An interspecies barrier to tetraploid complementation and chimera formation

To study development of the conceptus in xenogeneic environments, we assessed interspecies chimera formation as well as tetraploid complementation between mouse and rat. Overall contribution of donor PSC-derived cells was lower in interspecies chimeras than in intraspecies chimeras, and high donor chimerism was associated with anomalies or embryonic death. Organ to organ variation in donor chimerism was greater in interspecies chimeras than in intraspecies chimeras, suggesting species-specific affinity differences among interacting molecules necessary for organogenesis. In interspecies tetraploid complementation, embryo development was near normal until the stage of placental formation, after which no embryos survived.

placenta, the embryo proper cannot develop normally with 4N cells. Tetraploid complementation is therefore a technique that restricts the 4N host embryo to extraembryonic lineages, creating an "embryo proper niche" that can be filled with donor PSCs 9 . The offspring generated by tetraploid complementation are distinct from chimeras-although the conceptus is chimeric, the embryo proper is derived from a single genetic background.
Limitations of interspecies chimeras, including cell contribution and tissue distribution, have not been thoroughly explored. Identifying these limits will not only highlight the temporal and spatial occurrence of a xenogenic barrier, but may also provide clues regarding which cell types and organs are most amenable to interspecies chimerism and organogenesis. Here, we first use tetraploid complementation to explore development of PSC derived embryos with xenogenic extraembryonic tissues. We then probe the functional limits of various PSCs regarding their ability to contribute to intra-and interspecies chimeras across multiple tissues and developmental timepoints.

Development of PSC-derived embryos in a xenogenic environment.
To examine the developmental potential of PSC-derived embryos in a xenogenic environment, we first injected rat PSCs (rPSCs) into mouse tetraploid embryos.
After embryo transfer, we assessed development of the PSC-derived embryos at several stages. Injection of enhanced green fluorescent protein (EGFP) expressing rat iPSCs (T1-3) yielded, at embryonic day [E] 6.5, epiblasts that were composed entirely of EGFP + rat iPSC-derived cells and were surrounded by EGFP − mouse extraembryonic tissues (Fig. 1A). At E9.5, embryos enclosed in mouse extraembryonic tissues expressed EGFP ubiquitously from head to tail (Fig. 1A). 12 live embryos were curved ventrally, with red blood cells in the aortic region ( Fig. 1A and Table 1) and those embryos were correspond to E10.5 to 11 stage of rat embryo based on their forelimb bud morphology. Immunohistological study and flow cytometric analysis clearly showed that these embryos were entirely composed of EGFP + rat iPSC-derived cells (Fig. S1A,B). Immunofluorescent staining of embryonic tissue with antibodies against markers for the three representative germ layers revealed that rat iPSC-derived embryos had beta III-tubulin + neural cells, forkhead box protein A2 (FoxA2) + gut-endoderm cells, and platelet endothelial cell adhesion molecule 1 (PECAM-1) + mesoderm-oriented blood vessels (Fig. S1C). These data strongly suggest that rat iPSC-derived cells had normally differentiated into various types of tissues even in a mouse environment until E9.5.
To check for further developmental potential of rat PSCs in mice, we analyzed embryos at E11.5-13.5. However, no live embryos were found at E11.5 ( Fig. 1B and Table 1). At this timepoint the embryos exhibited substantial tissue degeneration (Fig. S1D) when compared with E11.5 embryos generated by injection of rat PSCs into rat tetraploid embryos 10 . On the other hand, the mouse tetraploid embryos could support full term development with mouse PSCs, confirming their developmental potential in an intraspecies setting (Fig. S2). This indicated that the mouse placental/maternal environment was not competent to support development of rat PSCs beyond E9.5.
We next attempted to develop mouse embryos in a rat environment by injection of tetraploid-competent mouse PSCs into rat tetraploid embryos. Although we could not obtain live pups with interspecies tetraploid complementation compared with intraspecies ( Table 2 and Fig. S2B), mouse PSC-derived embryos were able to develop until E11.5, were curved ventrally with red blood cells in the aortic region and those embryos were correspond to E9.5 stage of mouse embryo based on their forelimb bud morphology (Fig. 1C). Moreover, when we analyzed embryos at E14.5, we found further developed mouse ESC-derived embryos in rat (Fig. 1D).
