Formation of a thymus from rat ES cells in xenogeneic nude mouse↔rat ES chimeras

Authors


  • Communicated by: Hiroshi Hamada

okabe@gen-info.osaka-u.ac.jp

Abstract

Various conditions for differentiating embryonic stem (ES) cells or induced pluripotent stem (iPS) cells into specific kinds of cell lines are under intensive investigation. However, the production of a functional organ with a three-dimensional structure from ES or iPS cells is difficult to achieve in vitro. In the present paper, we describe the establishment of a green fluorescent protein-expressing rat ES cell line and production of mouse↔rat ES chimera by injecting rat ES cells into mouse blastocysts. The rat ES cells contributed to various organs in the chimera, including germ cells. When we injected ES cells into blastocysts of nu/nu mice lacking a thymus, the resultant chimeras produced thymus derived from rat ES cells in their bodies. The chimeric animals may provide a method for the derivation of various organs from ES or iPS cells.

Introduction

Chimeric animal production was initially reported as having been achieved through aggregation of two independent embryos of the same species (Tarkowski 1961; Mintz 1967). Later, chimeras between different species (Rossant & Frels 1980) or different genera were attempted in various laboratories (Zeilmaker 1973; Meinecke-Tillmann & Meinecke 1984). The first intergenera chimeric animal developing to term was derived from aggregating an uneven number of sheep and goat eight-cell embryos (3 : 1 or 1 : 3) and transferring them to a sheep or goat mother, respectively (Fehilly et al. 1984). However, the rat↔mouse chimera were reported to be embryo lethal (Gardner & Johnson 1973, 1975). The combination of a trophoblast and a recipient mother’s genotype is considered to be a critical factor for the success of interspecific murine chimeras (Rossant et al. 1982; Fehilly et al. 1984). Thus, one of the difficulties implied in the formation of rat↔mouse aggregation chimera might stem from a possible contribution of cells from different species in the placenta. The usage of ES or iPS cells can circumvent this potential problem, as shown in an accidental formation of mouse↔rat chimera (Brenin et al. 1997). ES cells are destined to contribute to the embryo proper, but not to trophoblast cells (Beddington & Robertson 1989).

In the present experiment, we report the establishment of ubiquitous promoter-driven, green fluorescent protein (GFP)-tagged ES cells from DA rat strain and the formation of mouse↔rat chimera by injecting rat ES cells into mouse blastocysts. For further application, we tried to produce an organ derived from rat ES cells in the xenogeneic chimera. Intensive research on differentiation of ES or iPS cells into specific cell lineages continues, but production of an organ with a three-dimensional structure has proved to be a challenge in vitro. In the present paper, we describe the formation of a thymus of rat ES origin with a combination of thymus lacking nu/nu-mouse blastocysts.

While we are preparing the present paper, a success in producing pancreas using a similar approach with rat iPS cells was reported by another group (Kobayashi et al. 2010). Thus, this is the second report of organ formation from ES or iPS cells. We report a formation of thymus from rat ES cells by making a chimera with nu/nu mice. Differing from the precedent xenogeneic chimera paper, a contribution of ES cells into germ cells was suggested.

Results

Production of green ES cells from rat

We established 14 rat ES cell lines from 38 blastocysts from the dark agouti (DA) rat (purchased from Japan SLC) having an agouti coat color by culturing in N2B27 containing 2i as reported previously (Buehr et al. 2008; Li et al. 2008) (Fig. 1A). One of the established cell lines, rES-DA1Osb (DA1), was electroporated with plasmid carrying CAG–EGFP and a puromycin selection marker and one of the green ES cell lines cloned [rES-DA1-Tg(CAG–EGFP)1Osb (DA1-GB1)] was used for the following experiments. Figure 1B,C shows the expression of stem cell markers such as alkaline phosphatase, Oct4, Nanog, Sox2 and SSEA1 in DA1-GB1 cell line at passage 11. The chimera-forming ability was examined by injecting ES cells into blastocysts of albino F344/N rat (Fig. 1D). Thus, prepared chimeric rats were fertile and the agouti pups (including ‘green fluorescent’ pups) were born after mating with albino Wistar rats, indicating the successful germ-line transmission from the ES cells (Fig. 1E).

Figure 1.

