FGF treatment of host embryos injected with ES cells increases rates of chimaerism

In spite of the emergence of genome editing tools, ES cell mediated transgenesis remains the most controllable way of creating genetically modified animals. Although tetraploid (4N) complementation of 4N host embryos and ES cells, is the only method guaranteeing that offspring are entirely ES cell derived, this technique is challenging, not always successful and difficult to implement in some laboratory settings. The current study shows that pretreatment of host blastocysts with FGF4 prior to ES cell injection can provide an alternative method for the generation of animals displaying high rates of chimaerism. Chimaerism assessment in E11 fetuses and born pups shows that a large percentage of resulting conceptuses show a high ES cell contribution from implantation onwards and that developing pups do not necessitate c-section for delivery.


Introduction
The capacity of murine embryonic stem (ES) (Evans and Kaufman 1981;Martin 1981) and induced pluripotent stem (iPS) cells to contribute to all embryonic lineages when injected into murine preimplantation embryos, has had a tremendous impact on transgenic research (Bradley et al. 1984). Over the years the technique has been adapted in order to increase the efficiency of transgenesis. The first improvement came in 1990 when Nagy et al. aggregated ES cells with tetraploid embryos (Nagy et al. 1990). As previous investigators had shown that the viability of tetraploid cells in pups is poor (Kaufman and Webb 1990;Lu and Markert 1980;Snow 1975), the assumption was that resulting offspring from an aggregation between tetraploid and ES cells would be merely derived from the ES cells. While initially there were no viable offspring, a later study improved the method by utilizing low passage number inbred ES cells (Li et al. 2007;Nagy et al. 1993). The researchers (Nagy et al. 1993) hypothesized, however, that not only the efficiency might be dependent on the passage number but also on the background of the ES line (inbred or hybrid). Eggan and colleagues confirmed this latter hypothesis and showed that hybrid ES cell lines were more potent in generating live offspring than inbred ES cell lines (Eggan et al. 2001). Nevertheless the efficiency of tetraploid complementation assays remains rather unsatisfying as the technique only works with certain ES cell lines and pups not only need to be delivered via c-section requiring additional nursing mothers but are also subject to various lethal disorders. As an alternative to tetraploid complementation, injection of ES cells into pre-compaction stage embryos has been introduced (Poueymirou et al. 2007). This technique permits the generation of highly chimaeric mice with less technical hurdles, but so far has not always been successful in different laboratory settings (Hu et al. 2013). The direct application of genome editing tools in oocytes (Shen et al. 2014;Singh et al. 2015;Wang et al. 2013) has briefly led to believe that ES cell mediated transgenesis would become needless. The difficulty to create mice with conditional or compound mutations via direct application of these genome editing tools in oocytes (Singh et al. 2015), however, stillhighlights the importance of ES cell mediated transgenesis.
The success of creating animals that are entirely ES cell derived through tetraploid complementation rests on the limited developmental potential of the epiblast from tetraploid host embryos. In normal circumstances, cells from the inner cell mass (ICM) differentiate to become either the primitive endoderm or the naïve epiblast. Cell fate determination in the ICM is governed by FGF signaling as FGF stimulates the formation of primitive endoderm (Yamanaka et al. 2010). Inhibition of FGF signaling in preimplantation embryos is therefore commonly utilized to derive ES cells (Gallagher et al. 2003). The current study shows that the contribution of ES cells injected into preimplantation embryos can be improved if the developmental potential of the epiblast of the host embryo is biased towards primitive endoderm following exposure to high concentrations of FGF4.

Materials and methods
Embryo collection Mice were exposed to an artificial 12 h light-dark cycle (with lights on from 7:00 to 19:00 h). The ovarian cycle of females (all strains, age between 6 and 10 weeks) was synchronized by intra-peritoneal injections of Folligonan and Chorullon (50U/ml, both one injection with 48 h in between, usually around 13.30 pm). Dosage of the hormones varied according to the age and strain of the females (Cast/Eij: 80-100 ll; all other strains:100 ll if 6 weeks, 125 ll if 7 weeks, 150 ll for females 8 weeks and older). Following the last injection each female mouse was mated with a male mouse. About 20 h after the last injection, mating was evaluated by looking for the presence of a copulation plug in the vagina. Females that had mated were euthanized at E2.5 to collect host embryos for ES cell injection or at E3.5 for ES cell derivation.
