle organoid infection studies to actionable surveillance tools remains undefined. The value of retaining this framework in the present review is prospective - to position NMS iPSC platforms as infrastructure with One Health utility before that utility is formally demanded - rather than to describe an established application.

3.5. Xenotransplantation

Porcine iPSCs (piPSCs) hold particular promise as a starting material for xenotransplantation, given the anatomical and physiological proximity of porcine and human organs. The iPSC route is advantageous because humanisation edits - introducing complement regulatory genes and knocking out xenoantigens - can be performed and validated at the stem cell stage before differentiation and organogenesis, rather than requiring a whole transgenic animal for each modification [134,135]. Clinical xenotransplantation from iPSC-derived sources remains a long-term prospect, but pig-to-human organ transplants using genetically modified animals are already entering early clinical evaluation, establishing the translational trajectory into which iPSC-derived porcine sources will eventually feed [136,137]. The same principle that makes porcine iPSCs valuable for xenotransplantation - that editing and quality control performed at the stem cell stage before any commitment to whole-animal production offers precision and scalability unavailable through direct embryo or primary cell approaches - applies with equal force to the disease modelling and infectious disease platforms described in the sections that follow.

4. Agriculture and Biotechnology

iPSC technology is increasingly being applied across livestock species to address core challenges in modern agriculture: improving productivity and disease resistance, preserving genetic diversity in the face of climate change, and enabling novel biotechnological applications including cultured meat. Across pigs, cattle, and chickens, iPSC platforms are beginning to move beyond proof-of-concept toward applied tools for sustainable and resilient food systems [138].

Porcine iPSCs (piPSCs) were first derived in 2009 from foetal and postnatal fibroblasts using OSKM factors [139] and have since been refined for targeted genetic modification aimed at both productivity enhancement and disease resistance [140,141,142,143]. Bovine iPSCs (biPSCs) offer a platform not only for genetic trait improvement but also for developing stem cell-based therapies for inherited disorders such as citrullinemia and leukocyte adhesion deficiency in cattle [144,145] . More broadly, ruminant iPSCs - spanning cattle, sheep, goats, and buffalo - support genetic improvement programs, disease resistance breeding, and biomanufacturing applications including cultured meat and dairy [138]. Avian iPSCs represent an emerging frontier: the chicken's short life cycle and well-characterised genome make it particularly tractable for stem cell studies, and chicken iPSCs are being applied to genetic engineering, production trait improvement, and germline transmission to fix desirable variants in breeding populations [146,147].

4.1. Disease Resistance and Modelling

iPSC platforms enable two complementary approaches to agricultural disease challenges: resistance breeding through precision genome editing, and in vitro disease modelling for therapeutic and vaccine development. The mechanistic basis of these platforms - particularly the use of iPSC-derived macrophages for host-restricted pathogen biology - is developed in detail in Section 3.3; the focus here is the resistance breeding logic and its agricultural implications.

The core argument is straightforward: editing immune-related loci in iPSCs, followed by SCNT or embryo production from edited lines, offers a route to generating livestock with heritable, precisely defined disease resistance without the stochasticity of conventional breeding. The iPSC route carries a critical advantage over direct embryo editing in this context: edits can be introduced, screened, and quality-controlled at the stem cell stage across large cell numbers before any commitment to embryo production, enabling multiplex strategies and rigorous off-target assessment that are impractical at the zygote level.

The strongest proof-of-concept comes from the CD163/PRRSV system, in which deletion of scavenger receptor cysteine-rich domain 5 confers complete and heritable PRRSV resistance in pigs, with FDA approval granted in April 2025 for commercial use [148,149]. While achieved through direct embryo editing rather than an iPSC pipeline, it validates the principle that editing a single immune-related locus can produce commercially viable, heritable disease resistance at scale. For ASFV - which has no licensed vaccine and has killed more than 200 million pigs worldwide - the SLA-DM locus identified through iPSC-derived macrophage screens (Section 3.3) represents the leading iPSC-specific editing target, with translation toward live animal models underway at the Roslin Institute [121]. This remains a cellular-level finding; in vivo confirmation in edited animals is the critical prospective step. The amenability of chicken iPSCs to targeted genome editing opens analogous resistance engineering opportunities in poultry - including potential targets relevant to highly pathogenic avian influenza, the most economically and epidemiologically consequential poultry disease globally -complementing conventional selective breeding approaches in avian production species [146,150].