 Establishment of a green fluorescent rat ES cell line and its germ-line transmission. (A) A colony of originally established rat ES DA1. (B) Green fluorescent protein (GFP)-expressing DA1-GB1 produced from DA1. (C) Expressions of stem cell markers in the colonies of DA1-GB1. The scale bar represents 100 μm. (D) Ten-day-old chimeric rats obtained by injection of DA1-GB1 rat ES cells into F344 rat blastocysts. The agouti coat color denotes the presence of DA1-GB1 rat ES cells against an albino F344 background. (E) The pups obtained by mating male chimeras with female Wistar rat of albino strain. The pigmented offspring indicate the germ-line transmission of the DA1-GB1 rat ES cells. The bold white arrow shows germ-line offspring with GFP transgene and bold black arrow without GFP transgene.

Production of mouse←rat ES chimera

We used this germ-line-competent DA1-GB1 cell line for the production of a mouse←rat ES chimera (here, the ‘←’ indicates the rat ES cells were injected into mouse blastocysts to make chimera (Fig. 2A,B)). The mouse←rat ES chimeric embryos were transplanted into the uterus of pseudo-pregnant female mice. A previous paper indicated that the rat↔mouse chimeric embryos survive longer when E4.5 rat inner cell masses (ICM) were injected into the E3.5 mouse blastocysts (Gardner & Johnson 1973). Our mouse←rat ES chimera method improved the precedent method, and we obtained live pups on day 19.5 pc, which is full term for mouse and 2 days short for rat. On the contrary, the rat←mouse ES chimeras using rat blastocysts and mouse ES cells transplanted into pseudo-pregnant rats were born on day 21.5 pc (data not shown). The contribution of rat cells in the mouse←rat ES chimeras ranged from negligible to at most 30–40%, estimated by their coat color. The resulting mouse←rat ES chimeric pups appeared healthy, were nursed by a mother mouse and grew like mice (Fig. 2D,E). Even when coat color showed that over one-third of the cells originated from rat, the body sizes remained in a slightly heavier range than wild-type mice (Fig. 2F). To analyze the contribution of rat cells in the mouse←rat ES chimera, we examined the GFP fluorescence and found GFP-positive cells in brain, liver, heart, kidney, lung, pancreas, thymus, spleen and testis (Fig. 3A). When we observed the organs, we found parts where the rat cells exclusively contributed to keeping the murine size. We assumed that mouse and rat cells form their organs according to their predesignated size, but mouse and rat cells seem to cooperate to form naturally shaped chimeric organs as shown in Fig. 3.

Figure 2.

 Production of mouse←rat ES chimera. (A) Strategy for the production of mouse←rat ES chimera. Two to six rat ES cells were injected into each mouse day 3.5 blastocyst and were transplanted to pseudopregnant mice immediately. The resultant chimeras were raised by foster mice. (B) Incorporation of rat DA1-GB1 ES cells into ICR mouse blastocysts incubated in vitro for 1 day after injection. (C) E10.5 heart beating chimeric embryo with high contribution of rat ES cells. Green fluorescence indicates green fluorescent protein signals from rat ES cells. (D) Bold arrows indicate the chimeric pups on day 3 postpartum judged by the green fluorescence. The empty arrow indicates the nonchimeric littermate. (E) The 5-week-old mouse←rat ES chimera. Agouti coat color originated from DA1-GB1 rat ES cells and the white coat color originated from ICR mouse strain. (F) The body weight of male mouse←rat ES chimeras at 10 weeks of age [50 ± 8 g (n = 15)] indicated as M←rES was slightly heavier than that of ICR male mice of the same age [42 ± 4 g (n = 8)] (mean ± SD). However, the difference was not statistically significant and was much smaller than the estimated body weight. (Used mice and rats at 10 weeks of age weigh 43.7 and 199 g as an average, respectively. If rat body occupies 30–40% of the chimeric body, the theoretical body weight could be somewhere from 90 to 106 g). The black transverse lines indicate the average. (G) #1–6 were mouse←rat ES chimeras identified from their agouti coat color. Tail tip genome from these chimeras was amplified by mouse Gapdh-(mGapdh)- or rat Gadph-(rGapdh)- specific primer sets. The contribution of rat ES cells was shown except in #2 chimera, where the rat ES cells failed to contribute in the tail tip.

Figure 3.