These embryos were mostly morphologically normal and showed pulsation of the heart (data not shown). These data clearly indicate that rat tetraploid embryos can support the development of mouse PSC-derived embryos until the E14.5 stage.
We followed the development of mouse embryos in the rat environment for three more days, however, we could not detect any live mouse embryos at E17.5 ( Fig. 1D and Table 2). This limitation resembles that encountered with rat PSC-derived embryos in the mouse environment.
Ability of PSCs to contribute to interspecies chimeras. To investigate the limits of mouse and rat PSC contributions to forming interspecies chimeras, we generated rat-mouse interspecies chimeras by injecting rat or mouse PSCs into mouse or rat blastocysts and analyzed whole-embryo chimerism at several different developmental stages ( Table 3). The average chimerism of E9.5 embryos generated by injection of rat PSCs into mouse blastocysts was 24.3% using ESCs and 52.7% using iPSCs. In both groups, chimerism declined to less than 11% as the developmental stage advanced ( Fig. 2A) as did the chimera survival rate. At E9.5, survival was 57% for the ESC group and 75% for the iPSC group, falling to less than 25% at E14.5 in both group (Fig. 2B).  (7) E14.5 101 Morphological analysis revealed normal development at E9.5 in 50% of chimeric embryos generated by injection of rat PSCs into mouse blastocysts. However, development was abnormal in almost all embryos (88%) at E11.5, with highly chimeric embryos tending to be malformed (Fig. 2E). By contrast, average chimerism of embryos and survival rate of chimeras generated by injection of mouse PSCs into rat blastocysts were high (28% and 79% respectively) at E15.5 (Fig. 2C,D) and the chimeric embryos were morphologically normal (Fig. 2F). These results indicate that mouse and rat PSCs have similar abilities to contribute to chimera formation when assessed in early embryonic interspecies chimeras, and that mouse PSCs maintain this contribution level during the prenatal period. However, interspecies chimeras with a high contribution of rat PSCs were absorbed as development advanced.

Developmental capacity of PSC derived cells in xenogenic organs.
We also analyzed chimerism in various organs at E14 to E15. In embryos generated by injection of rat ESCs into mouse blastocysts, relatively high chimerism was observed in heart (13.1 ± 4.0%), lung (19.0 ± 4.1%), and intestine (19.4 ± 3.1%). The same trend was observed in rat iPSC-injected chimeric embryos (heart 6.5 ± 2.9%, lung 8.5 ± 4.3%), and intestine 8.3 ± 0.9%) (Figs 3A and S3A). To determine if the rat PSC distribution differences among organs was caused by aberrant methylation patterns in the rat PSCs 10 , we injected rat ICMs into mouse blastocysts and analyzed chimerism in various organs of the embryos produced. Although overall chimerism was higher than in embryos generated by injecting rat PSCs, contribution patterns were similar except for brain (Fig. S4).
These data indicate that developmental limitations of PSCs in interspecies chimeras differ in each organ, and that the contribution pattern varies with species of injected PSC and the corresponding interspecies blastocyst. Moreover, trends in contribution rates of rat PSCs to each organ except the brain were not due to aberrant methylation patterns.
On histopathologic examination, lymphoid cell infiltration and dilated capillaries were present in malformed digits and tails. As no bone abnormality was observed, the malformations were ascribed to synovitis (Fig. 4B). In skin, dyskeratosis, blister formation, acanthosis, lymphoid infiltration and hypertrophy of the epidermis indicated dermatitis (Fig. 4B). Urine flow testing, with microscopy of the ureter and the bladder, revealed that nephromegaly was caused by hydronephrosis induced by ureteral obstruction (Fig. 4F). To examine whether ureteral obstruction was caused by existence of xenogeneic cells in nephric duct during the development, we generated interspecies chimera by injection of Pax2 and Pax8 double knockout ES cells into Rat blastocyst. Previous reports showed that Pax2 and Pax8 double deficient mouse embryo were unable to form nephric duct because of deficient of the mesenchymal-epithelial transitions of the intermediate mesoderm 11 . Immunofluorescent staining of nephric duct with antibodies against markers for transitional epithelium showed that Pax2 and Pax8 double mutant cells did not contribute transitional epithelium in interspecies chimera. Urine flow testing revealed that no ureteral obstruction was observed in Pax2 and Pax8 double mutant ESCs injected interspecies chimera (Fig. 4F). These results indicate that developmental defects in interspecies chimera was caused by existence of xenogenic cells during the development.  Table 3. Results of intraspecies chimera formation by injection of rat or mouse PSCs into mouse or rat blastocysts. a Frequencies of embryos (%) were determined by dividing the number of implanted embryos by the number of embryos. b Frequencies of PSC-derived live embryos or pups (%) were determined by dividing the number of implanted embryos by the numbers of PSC-derived live embryos or pups.