4.2. Climate Resilience and Genetic Diversity Preservation

Climate change poses an escalating threat to livestock productivity, particularly through heat stress, shifting disease landscapes, and the erosion of locally adapted breeds. iPSC technology offers two distinct strategies to address this.

The first is active genetic intervention. The "slick" gene variant in cattle, which produces a short coat and measurably reduces heat stress, illustrates how iPSC-based editing can accelerate the introgression of climate-adaptive traits: the variant can be introduced by gene editing in bovine iPSCs, with edited embryos transferred to surrogates producing phenotypically heat-tolerant calves [151]. The recent generation of stable bovine induced totipotent stem cells (iTPSCs) has further expanded this toolkit [152]. Unlike conventional iPSCs, which are restricted to embryonic lineages, iTPSCs retain the capacity to contribute to both embryonic and extraembryonic tissues - a property that enables more efficient whole-animal cloning from edited cell lines and broadens the reproductive deployment options for climate-adaptive variants. Derived using specific signalling inhibitors alongside the SV40 large T antigen, bovine iTPSCs provide an inexhaustible, cryopreservable cellular substrate for climate-adapted breeding programmes that is not constrained by gamete availability or collection logistics.

The second strategy is preservation. A living iPSC biobank - in which somatic cells from locally adapted breeds are reprogrammed and banked as iPSCs rather than as terminally differentiated cells - creates a living, expandable genetic resource that can be differentiated into germline-competent cells and reintroduced into breeding programs as needed [153]. Unlike conventional cryobanking of sperm or embryos, iPSC banks are not limited by collection logistics or gamete availability, and the cells remain amenable to future editing as new climate-adaptive targets are identified.

4.3. Cultured Meat

iPSCs are increasingly positioned as a foundational cell source for cultured meat production, offering an indefinitely self-renewing starting material that can be differentiated into skeletal muscle and adipose tissue at scale [154]. Species-specific iPSC lines from cattle [155] and pigs [156] have been expanded in serum-free suspension cultures - a prerequisite for cost-effective biomanufacturing - and companies including MyriaMeat and Magic Valley are already leveraging iPSC-derived platforms to produce cultivated meat products [157].

The sustainability case for iPSC-derived meat is substantial: production from a single harmless biopsy eliminates slaughter, reduces antibiotic use, and carries a significantly lower land, water, and greenhouse gas footprint than conventional livestock farming [154]. For wildlife conservation, the implications extend further: iPSC-derived cultured meat from abundant species could in principle reduce hunting pressure on overexploited wild populations, and species-specific iPSC lines from animals that cannot be farmed conventionally, including bushmeat species under poaching pressure, could provide culturally relevant protein alternatives without any wild harvest. These applications remain speculative but are conceptually grounded in the same iPSC differentiation logic that underpins the agricultural platforms reviewed above.

Large-scale viability, however, depends on overcoming persistent and interconnected challenges. On the production side, the cost of growth media - particularly the removal of animal-derived serum and the replacement of recombinant growth factors with cost-effective alternatives - remains the dominant bottleneck: cultivated meat is projected to remain more expensive than conventional meat [158]. Scaffold technology for replicating the fibrous texture of whole-cut muscle, rather than ground or processed products, remains an unresolved materials science and bioengineering challenge. Consumer acceptance presents an additional non-technical barrier, with surveys consistently identifying naturalness concerns and labelling ambiguity as primary sources of reluctance across major markets [158].

The regulatory landscape is fragmented and rapidly evolving. Singapore was the first jurisdiction to approve a cultivated meat product for commercial sale in 2020, followed by the United States, where GOOD Meat and Upside Foods received joint FDA/USDA clearance in 2023 enabling limited commercial sales. The European Union is reviewing applications under its Novel Food Regulation with no approvals to date, while Italy has moved in the opposite direction, legislating a ban on production and sale - illustrating the patchwork of national positions that complicates multinational commercialisation [158]. For iPSC-specific products, an additional regulatory layer applies: no jurisdiction has issued specific guidance on iPSC-derived food ingredients, and open questions regarding residual pluripotency, genomic stability of extensively passaged lines, and the acceptability of reprogramming vectors in a food production context will need to be resolved before iPSC-derived cultured meat can advance through approval pipelines. Engaging regulators early to define the required evidence standards will be as important as solving the remaining biological challenges.