 Contribution of rat ES-derived cells in various organs in chimera. (A) Comparative view of whole organs excised out from 1-day-old mouse←rat ES chimera (right) and from the ICR mouse (left). The scale bar represents 2.5 mm. The ‘testis’ indicates the testicular section of the 2-day-old mouse←rat ES chimera. Both mouse-(indicated by arrowhead) and rat-(indicated by arrows) derived cells were found in seminiferous tubules. The blue signals indicate nuclei stained by Hoechst33342. The scale bar represents 50 μm. (B) Testicular cells were squeezed out from the seminiferous tubules of the 13-week-old mouse←rat ES chimera (REF Palvinen, Nature Method) and were observed under phase contrast microscope (middle). Both mouse (arrowhead) and rat sperm (arrow) were observed among the cells. Germ-line cells from mouse and rat seminiferous tubules prepared in the same manner are indicated in the left and right panels, respectively. Note the difference of sperm head shape. The scale bar represents 10 μm. (C) Mouse (arrowhead) and rat (arrow) sperm could be separated by the propidium iodide (PI) staining after freezing and thawing of collected epididymal sperm.

Contribution of ES cells into germ-line cells in chimeras

The fertility of mouse←rat ES chimeric animals is difficult to define, but the chimerism in brain seems not to affect the sexual behavior of the recipient mice, and we found that male chimeras mated with female mice and showed fertilizing ability and, alternatively, female chimeras mated with male mice and delivered mouse pups. Recently, another group also reported chimeras by injecting rat iPS cells into mouse blastocysts (Kobayashi et al. 2010), and they described the no-germ-line contribution of iPS cells in their chimeras. However, in our case, the ES cells contributed to the testis (Fig. 3A), and rat sperm was observed in the mouse←rat ES chimeric testis (Fig. 3B). Although the rat sperm population was not large enough to be detected as a distinguishable peak by the fluorescent-activated cell sorter analysis, we could find rat sperm in the epididymis of mouse←rat ES chimeras (Fig. 3C).

Production of rat thymus in a chimeric environment

We thought one possible application of mouse←rat ES chimeric animals might be organ formation from ES (or iPS) cells as Kobayashi et al. (2010) independently reported. To this end, it is necessary to prepare animals lacking an organ as an ES recipient species used for chimera formation. We chose nude mice (nu/nu mice) that lack thymus as recipient blastocysts for rat ES cells. It was reported that thymic epithelial cells in aggregation chimeras (B6-nude↔BALB/c and B6↔BALB/c-nude) consisted of thymic epithelial cells exclusively from the non-nude donor (Martinic et al. 2003), indicating that cells from nu/nu mice cannot differentiate into thymic epithelial cells. By injecting rat ES cells, we could observe formation of thymus in the nu/nu-mouse←rat ES chimera (Fig. 4). Although the size was smaller than the normal mouse thymus, the thymi showed a typical three-dimensional cell structure (Fig. 4B,C). In our nu/nu-mouse←rat ES chimera, the lymphocytes were dominated by mouse-originated cells and we were not able to demonstrate the presence of CD4+ or CD8+ single-positive rat cells. Instead, mouse lymphocytes of CD4+ or CD8+ single-positive (with T-cell marker, CD3 positive) populations appeared in nu/nu-mouse←rat ES chimera (Fig. 4E,F).

Figure 4.

 Development of thymus in nu/nu-mouse←rat ES chimera. (A) Thymus in 1-week-old ICR (mouse), nu/nu-mouse (nu/nu) and nu/nu←rat ES chimera. The dotted circular line indicates thymus. H, heart; L, lung; GFP, green fluorescent protein. Thymic structure was observed in four of 25 nu/nu←rat ES chimeras. (B) Thymus of 1-week-old ICR (left) and the nu/nu-mouse←rat ES chimera (right). BF, bright field; HE, hematoxylin and eosin stained. (C) Typical thymic epithelium structure consisted of exclusively green fluorescent rat cells in nu/nu←rat ES chimera. The scale bar represents 100 μm. (D) 1-week-old nu/nu←rat ES chimera. (E) Dot plot analysis of anti-CD4 and anti-CD8 staining of splenocytes. mCD3, mouse CD3; mCD4, mouse CD4; rCD3, rat CD3; rCD4, rat CD4. (F) The mouse CD4+ or CD8+ splenocytes demonstrated in (D) were expressing T-cell marker, CD3.