Host embryos for ES cell injection ( Fig. 1) Host embryos were obtained from inbred crosses of C57BL/6JOlaHsd (Envigo, The Netherlands) mice. Embryos were flushed from the oviducts at E2.5 and randomly placed into pre-equilibrated (approximately 2 h) mineral oil covered culture drops containing either KSOM or KSOM supplemented with FGF4 (1000 ng/ml, Sigma) and heparin (1 lg/ml, Sigma) and kept in culture at 37°C and 5% CO 2 . The following day, at E3.5, 15-25 ES cells were injected into the blastocoel of cavitated embryos. The embryos were placed back into their respective culture drops and transferred if a blastocoel reformed. Approximately 5-10 embryos were transferred to the uterine horns of pseudopregnant (cross between CBA/J from Charles River Laboratories, The Netherlands and C57BL/10ScOlaHsd.from Invigo, The Netherlands) either the same day within a few hours (E3.5) or early in the morning of the day following (E4.5) the blastocyst injection. The pseudoporegnant females underwent general anaesthesia using an intra-peritoneal injection of Ketamine (120 mg/kg) and Rompun (7.5 mg/kg) and subsequently received a pain killer (Rimadyl Cattle 5 mg/kg, subcutaneous). Uteri were exposed through a small incision on the back of the mouse and embryos were mouth pipetted into the uteri through an opening made with a needle. Following the transfer, uteri were pushed back in the abdomen and a suture was made around the incision. To prevent a drop in body temperature during wakening, the mice were placed on a heated mat.
FACS analysis of E11 fetuses E11 fetuses were removed from the uterine horns and dissociated to single cells by cutting each fetus into Embryos were flusded at E2.5 around compaction and were immediately placed into either regular KSOM or KSOM supplemented with FGF4 (1000 ng/ml) and heparin (1 lg/ml). At E3.5, blastocysts were injected with ES cells (inbred or hybrid) and placed into their respective culture medium. Embryos that reformed the blastocoel were transferred to the uterine cavity either the same day (E3.5) of the injection or the following day (E4.5) Transgenic Res (2017) 26:237-246 239 small pieces, exposing the pieces to pre-warmed (37°C) 0.25% Trypsin/EDTA (Life Technologies) for 10 min followed by their aspiration into 2 ml syringes using needles of various diameter. The suspension (approximately 2 ml) was subsequently collected in a tube containing 3 ml of PBS with 20% FCS in order to inactivate the 0.25% Trypsin/EDTA. Following centrifugation (1000 rpm for 5 min), the pellet was dissolved into 1 ml of PBS with 2% FCS and passed through a cell strainer before the percentage of GFP positive cells was assessed via FACS.

GFP contribution in adult offspring
Lungs, heart and brain representing respectively endoderm, mesoderm and ectoderm were dissected from adult offspring (age over 30 weeks) with a chimaeric coat colour and the contribution of GFP relative to H2A was assessed via real-time PCR. Non-chimaeric offspring were not included in this analysis. Genomic DNA was extracted via a phenol chlorophorm extraction and DNA was precipitated through a standard ethanol precipitation. For the genotyping PCR, all samples were analyzed were run in a 10 ll final reaction volume using the BioRad CFX 384 Real-time System. The reaction mixture contained 29 SYBR Green PCR Master Mix (Applied Biosystems, #4309155), primer (either GFP or H2A) and 2 ll of genomic DNA (approximately 10 ng). The sequence of the primers were as follows: H2A Forward primer CTTTGCGCTTTCGTGATGTC, H2A Reverse primer CCCCACCGGGAACTGTAG, GFP Forward primer CCTACGGCGTGCAGTGCTTCA GC GFP Reverse primer CGGCGAGCTGCACGCTG CGTCCTC. After an initial hold at 95°C for 10 min, reaction mixtures underwent 40 cycles of 30 s at 95°C, 30 s at 57°C and 1 min at 72°C. Standard curves were generated using mouse ES cell genomic DNA from an Actin GFP ES cell line and efficiencies used to calculate the abundance of GFP.