5. Environmental Protection

Conventional ecotoxicological assessment has relied on a narrow repertoire of model organisms - zebrafish embryos, Daphnia, rat hepatocyte lines, and standardised Organisation for Economic Co-operation and Development (OECD) bioassays - that inadequately represent the physiological and metabolic diversity of wild species [159,160]. Extrapolating toxicity data from these models to ecologically relevant wildlife involves assumptions about receptor conservation, metabolic equivalence, and developmental timing that are frequently violated across the phylogenetic distances involved. Non-human iPSC-derived cells offer a route to interrogate contaminant effects in a species-, tissue-, and developmental-stage-specific manner, with the capacity to resolve mechanisms that whole-animal tests obscure and to generate data directly relevant to species of conservation concern.

Four toxicological mechanisms are particularly well-suited to iPSC-based investigation in wildlife contexts: AHR-mediated toxicity (dioxins, PCBs, PAHs), endocrine disruption (EDCs via nuclear receptor signalling), oxidative stress and heavy metal toxicity (NRF2/KEAP1 pathway), and developmental toxicology (WNT/BMP/NODAL signalling, gastruloid models).

5.1. iPSC-Derived Cell Systems as Species-Authentic Reporters of Organic Pollutants

Dioxins, PCBs, and polycyclic aromatic hydrocarbons represent among the most ecologically consequential persistent organic pollutants in aquatic and terrestrial environments, and their primary mechanism of cellular toxicity - activation of the aryl hydrocarbon receptor (AHR) - is one of the least adequately captured by current standard assay species. The AHR axis operates via bHLH/PAS-domain-dependent nuclear translocation, ARNT dimerisation, and transcriptional induction of CYP1A1 and CYP1B1, with downstream consequences that extend beyond xenobiotic metabolism to include mitochondrial dysfunction and G1/S cell-cycle arrest via direct interaction with hypophosphorylated retinoblastoma protein - effects with direct implications for tissue renewal and reproductive output [161,162,163,164]. The central problem for ecological risk assessment is that AHR ligand affinity varies up to 1,000-fold across vertebrate species, meaning that a contaminant concentration tolerated by a rat hepatocyte may be acutely toxic to a sentinel fish, raptor, or marine mammal - divergences that human or rodent-based regulatory assays are structurally unable to detect [165]. No NMS iPSC-derived cell system has yet been applied to this problem, despite the fact that the scientific rationale is among the strongest of any mechanism reviewed here. iPSC-derived hepatocyte-like or gill-like cells from species with characterised AHR polymorphisms would enable population-specific mechanistic interrogation at precisely the level of biological resolution that risk assessment requires. Atlantic killifish populations independently adapted to Superfund-site PCB contamination via convergent AHR pathway desensitisation represent a particularly compelling proof-of-concept target: the genetic basis of tolerance is characterised, the ecological stakes are high, and the contrast between sensitive and resistant populations within a single species offers a controlled system for dissecting AHR-mediated toxicity without cross-species confounding [166,167].

5.2. Organic Pollutants and Endocrine Disruption

As with AHR toxicology, no NMS iPSC-derived steroidogenic or thyroid lineage has been deployed in an Endocrine-Disrupting Chemicals (EDCs) assay context. The gap this creates is significant. EDCs perturb nuclear receptor signalling through ERα, AR, and TR pathways [168], as well as non-nuclear receptor mechanisms including AHR-ERα cross-talk, in which ligand-bound AHR directly suppresses oestrogenic gene expression [169]. Most validated in vitro EDC assays use rodent or human receptor constructs that may substantially misrepresent potency in wildlife species with divergent ligand-binding domains [170]. Species-specific iPSC-derived steroidogenic or thyroid lineages could directly address this limitation [171], though validated differentiation protocols for any wildlife species do not yet exist and represent a significant unmet technical need.