To analyze whether the thymi formed in nu/nu-mouse←rat ES chimera were functional, the thymus was transplanted under the renal capsule of rnu/rnu rat lacking thymus. There were no CD4+ or CD8+ single-positive (with T-cell marker, CD3 positive) cells before implantation, but these cell populations appeared 8 weeks after the thymic transplantation into rnu/rnu rat (Fig. 5).

Figure 5.

 CD4+ or CD8+ single-positive cells in thymus-transplanted rnu/rnu rat. (A) A schematic diagram for the transplantation of thymus from 0-day postpartum nu/nu-mouse←rat ES chimera to rnu/rnu rat. (B) The CD3+CD4+ or CD3+CD8+ populations appeared after transplantation of thymus formed from rat ES to rnu/rnu rat.

Discussion

Precedent papers reported that the rat↔mouse chimeric embryos could not develop to term (Gardner & Johnson 1973, 1975). However, there is a report that trophoblast genotype is important for the survival of interspecific murine (Mus musculusMus caroli) chimeras (Rossant et al. 1982). Needless to say, the role of placenta in protecting the fetus from the mother’s immunological attack is very important. We thought if we injected ES or iPS cells into blastocyst and made chimeras, a placenta might be formed from a single species matching the mother, because ES and iPS cells are destined to contribute to the embryo proper, but not to trophoblast cells (Beddington & Robertson 1989; Brenin et al. 1997). As we expected, we could obtain viable chimeras in both mouse←rat ES and rat←mouse ES chimeras when transplanted to mouse and rat, respectively. We assume this could be an important rationale for the success of obtaining mouse←rat ES chimera.

Although production of xenogeneic chimeric animals from the embryonic stage was difficult, there are many papers reporting formation of a partial xenogeneic chimera in various combinations. Successful examples of organ transplantations from one species to another which survive and regrow over genera (or orders) are reported. For example, an interorder testis grafting, such as pig and goat neonate testis into mouse body (Artiodactyla to Rodentia), is reported to support complete spermatogenesis in mouse body (Honaramooz et al. 2002). Human liver cells (Sato et al. 2008) and human endometrial cells (Masuda et al. 2007) (Primate to Rodentia) are also known to be transplantable into mice. Thus, in general, the cells of different species are able to adjust to their environmental conditions rather flexibly and remain functional if the cells can avoid immunological attack.

Recently, another group also reported chimeras by injecting rat iPS cells into mouse blastocysts (Kobayashi et al. 2010), and they described the no-germ-line contribution of iPS cells in their chimeras. However, in our case, the rat ES cells contributed to the testis, and rat sperm was observed in the chimeric testis and epididymis as shown in Fig. 3B,C. This is natural because mouse testis supports rat spermatogenesis, as examined by transplantation of rat spermatogonia into mouse testis (Clouthier et al. 1996; Shinohara et al. 2006). Although the contribution of iPS cells to germ cells was negated in mouse←rat iPS chimera (Kobayashi et al. 2010), when the iPS cells share a similar characteristic nature with the ES cells, they were able to differentiate into germ cells (Okita et al. 2007). Thus, their notion may not be applicable for a basic rule.

In any case, developing an organ of a certain species in a body of a different species was demonstrated to be possible if suitable conditions are established: thymus in the present paper and pancreas in the precedent paper (Kobayashi et al. 2010). We have not carried out a detailed analysis of the function of rat ES-derived thymus and it is not clear how the mouse T-cells were ‘matured’ in a rat ES-derived thymus. However, there is a report that mouse T-cells in nu/nu mouse can be educated by the grafted rat thymi (Nishigaki-Maki et al. 1999). Moreover, there is a report indicating the T-cell maturation can be achieved independent from thymic epithelial MHC (Martinic et al. 2006). It would be interesting to study what is ‘self’ and what is ‘nonself’ in the chimeric mouse with a mature mouse T-cell educated by rat thymus.

Both mouse and rat are the most commonly used experimental animals and the mouse←rat ES chimeras are easy to produce. The chimera formation system described in the present paper will open the door for us to solve various fundamental questions in many biological fields, such as development (body size, shape, etc.), immunology (self and nonself) and nerve systems (instinctive actions), together with the clinical usage of producing an organ from ES or iPS cells in the future.