Results
E11 chimaerism rate improved following exposure to FGF4 ICM cells of preimplantation embryos exposed to high concentrations of FGF4 and heparin are forced towards development of extra-embryonic lineages (Yamanaka et al. 2010). This study indicated that ICM cells are entirely biased to develop towards the primitive endoderm when the FGF4 concentration approximates 1000 ng/ml. To test whether the reduced developmental fate of host embryos exposed to FGF4 (1000 ng/ml) and heparin (1 lg/ml) is beneficial for the production of chimaeric mice, E11 fetuses created following injection of ES cells into host embryos, that either were or were not exposed to FGF4 and heparin were analyzed. Embryos were recovered at E2.5 and placed in culture. The following day, embryos had attained the morula or blastocyst stage. Only blastocysts were injected with ES cells and were re-placed into their respective culture medium. Both inbred C57BL/6JOlaHsd and hybrid 129S2/SvHsd/C57BL/ 6JOlaHsd Actin GFP ES cell lines were injected into blastocysts cultured in regular KSOM or KSOM supplemented with FGF4 (1000 ng/ml) and heparin (1 lg/ml). In total, respectively, 52 and 58 blastocysts, were transferred on the same day of ES cell injection; whereas 75 and 86 blastocysts were transferred in the morning of the day following the blastocyst injection. At E11, fetuses were recovered (Fig. 2) (Fig. 3).
Pup chimaerism rate improved following exposure to FGF4 We tested subsequently whether the high proportions of ES cell contribution in E11 fetuses obtained from embryos that were cultured in KSOM supplemented with FGF4 (1000 ng/ml) and heparin (1 lg/ml) also translated in the birth of highly chimaeric offspring. Various hybrid ES cell lines (129S2/SvHsd/C57BL/ 6JOlaHsd (n = 2), 129S2/SvHsdCastEij and CastEij/ C57BL/6JOlaHsd) were injected into embryos that were cultured either in KSOM or KSOM supplemented with FGF4 (1000 ng/ml) and heparin (1 lg/ ml). Since previous fetal data had shown that the efficiency per transferred blastocyst is rather low when ET occurs at E4.5, all embryos were retrieved at E2.5 and transferred the same day of the blastocyst injection (E3.5). Significantly more pups, were chimaeric (coat colour) ( Table 2) (v 2 , p value\0.02) when transferred blastocysts had been cultured in KSOM supplemented with FGF4 (84%) (1000 ng/ml) and heparin (1 lg/ml) instead of regular KSOM (53.8%). The pups were naturally delivered without c-section and most of them reached adulthood. All the chimaeric mice that were set up for breeding tested positive for germline transmission in the first litter obtained. The percentage of pups relative to the number of transferred blastocysts was lower than reported for the E11 fetuses. Nevertheless, the fetal loss was similar in both tested groups and can therefore not be attributed to FGF4 supplementation to the culture medium. Figure 4 summarizes the percentage of chimaeric fetuses and pups relative to the total amount of fetuses and pups. The contribution of ES cells to offspring (chimaeric as assessed via coat colour contribution) was determined, for those ES cell lines carrying a GFP reporter (129S2/SvHsd/C57BL/6JOlaHsd (n = 1) and Cas-tEij/C57BL/6JOlaHsd), by quantifying GFP contribution in the lungs, heart and brain, As such, 9 chimaeric offspring from the control group were compared to 15 chimaeric offspring from the treated group. The contribution of GFP was estimated by genomic QPCR using primers for GFP and primers for H2A as a reference. For lungs, heart and brain of chimaeric offspring created following the conventional way, the median GFP contribution was 53, 62 and 43%. For offspring derived from blastocysts cultured in KSOM supplemented with FGF4 (1000 ng/ml) and heparin (1 lg/ml) these results amounted to 83, 80 and 23% (Fig. 5). The results show that the rate of chimaerism in lung and heart of adult offspring is higher when host blastocysts were exposed to FGF4 (1000 ng/ml) and heparin (1 lg/ml). Our results obtained with brain, in contrast, show that the high ES cell contribution in lungs and heart cannot be extrapolated to the brain.