5.3. iPSC-Derived Platforms for Oxidative Stress and Heavy Metal Toxicity

ROS-mediated oxidative stress induced by heavy metals, nanomaterials, and pesticides triggers NRF2/KEAP1 stress responses, mitochondrial dysfunction, and apoptosis [172,173,174]. Here, the potential contribution of NMS iPSC platforms is practical as much as mechanistic. Primary cells from rare or protected species are difficult to obtain in quantities sufficient for replicated dose-response experiments, and are legally prohibited to collect repeatedly in most jurisdictions. iPSC-derived cells would provide a renewable, genetically defined alternative that overcomes both the supply constraint and the inter-preparation variability inherent to primary cell isolation [175,176]. While NRF2 pathway characterisation using iPSC-derived cells has been demonstrated in human and rodent systems [177], its application in wildlife iPSC platforms remains largely unrealised; establishing baseline NRF2/KEAP1 signalling profiles in iPSC-derived cells from species with documented heavy metal exposure histories - cetaceans, raptors, and freshwater fish from contaminated catchments - represents a tractable near-term priority [178,179,180].

5.4. iPSC-Based Gastruloid Models for Developmental Toxicology

iPSC-based embryoid body and gastruloid models have been used to assess teratogenicity via perturbation of WNT, BMP, and NODAL/ACTIVIN signalling axes in human and rodent systems, providing an alternative to whole-embryo tests that is both more mechanistically informative and more ethically tractable for protected species [181,182]. Extension to NMS has not been attempted. Doing so requires species-specific baseline characterisation before such models can be deployed with confidence - specifically, confirmation that WNT/BMP signalling dynamics during gastruloid formation mirror in vivo developmental gene expression in the target species, which in turn requires single-cell transcriptomic reference data from early embryos that exist for very few NMS [183,184,185,186]. As these genomic resources improve [187,188], developmental iPSC models will become increasingly deployable across ecologically relevant taxa - but this remains a second-order priority contingent on first establishing stable, validated iPSC lines and basic differentiation capacity in the species of interest.

5.5. From Aspiration to Action: Prioritising NMS iPSC Investment Across Toxicological Mechanisms

Across these four mechanisms, a common thread is that the mechanistic specificity enabled by iPSC-derived wildlife cells represents a qualitative advance over what conventional ecotoxicological models can provide [160,189]: not merely a more convenient assay format, but access to species-authentic receptor biology, developmental context, and tissue-specific metabolism that are invisible to standardised bioassays [190]. The honest picture, however, is that none of these applications yet exists in practice for NMS. The most tractable near-term entry points are those where the upstream requirements are already partially met: mammalian wildlife iPSC lines that exist today could be directed toward NRF2/KEAP1 characterisation without new reprogramming investment; killifish iPSC derivation, if achieved, would immediately unlock AHR population toxicology; and EDC receptor assays could be built on primate iPSC lines already established for conservation purposes. Progress across all four mechanisms is contingent on the same upstream investments: improved genome annotation for non-model species [191], development of species-specific differentiation protocols [184] and establishment of in vivo transcriptomic reference datasets against which iPSC-derived cell identity can be validated [192,193]. Until these foundations are in place, wildlife iPSC-based ecotoxicology will remain largely aspirational - but the scientific case for building them is compelling, and the tools to do so are rapidly maturing.

5.6. Taxonomic Readiness for iPSC-Based Ecotoxicology: Mammals, Fish, and Amphibians

The practical utility of iPSC-based ecotoxicology depends entirely on the availability of stable, validated cell lines, and this varies enormously across taxonomic groups. In mammalian wildlife, robust iPSC derivation and partial differentiation have been achieved across a growing number of species, and the prospect of "living biobanks" that integrate toxicological screening capacity with genetic preservation is realistic [7]. Here, iPSC platforms offer the most immediately actionable value, enabling chronic, sub-lethal, multi-compound exposure studies in species of direct conservation concern without requiring repeated tissue collection from protected animals.

