Improvement of in vitro and early in utero porcine clone development after somatic donor cells are cultured under hypoxia

Abstract Genetically engineered pigs serve as excellent biomedical and agricultural models. To date, the most reliable way to generate genetically engineered pigs is via somatic cell nuclear transfer (SCNT), however, the efficiency of cloning in pigs is low (1–3%). Somatic cells such as fibroblasts frequently used in nuclear transfer utilize the tricarboxylic acid cycle and mitochondrial oxidative phosphorylation for efficient energy production. The metabolism of somatic cells contrasts with cells within the early embryo, which predominately use glycolysis. We hypothesized that fibroblast cells could become blastomere‐like if mitochondrial oxidative phosphorylation was inhibited by hypoxia and that this would result in improved in vitro embryonic development after SCNT. In a previous study, we demonstrated that fibroblasts cultured under hypoxic conditions had changes in gene expression consistent with increased glycolytic/gluconeogenic metabolism. The goal of this pilot study was to determine if subsequent in vitro embryo development is impacted by cloning porcine embryonic fibroblasts cultured in hypoxia. Here we demonstrate that in vitro measures such as early cleavage, blastocyst development, and blastocyst cell number are improved (4.4%, 5.5%, and 17.6 cells, respectively) when donor cells are cultured in hypoxia before nuclear transfer. Survival probability was increased in clones from hypoxic cultured donors compared to controls (8.5 vs. 4.0 ± 0.2). These results suggest that the clones from donor cells cultured in hypoxia are more developmentally competent and this may be due to improved nuclear reprogramming during somatic cell nuclear transfer.


| INTRODUCTION
For two decades, since the creation of the cloned first lamb using a somatic cell donor, the scientific community has strived to improve the efficiency of somatic cell nuclear transfer (SCNT) in mammals (Campbell, McWhir, Ritchie, & Wilmut, 1996;Gábor, 2018). Notably, numerous live animals from 23 species have been successfully produced by SCNT due to an amalgamation of efforts from several dedicated researchers discovering new strategies (reviewed in Loi, Iuso, Czernik, & Ogura, 2016). Therefore efficiency has improved over time and breakthroughs are still occurring, where even the most challenging species are proving to be clonable (Liu et al., 2018). While other emerging techniques are improving, such as zygote injection of CRISPR/ Cas9, SCNT is still one of the most efficient ways to reliably create genetically engineered animals. Despite the successful induction protocols for ESC and iPSC in rats (Liao et al., 2009) and mice (Bryja, Bonilla, & Arenas, 2006), creating reliable protocols has proven to be more challenging in the larger domestic species. A major hurdle for cow and pig iPSCs is that both 1) require continued expression of the introduced transgenes to maintain some degree of pluripotency, 2) have limited passage life, and 3) are not able to form teratomas in immunodeficient mice (reviewed in Ezashi, Yuan, & Roberts, 2016).
Due to these challenges, SCNT is still popular in swine.
Gene-edited and/or transgenic (GET) pigs serve as excellent models to study disease progression and develop treatments for human genetic disorders. The pig, in particular, conveys great similarity to humans in their anatomy, physiology, and genomics, therefore, allowing them to exhibit symptoms of human pathologies more accurately and reliably than other animal models (reviewed in Fan & Lai, 2013;Prather, Lorson, Ross, Whyte, & Walters, 2013;Walters et al., 2012). For several reasons, GET pigs are the most likely donor option for future xenotransplantation of organs; therefore great efforts have been underway to make this possibility a reality (Bottino et al., 2014;Ekser, Rigotti, Gridelli, & Cooper, 2009;Klymiuk, Aigner, Brem, & Wolf, 2010;Lai et al., 2002;Lavitrano et al., 2002;Lutz et al., 2013;Mohiuddin et al., 2014;L. Yang et al., 2015). Due to the usefulness of GET pigs in both biomedicine and agriculture, there is a growing need for the creation of new or improved upon models, which at this time, is still predominately achieved via SCNT. Currently, the efficiency of SCNT in pigs is approximately 1-3% (Whitworth & Prather, 2010). This percentage varies slightly depending on if investigators calculate success rate based on total embryos reconstructed, total embryos transferred, or pregnancy rate of surrogates to which embryos were transferred.
Nevertheless, this rate is low and there is a substantial need for improvement. When blastomeres are used as donor cells for SCNT the rate of success is significantly improved (Mitalipov, Yeoman, Nusser, & Wolf, 2002). Therefore we speculated that if somatic cells could be induced to be more blastomere-like, the cloning efficiency may be greatly improved.
There is growing evidence that cellular reprogramming is facilitated in part by upregulation of glycolysis (Folmes et al., 2011;Kondoh et al., 2005;Moussaieff et al., 2015;Zhu et al., 2010). In light of this, we reasoned promoting a highly glycolytic metabolism could facilitate nuclear reprogramming. Somatic cells predominately use mitochondrial oxidative phosphorylation and the citric acid cycle for the production of energy whereas the metabolism of preimplantation embryos is evidenced to be more glycolytic and Warburg effect-like in nature (Krisher & Prather, 2012). We hypothesized that if donor fibroblasts were cultured under hypoxic conditions, it would elicit the higher glycolytic activity thereby promoting metabolism exhibited in the blastomeres of early embryos. We speculated this would aid in the facilitation of nuclear reprogramming and improve cloning efficiency. In this study, we investigated if restricting oxygen from donor fibroblasts during cell culture would improve measures of in vitro developmental quality in reconstructed pig embryos and improve survival through early gestation.