Discussion
In this study, we demonstrated the existence of a barrier to interspecies tetraploid complementation and to interspecies chimera formation. In mouse and rat development, placenta formation begins around E9.5 and ends by E14. During this time, the placenta acquires the important role of sustaining late-stage embryo development. In both the mouse and rat interspecies tetraploid complementation experiments, the embryo proper could develop until at least E9.5.  It was not until after initiation of placentation that the xenogenic barrier became explicit, ultimately resulting in embryonic death.
In our interspecies tetraploid complementation setting, the rat placenta formed nicely by E14.5 as shown in Fig. 1C. The size and blood flow of the placenta appeared normal, however it did not support development of late-stage mouse embryos. Whole embryo culture in vitro is similarly limited, sustaining development without the placenta to around E11-E13.5 12 . Therefore, mouse embryonic death in the case of interspecies tetraploid complementation is likely due to inappropriate function of the placenta. Functional aspects such as exchange of gases and supply of nutrition or other supportive factors are likely insufficient, particularly if those factors are species-specific.
By contrast, when rat iPSCs were injected into mouse tetraploid blastocysts, none of the rat embryos survived to E11.5. Placental function therefore cannot be assessed at this timepoint, as the placenta is not fully developed. Regardless, early incompatibilities among the developing xenogenic placenta and embryo proper could be the cause of embryonic lethality since development is normal until the onset of placentation.
Similar developmental arrest and abnormalities were found in interspecies chimeras generated by injecting rat PSCs into mouse blastocysts. As shown in Supplementary Fig. 4, chimerism was higher when rat ICM cells instead of rat PSCs were injected into mouse blastocysts. Although these rat PSCs generated higher chimerism when injected into rat blastocysts, these data suggest that they lost or failed to utilize some of the properties of ICM cells necessary in the xenogenic environment. One possible reason is an aberrant epigenetic status of the injected rat PSCs. Methylation patterns of differentially methylated regions in a normally developing rat differ not only from rat PSC-derived embryos, but also from the PSCs themselves 10 . Aberrant methylation status in mouse ESCs also is associated with abnormalities and lethality in chimeric mice 13 . This phenomenon in rat PSCs might also be developmentally lethal in the xenogenic environment.
Interestingly, we found drastic organ-to-organ variation in donor chimerism. Environmental factors thus must influence differentiation of xenogenic PSC-derived cells. These data suggest species-specific affinity differences among interacting molecules necessary for development of germ cells.
Various malformations and anomalies were observed in interspecies chimeras. The frequencies of malformation and anomalies appeared proportional to the level of donor chimerism. Because these malformations were observed even when ICM cells were injected, and because malformations were rare in intraspecies chimeras and no nephric duct abnormality was observed when the nephric duct was not interspecies-chimeric, this was likely not due to abnormalities in PSCs but instead to xenogenic incompatibilities such as inappropriate timing and site for cellular interactions, mismatched ligand-receptor interactions, affinity differences in adhesion molecules, and other cell intrinsic or extrinsic factors.
The malformations and anomalies observed in these chimeras may therefore provide important clues to understanding organogenesis at the molecular level. Comprehensive understanding of these molecular pathways could not only unlock mysteries to evolutionary and developmental biology, but also provide tools for regenerative medicine and ultimately human organogenesis.

Materials and Methods
Animals. All mice and rats were maintained under specific pathogen-free conditions. All animal experiments were performed in accordance with the guidelines of the University of Tokyo and the National Institute for Physiological Sciences, Japan.
All animal and recombinant DNA experimental protocols were approved by The University of Tokyo Animal Care and Use Committee (no. A16-22, A16-61 and A16-66).