According to the data from the E11 fetuses, 77% of the fetuses were entirely chimaeric as opposed to 20% in the untreated group. These data are better than the reported numbers of entirely E9 transgenic fetuses conceived following injection of ES cells into precompaction preimplantation embryos (Cox et al. 2010). Most chimaeric recovered pups from both experimental conditions displayed a relative high rate of chimaerism as most of them presented a uniform brown colour. The fact that the utilized ES cell lines had a low passage number and were derived and cultured in 2i conditions may provide an explanation for the high contribution efficiency. This high rate also resulted in germline transmission in all chimaeric animals including the animals belonging to the control group. Since the coat colour of the chimaeric mice did not provide discernable differences between both experimental groups, chimaerism was subsequently quantified by establishing the GFP contribution using quantitative RT-PCR. We found that the standard deviation of this last technique was large, and these results should therefore considered as an indicator instead of an exact quantifier. Median values of GFP in lungs and heart contribution approximated 80% when host blastocysts were exposed to FGF4 (1000 ng/ml) and heparin (1 lg/ml) versus 60% in controls. Surprisingly, GFP contribution in brain tissue, which is ectoderm derived, showed the lowest rate of chimaerism with a decreasing trend for animals belonging to the treated group. The latter results are difficult to explain as the coat colour of the mice indicated a high rate of chimaerism and no visual differences in GFP contribution in the head of treated and control E11 fetuses was observed. The reversion of some primitive endoderm biased host ICM cells to the pluripotent fate may explain this finding as these cells are more likely to be found adjacent to the remaining primitive endoderm cells. Indeed, the primitive endoderm forms during further development the visceral endoderm which induces the formation of the neurectoderm from the bordering epiblast cells. Alternatively, the neurectoderm may contain cells with an extra-embryonic origin. ES cell lines carrying specific quantifiable markers should be used in order to quantify the exact rate of chimaerism during subsequent experiments. These experiments would also be helpful to test other ES cell lines in general, since the technique has only been validated for one inbred and four hybrid ES cell lines. In addition, it should be highlighted that the most effective concentration of FGF4 and heparin, needed to bias the host ICM towards PE without compromising viability, will most Chemical induced segregation of the ICM towards PE via FGF4 provides an easy and more feasible alternative to generate mice with high rates of chimaerism in laboratory settings where either tetraploid complementation or ES cell injection into cleavage stage embryos is not effective. Pups can be born via natural delivery and a large percentage of pups are entirely transgenic for tissues from the mesodermal and endodermal lineage. Since the presented technique only relies on biasing cell fate, in contrast to a progressive negative developmental selection of unviable 4 N cells, host embryos in the presented method remain viable. Consequently, Efficiency was calculated in relation to the amount of injected embryos transferred () and in relation to the amount of fetuses retrieved (()) Fig. 4 Summary of percentage of chimaeric E11 fetuses and pups. E11 fetuses and pups resulting from blastocysts injected with ES cells (various ES cell lines) respectively compared to the total amount of fetuses and pups. Host embryos were either cultured in KSOM or KSOM supplemented with FGF4 (1000 ng/ml) and heparin (1 lg/ml) prior to and after injection with ES cells on E3.5 and transferred within a few hours (2-8 h) following injection. Significance tested via v 2 statistical test whereas no pups will be born via tetraploid complementation if the quality of the utilized ES cells is not good, pups with no or low levels of chimaerism can be born when using the current presented method. Additionally, in contrast to conceptuses created through tetraploid complementation that display high rates of host contamination at implantation (Eakin et al. 2005), the current method provides conceptuses that are merely transgenic from implantation onwards, which makes this technology a very powerful tool to study gene mutations at different stages of embryonic development and in adults.