Experimental procedures

Animals

The handling and surgical manipulation of all experimental animals were conducted according to the guidelines of the Committee on the Use of Live Animals in Teaching and Research of Osaka University. DA/Slc rats for the establishment of rat ES cell lines, F344/N Slc, Wistar and SD rats for production of chimeric rats, were purchased from Japan SLC, Inc., and rnu/rnu rats (F344-nude; CLEA Japan, Inc.) were used for the thymus transplantation. ICR (Japan SLC, Inc.) and CD1-Foxn1nu (ICR-nude; Japan Charles River) mice were used for the production of mouse←rat ES chimeras.

Establishment and maintenance of rat ES cell lines

Rat ES cells were established and maintained in 2i containing medium as described (Li et al. 2008), with a slight modification. Briefly, rat blastocysts were gently flushed out from the uterus of E4.5 pregnant DA/Slc rats with R2ECM (Goh et al. 2000). After the removal of the zona with acid Tyrode’s solution (Sigma), whole blastocysts were transferred into gelatin-coated 60-mm dishes and cultured on MEFs with N2B27 medium supplemented with 3 mm CHIR99021 and 1 mm PD0325901 (Axon). After 7 days, the outgrowths of blastocysts were disaggregated by 1 mm EDTA–PBS(−) by pipetting. The cells were replated into 0.2% matrigel-(Becton, Dickinson and Company) coated dishes and cultured on SNL cells (Thomas & Capecchi 1987) with 2i containing medium. Emerging ES cell colonies were then trypsinized and expanded. Established rat ES cell lines were routinely maintained in 2i conditions at 37 °C in a humidified 5% CO2 incubator. The medium was changed daily, and cells were split with 0.05% trypsin every 2–3 days.

Establishment of GFP-expressing rat ES cells

Eight × 105 rat ES cells were electroporated in calcium- and magnesium-free phosphate-buffered saline [PBS(−)] with 10 μg PvuI linearized pPGKpuro-CAG–EGFP plasmid using the Bio-Rad Gene Pulser (0.25 kV, 0.5 mFD). The electroporated cells were plated into a 60-mm dish containing 5% FCS in 2i medium. The following day, the medium was changed to serum-free 2i. To clone the puromycin-resistant colonies, puromycin (1 μg/mL final concentration) was added to the medium 72 h after the electroporation. The resistant colonies were picked up and cultivated in 96-well plates for another 10 days to expand.

Alkaline phosphatase staining and immunostaining

Alkaline phosphatase staining was carried out with a Histofine Fuchsin Substrate kit for alkaline phosphatase (Nichirei) according to the manufacturer’s instructions. Immunostaining was carried out via standard protocols. Briefly, rat ES cell colonies were fixed in 4% paraformaldehyde. Fixed colonies were washed in PBS(−) and then blocked with BlockAce (Dainippon Seiyaku). Primary antibodies used include the following: Oct4 (H-134, 1 : 100; Santa Cruz), Nanog (ab80892, 1 : 100; Abcam), Sox2 (D-17, 1 : 50; Santa Cruz) and SSEA1 (MC-480, 1 : 200; R&D). Alexa Fluor fluorescent secondary antibodies (Invitrogen) were used at 1 : 500 dilution. Nuclei were stained by Hoechst 33342.

Production of rat ES chimeras and testing germ-line transmission

Blastocysts from E4.5 pregnant F344/N Slc rats were placed in kSOM medium (Ho et al. 1995), covered with mineral oil and incubated at 37 °C with 5% CO2 in air. Eight to 12 GFP-marked rat ES (DA1-GB1) cells (passage numbers between 9 and 11) were injected into each blastocyst and incubated in kSOM medium until transfer into uterus. A total of 67 chimeric blastocysts were transferred into the uterine horn of E3.5 pseudopregnant SD rat. Chimeric rats were recovered by natural delivery on 21.5 or by caesarian section on E21.5. All of the six chimeric rats were successfully weaned. Chimeric rats were identified by GFP fluorescence and by coat color. Germ-line transmission was tested by mating chimeras with Wistar rats.