Fish-derived iPSC platforms would substantially extend ecotoxicological capabilities to aquatic species of conservation relevance, enabling tissue-type-specific interrogation of contaminant exposures that accumulate across developmental time [194]. However, whether this is achievable in the near term depends critically on establishing stable, well-validated fish iPSC lines - a goal not yet demonstrably met for any species. Recent progress in chemical reprogramming of koi caudal fin fibroblasts using a staged chemically defined medium represents meaningful but preliminary progress; pluripotency validation and directed differentiation into ecotoxicologically relevant cell types remain to be demonstrated [195].

The case for amphibian iPSC development is arguably the most urgent. The permeable skin, biphasic life history, and exquisite sensitivity of amphibians to contaminants, thermal stress, and pathogens such as Batrachochytrium dendrobatidis make them ideal cellular sentinels for environmental monitoring. Recent frog primary fibroblast cultures applied to ecotoxicological assays [196] represent genuine but limited progress - these are not renewable iPSC platforms, and the fundamental reprogramming milestones remain unachieved for any amphibian species. This is the largest gap between ecological urgency and technical readiness in the entire ecotoxicology section and closing it should be treated as a priority for the field.

Scaling iPSC-based environmental monitoring across the hundreds of threatened species for which such platforms would be most valuable faces formidable logistical challenges: each species requires bespoke reprogramming optimisation, pluripotency validation, and differentiation protocol development. Prioritisation frameworks - identifying sentinel species that are both ecologically informative and technically tractable - will be essential if the field is to allocate limited resources effectively.

5.7. iPSC-Based Environmental Toxicology in NMS: What Is Demonstrated, What Remains Aspirational

Non-human iPSC technology occupies an intellectually compelling but technically uneven position in environmental science. Its most clearly demonstrated value lies in mammalian wildlife, where robust derivation and practical biobanking are achievable today, and where the path from existing iPSC lines to rudimentary toxicological assays is shorter than in any other taxonomic group. For aquatic and amphibian taxa - arguably the most ecologically urgent targets - iPSC platforms remain aspirational, constrained by unresolved reprogramming challenges and the absence of validated pluripotency standards. The mechanistic arguments for deploying these systems in AHR toxicology, EDC assessment, developmental toxicology, oxidative stress and heavy metal toxicity are well grounded, but none has been realised in practice for NMS, and realising them requires species-specific technical investment not yet commensurate with the ambition. Used critically and with transparency about current limitations, non-human iPSC technology has genuine potential to strengthen environmental protection; used uncritically, it risks overpromising at the expense of the field's credibility.

6. Conclusions

The trajectory of iPSC science across NMS over the past fifteen years is one of genuine, if uneven, progress. From the first proof-of-concept lines derived from a drill and a Northern white rhinoceros in 2011, the field has expanded to encompass megafauna, marsupials, monotremes, primates, birds, bats, and early forays into fish - spanning applications from wildlife genetic rescue and veterinary regenerative medicine to agricultural disease resistance, cultured meat, and ecotoxicological assessment. Taken together, this body of work makes one conclusion difficult to avoid: the principal barriers to extending iPSC technology across the tree of life are not biological in nature. The core pluripotency network is conserved across vertebrates. Reprogramming is, in principle, achievable in any somatic cell. What limits the field is not biology but infrastructure - the absence of annotated reference genomes, validated species-specific reagents, standardised culture conditions, agreed quality control frameworks, and the institutional funding mechanisms that would make systematic protocol development feasible at scale.

This infrastructure deficit manifests at every stage of the iPSC workflow. Pluripotency state diversity across vertebrates remains poorly characterised: the naïve, formative, and primed states defined in mouse and human have no confirmed equivalents in most NMS, and without single-cell embryogenesis atlases to define the in vivo target, reprogramming outcomes in novel species cannot be benchmarked against a biological ground truth. Culture incompatibility compounds this uncertainty - commercially available maintenance media optimised for human or murine cells routinely fail in NMS, while the species-specific growth factor requirements that would guide rational medium design are almost entirely unknown. Factor choice presents a parallel challenge: canonical OSKM combinations frequently underperform in species whose pluripotency networks have diverged from the mammalian blueprint, and the identification of optimal species-matched or species-supplemented factor sets requires genomic and transcriptomic resources that most NMS lack. The shift toward non-integrative delivery - Sendai virus, episomal vectors, and chemically defined reprogramming - has meaningfully reduced the genomic footprint of NMS iPSC lines and brought conservation and reproductive applications closer to practical feasibility. But non-integrative efficiency remains lower than viral methods, and optimisation across taxonomically diverse cell types is far from complete.