| RESULTS
Fusion of reconstructed embryos was not significantly different (p = 0.10) between hypoxic (HYP) and control (CON) cultured fibroblast donors and was not different amongst cell lines (p = 0.44). While the total proportion of embryos cleaved was not different between donor fibroblast oxygen culture treatments (p = 0.75), HYP reconstructed clones had a higher (p = 0.03) proportion cleave earlier (within the first 24 hr) compared to control in both experiments (Table 1)
In the present study, we did not note any differences in blastocyst development or Day 35 pregnancy rates between the use of hypoxic atmosphere incubators or hypoxic chambers. We switched to hypoxic chambers due to the feasibility of use as well as cost. One limitation is that we did not have a system to monitor how much oxygen was in  (Mordhorst, Murphy, Schauflinger, et al., 2018;Prather, Rowland et al., 2013). A cryogenic vial of fibroblasts (0.5 ml aliquots;~1.5 million/ml in media containing 90% fetal bovine serum (FBS) and 10% dimethyl sulfoxide) was defrosted from liquid nitrogen storage for every replicate in the experiment. Cells were thawed and cultured in Dulbecco's modified Eagle's medium (1 g/L glucose with phenol red) supplemented with 15% FBS (Corning, Corning, NY) for seven days in T25 flasks (Corning). Cells were either treated as controls (CON) cultured in 5% oxygen for the duration of 7 days or cultured in step-wise decreasing concentrations of oxygen (HYP) where for 2 days they were maintained at 5% oxygen, on the third day cultured at 2.5% oxygen, and from the fourth to the seventh days cultured at 1.2% oxygen. Hypoxic culture conditions were achieved one of two ways; either via increased nitrogen gas concentration in a standard mixed gas incubator or by using purchased preanalyzed mixed gas cylinders to fill humidified modular incubator chambers, which were placed in incubators.
When whole incubators were used (three initial in vitro experiment replicates and two replicates of embryo transfer experiments), a system of four nitrogen tanks connected by one gas line was assembled to ensure adequate nitrogen gas was available to incubators. The regulators of nitrogen tanks were set to sequentially empty the tanks one by one as culture experiments required a large quantity of nitrogen gas, and at least one of the tanks had to be exchanged for a full tank daily. During this, oxygen concentration of incubators was regularly monitored by a handheld gas monitor (model: InControl 1050; Labotect Labor-Technik-Göttingen GmbH; Göttingen, Germany) which was professionally tested by Utility Lab Services (part of Thermo-Fisher Scientific; Waltham, MA) and proved to have an error of ±0.05% oxygen.
Therefore the low oxygen treatment may have been as high as 1.25% or as low as 1.15% oxygen. Incubators were maintained at 38.5°C

| Somatic cell nuclear transfer and embryo culture
Methods for SCNT were previously reported by (Mordhorst, Murphy, Schauflinger, et al., 2018). Sow-derived oocytes were purchased from DeSoto Biosciences (Seymour, TN) and shipped overnight in maturation medium (90% M199 and 10% follicular fluid supplemented with 0.57 mM cysteine, 5 μg/ml insulin, 10 ng/ml epidermal growth factor, 5 μg/ml LH and FSH). After maturation ( for 30 min (Machaty, Wang, Day, & Prather, 1997). Embryos were then incubated in MU1 (Redel, Tessanne, Spate, Murphy, & Prather, 2015) with a histone deacetylase inhibitor 0.5 µM Scriptaid, for 14-16 hr in a 5% carbon dioxide (atmospheric oxygen) incubator (Whitworth, Zhao, Spate, Li, & Prather, 2011;Zhao et al., 2009 Mordhorst, Murphy, Ross, et al., 2018). In embryo transfer experiments, cell lines proven to produce live offspring were used to distinguish (via genotyping; Figure 2)   | 563 test of normality values were generated using the Univariate procedure of SAS (SAS, Cary, NC); log or square root transformations were made before statistical analysis where necessary to meet assumptions for analyses. All data were analyzed using a MIXED procedure in SAS for main effects of donor cell treatment and cell line; interactions were assessed however this effect was removed from analyses when found to be insignificant. For embryo transfer data, blastocysts transferred were nested within surrogate gilt and a random effect of replicate was used (mainly because it corresponds to the time/season differences for this study duration). Survival probability was considered the number of favorable outcomes (viable fetuses) from the total (embryos transferred). Analyzed variables were considered statistically different if the p < 0.05.