Generation of mouse←rat ES chimera

Superovulated ICR or ICR-nu/nu females were mated with ICR or ICR-nu/nu males, respectively. Two- or four-cell stage embryos were collected from E1.5 pregnant mice and placed in kSOM medium, covered with mineral oil and incubated for 48 h at 37 °C with 5% CO2 in air. The well-expanded blastocysts were used for microinjection. Two to six DA1-GB1 cells (passage numbers between 11 and 12) were injected into each blastocyst and incubated at 37 °C in kSOM medium until transplantation into uterus. A total of 1279 blastocysts were transferred into the uterine horn of E2.5 pseudopregnant ICR mice. Mouse←rat chimera were recovered by natural delivery on E19.5 or by caesarian section on E19.5. Seventy-four live-born mouse←rat ES chimeras were obtained, among which seven pups were killed before weaning and 52 pups were successfully weaned. Mouse←rat ES chimeras were identified by GFP fluorescence and by coat color.

Genotyping

The genomic DNA was PCR amplified using the following primers: 5′-gccaaaagggtcatcatctccg-3′ and 5′-gtccactcatggcagggtaagataag-3′ for mGadph, 5′-ctggctcttgagagtaactgaagg-3′ and 5′-tggggactcctcagcaactg-3′ for rGadph. DNA fragments were amplified using Ex Taq (Takara) for 40 cycles under the following conditions: 94 °C for 30 s, 62 °C for 30 s and 68 °C for 30 s.

Histology

Chimeras were killed by cervical dislocation and the organs were fixed in 4% paraformaldehyde overnight, then rinsed with PBS(−) for 1 h, dehydrated in acetone for 1 h and embedded in glycol methacrylate (Technovit 8100; Heraeus Kulzer GmbH, Germany). Sections were prepared at a thickness of 5 μm. Fluorescent photomicrographs were obtained using a fluorescein epifluorescence filter set, and then the sections were stained with hematoxylin and eosin photographed again using bright field optics.

Preparation of epididymal sperm for FACS analysis

Epididymis from mouse, rat and mouse←rat chimera were cut in several places and placed in TYH medium (Toyoda et al. 1971) to disperse sperm. Each sperm suspension was stained by propidium iodide (final concentration was 10 μg/mL) after freezing and thawing. Multicolor flow cytometry analysis was carried out with FACS-Calibur (Becton Dickinson). Data were obtained with flowjo software. Rat sperm from mouse←rat ES chimera were sorted using FACS-Aria (Becton Dickinson).

Thymus transplantation

Thymus was aseptically dissected from 0-day-old nu/nu-mouse←rat ES chimera. The thymus was separated into two lobes and one lobe was transplanted into the right renal capsule of rnu/rnu rat. Eight weeks after transplantation to rnu/rnu rats, part of the spleen was dissected out and cells were subjected for the flow cytometric analysis.

Flow cytometric analysis

Under anesthesia, a part of the spleen was cut out from 3-week-old mouse, rat, nu/nu-mouse and nu/nu-mouse←rat ES chimeras, and adult rat, rnu/rnu rat and thymus-transplanted rnu/rnu rat and the cells were dispersed and were analyzed by a flow cytometer. Single cell suspensions from spleen were stained with the following antibodies for 30 min at 4 °C. The antibodies used were APC-labeled anti-mouse CD3ε (145-2C11, 1 : 50; BioLegend), anti-mouse CD4 (RM4-5, 1 : 50; BD Pharmingen) and anti-rat CD4 (OX-35, 1 : 50; BD Pharmingen), PerCP-labeled anti-mouse CD3ε (145-2C11, 1 : 50; BioLegend), anti-mouse CD8a (53-6.7, 1 : 50; BD Pharmingen) and anti-rat CD8a antibody (OX-8, 1 : 50; BioLegend), and biotin-conjugated anti-rat CD3 antibody (G4.18, 1 : 50; BD Pharmingen). The biotinylated antibody was detected by Cy2-labeled streptavidin (PA42001, 1 : 100; GE Healthcare) for 30 min 4 °C. Multicolor flow cytometry analysis was carried out with FACS-Calibur (Becton Dickinson). Data were obtained with FlowJo software for viable cells that were determined based on measurements of forward and side light scatter intensity and propidium iodide exclusion.

Acknowledgements

We are very grateful to Adam M. Benham for helpful discussion, critical reading of the manuscript and useful advice. This work was supported in part by grants from the Ministry of Education, Science, Sports, Culture and Technology of Japan.

Ancillary