Functional validation represents perhaps the deepest unresolved challenge. In the absence of germline-competent chimera formation - a benchmark achieved for NMS in only a handful of cases - the developmental potency of most published NMS iPSC lines remains formally unverified. Surrogate validation criteria such as pluripotency marker expression, embryoid body formation, and teratoma assays, while informative, are insufficient to exclude the possibility that cells represent partially reprogrammed intermediates. The field would benefit substantially from agreed minimum reporting standards that specify which combination of molecular, functional, and genomic quality control steps constitutes an acceptable evidence base for a published NMS iPSC line -standards calibrated to be scientifically rigorous while remaining practically achievable given the logistical realities of working with protected or rare species.

The applications reviewed here are not equally mature. In mammalian wildlife conservation, iPSC derivation and early downstream applications - PGCLCs induction, organoid generation, naïve pluripotency capture - are demonstrably achievable today, and the case for building living cellular biobanks as a genetic safety net for critically endangered species is scientifically and practically compelling. In veterinary medicine, iPSC-based disease modelling for companion animal conditions is advancing toward mechanistic utility, and the one medicine framework positions these platforms as genuinely bidirectional contributors to both veterinary and human biomedical knowledge. In agriculture, the CD163/PRRSV and SLA-DM/ASFV proof-of-concept findings demonstrate that iPSC-mediated genome editing can identify and validate disease-resistance targets at cellular resolution before any commitment to animal production - a workflow advantage over embryo-level editing that will only grow as multiplex editing demands increase. In ecotoxicology and comparative medicine, the scientific rationale for species-authentic iPSC-derived cell systems is well-grounded, but the platforms themselves are largely aspirational for non-mammalian taxa, constrained by the same upstream deficits in reprogramming optimisation and genomic annotation.

Looking forward, four investments are most likely to determine the pace at which iPSC technology fulfils its potential across NMS. First, expanded genomic and transcriptomic infrastructure - high-quality reference genome assemblies, single-cell preimplantation atlases, and comparative pluripotency gene regulatory network maps - will simultaneously accelerate factor design, culture optimisation, and reprogramming fidelity assessment across taxonomic groups. Initiatives such as the Earth BioGenome Project are directly enabling this, and the iPSC field should ensure it is positioned to translate genomic outputs into reprogramming reagents and validated protocols as reference data become available. Second, the development of species-specific differentiation protocols - particularly for germline, hepatic, neural, and immune lineages - will determine whether iPSC lines remain cellular curiosities or become functional platforms for conservation, disease modelling, and therapeutic development. Third, governance frameworks suited to the novel biological entities produced by NMS iPSC research - blastoids, genome-edited surrogates, iPSC-derived gametes from functionally extinct species - must be developed in parallel with the science rather than after it. The current absence of regulatory guidance for these entities in most jurisdictions represents a structural vulnerability that could impede translation at precisely the moment when biological feasibility is achieved. Fourth, strategic prioritisation of species for iPSC investment - concentrating effort on taxa that are conservation-critical, technically tractable, and capable of generating broadly transferable methodological insights - will be essential in a field where per-species protocol development remains resource-intensive.

The iPSC revolution that transformed human biomedical research over the past two decades is now, demonstrably, extending beyond its original boundaries. The cells, the concepts, and increasingly the tools are in place. What remains is the systematic, coordinated, and adequately funded effort to build the species-specific knowledge base that will allow those tools to be used with the precision and confidence the stakes demand. For species on the edge of extinction, for companion animals whose diseases mirror our own, for livestock systems under pressure from climate and pathogen alike, and for ecosystems whose sentinel species are being silently compromised by chemical exposures our current assays cannot detect - the case for making that investment is not merely scientific. It is urgent.