For preparation of donor ICM, immunosurgery of rat blastocysts was performed as described 18 . Briefly, the zonae pellucidae were removed from E5.5 rat blastocysts with acidic Tyrode's solution (Sigma-Aldrich, St. Louis, MO, USA). 20% anti-rat whole serum (ROCKLAND Immunochemicals INC. Limerick, PA, USA; 112-4101)/ DMEM (Sigma-Aldrich) for 3 hr. The blastocysts were washed once with DMEM and incubated for 20 min in 100% rat serum as the complement source. The trophectoderm cells were removed by gentle pipetting.
Tetraploid complementation. Tetraploid complementation was performed as described 10,15 . Briefly, for production of tetraploid embryos, 2-cell stage diploid embryos were collected in M2 medium (Merck Millipore, Darmstadt, Germany, MR-015P-D) from oviducts of mice 1.5 days postcoitum (dpc). These embryos were washed 3 times with medium containing 0.01% polyvinyl alcohol (Sigma-Aldrich; P8136), 280 mM mannitol (Sigma-Aldrich; M4125), 0.5 mM Hepes (Sigma-Aldrich; H4034), and 0.15 mM MgSO 4 (Wako, Osaka, Japan; 131-00405). Electrofusion of blastomeres to produce tetraploid embryos was carried out using a DC pulse (100 V/mm, 30 µsec, 1 time) followed by application of AC pulses (5 V/mm, 10 sec) using an ECM 2001 (BTX, Holliston, MA). These tetraploid embryos were transferred into potassium simplex optimized medium with amino acids (KSOM-AA) (Merck Millipore; MR-020P-D) and were cultured for 24-48 h for 4-cell/morula or blastocyst injection. and 8 mutated mouse ESCs into rat blastocyst (lower). Sections were stained with antibodies against e-cadherin (red) and EGFP (green). Cell nuclei were stained with DAPI. Muscle layer (M), lamina propria (L) and transitional epithelium (T) are abbreviated as indicated. Scale bars: 50 um (F) Urine flow testing of interspecies chimera generated by injection of wild type mouse ESCs into rat blastocyst (left) and interspecies chimera generated by injection of Pax2 and 8 mutated mouse ESCs (right). Lower column represents the results of flow testing (number of chimera with ureteral obstruction/number of total chimera). For micro-manipulation, ESCs/iPSCs were trypsinized and suspended in their culture medium. A piezo-driven micro-manipulator (Prime Tech, Tokyo, Japan) was used to drill the zona pellucida under the microscope and 5-10 cells were introduced into the perivitelline space of 4-cell/morula stage tetraploid embryos. ICMs were introduced into blastocyst cavities near the inner cell mass. After injection, tetraploid embryos underwent follow-up culture to blastocyst stage. The blastocysts then were transferred into the uteri of pseudopregnant recipient ICR female mice (2.5 dpc).
Intraspecies and interspecies chimeras. Generation of intraspecies and interspecies chimera was performed as described 8,16 . Briefly, mouse 8-cell/morula stage embryos were collected in M2 medium from oviduct and uterus of mice 2.5 dpc. These embryos were transferred into KSOM-AA and were cultured for 24 hr for blastocyst injection.
Wild-type rat blastocysts were collected in a bicarbonate-buffered medium composed of Roswell Park Memorial Institute medium (RPMI) 1640, Eagle's solution, and Ham's F12 containing 18% FBS medium 19 from oviduct and uterus of rats 4.5 dpc. These embryos were transferred into modified rat 1-cell embryo culture medium 20 containing 80 mM NaCl (Wako) and 0.1% polyvinyl alcohol (Sigma) and were cultured for about 1 hr until injection.
For micromanipulation, ESCs or iPSCs were trypsinized and suspended in ESC or iPSC culture medium. A piezo-driven micromanipulator (Prime Tech, Tokyo, Japan) was used to drill the zona pellucida and trophectoderm under the microscope and 5-10 ESCs or iPSCs were introduced into blastocyst cavities near the inner cell mass. After blastocyst injection, embryos underwent follow-up culture for 1-2 hr. Mouse blastocysts then were transferred into the uteri of pseudopregnant recipient ICR female mice (2.5 dpc) and rat blastocysts were transferred into the uteri of pseudopregnant recipient Wistar female rats (3.5 dpc). To generate interspecies chimeras for morphologic analysis, we used blastocysts obtained from wild type mice, wild type rats, and Pdx1 heterozygous mutant rats 21 .