Resurrection of chromosomes from frozen animals by single chromosome transfer into mouse oocytes

Abstract

Reviving extinct animals offers a crucial opportunity to recover lost or unknown genetic resources, yet cloning methods are unsuitable because they depend on intact donor nuclei and abundant oocytes or recipients from closely related species. To overcome these constraints, we explored a chromosome level revival strategy. Blood cells from rat carcasses stored at − 30 °C for over one year were introduced into enucleated mouse oocytes, where the rat nuclei underwent premature chromosome condensation. Microtubule polymerization inhibition enabled dispersion of rat chromosomes within the ooplasm, allowing isolation of individual chromosomes by micromanipulation. Each chromosome was subsequently transferred into an intact mouse oocyte, followed by intracytoplasmic sperm injection using GFP-transgenic mouse sperm. Embryos were cultured to the blastocyst stage, yielding 17 ES cell lines, two of which carried 41 chromosomes. Spectral karyotyping confirmed the presence of rat chromosome 9 alongside a full set of normal mouse chromosomes. These ES cells generated chimeric mice exhibiting GFP based chimerism across multiple organs. Histological analyses further demonstrated expression of numerous genes located on rat chromosome 9 within chimera mouse. This study demonstrated that a single chromosome from a frozen extinct species can be functionally revived and its transcriptional activity assessed within an interspecies oocyte.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-55500-1.

Introduction

Reviving extinct species is not merely human ambition but a potentially transformative approach with tangible implications for biodiversity conservation and the future of agriculture^1–3. Extinct genomes may harbor numerous unknown genes, and elucidating their functions could provide the blueprints for improving livestock, engineering resilience to environmental change, and even providing critical insights into the causes of extinction. Recent advances have enabled the whole-genome sequencing of specimens discovered in permafrost or preserved as pelts and taxidermy mounts, thereby recovering their DNA sequences^4–7. However, determining the actual functions of genes identified from these sequences is extremely difficult^8,9. When viable cells are available, such as in endangered species or in individuals sampled shortly after death, somatic cell nuclear transfer offers a feasible strategy for genetic rescue^10–12. Remarkably, cells obtained noninvasively, such as somatic cells from urine¹³ or milk¹⁴, can be used for cloning without harming the donor animal. However, even well-preserved frozen carcasses of truly extinct species lack living cells. Nevertheless, healthy cloned mice have been generated from carcasses stored at − 30 °C for up to 16 years¹⁵, and even from freeze-dried somatic cells¹⁶, provided that nuclear DNA remains intact. These findings demonstrate that although cellular viability is lost, genomic integrity can persist, enabling the restoration of individuals, thereby facilitating de-extinction and genetic resource preservation.

However, nuclear transfer faces two major challenges. First, the DNA of excavated extinct animals is often severely damaged, impeding the possibility of cloning^2,17. Second, large numbers of oocytes must be obtained from the same or another closely related species, and many recipient females must be available and capable of carrying the cloned embryos to term. For endangered species such as the gaur and wolf, cloning was achieved using oocytes from related species, such as cattle¹¹ and dogs¹⁸, and surrogate mothers. However, for extinct species, not only are conspecifics unavailable, but their closest living relatives are often endangered, precluding their use as surrogates. Attempts to perform nuclear transfer using oocytes from unrelated species have proven unsuccessful, with embryonic development failing due to multiple incompatibilities, including differences in placental physiology, nuclear–mitochondrial incompatibility, and variations in gestation length between the recipient uterus and the cloned fetus^19–21. Even if suitable related species were available, the inherent inefficiency of current cloning techniques is a hindrance. Consequently, to revive the mammoth via somatic cell nuclear transfer, obtaining a vast number of elephant oocytes and recipient elephant females would pose a significant obstacle. Elephants ovulate infrequently^22,23, meaning that securing sufficient oocytes would require maintaining an impractically large and costly herd.

If the primary objective is to identify and characterize unknown genes from an extinct species, a viable alternative to whole-organism cloning might be the restoration of cells via interspecies cell fusion. Indeed, hybrid cells combining human and mouse genomes have been successfully generated^24,25. However, such hybrid cell lines are notoriously difficult to maintain in culture, as they often lose chromosomes from one species over time. In human–mouse hybrid cells, human chromosomes can disappear during prolonged culture²⁶. Moreover, cells from extinct species preserved in permafrost or taxidermy specimens are invariably dead with disrupted cell membranes, making fusion with living cells impossible. Furthermore, even if stable hybrid cell lines could be established, they would typically originate from differentiated cells (e.g., fibroblasts), which cannot recapitulate the full spectrum of tissue-specific gene expression found in a whole organism.

To address these limitations, we developed a technique for isolating a single chromosome from somatic cells of a frozen rat carcass, which is subsequently injected into a mouse oocyte. The oocyte is fertilized with mouse sperm to form a blastocyst, which is then used to establish mouse ES cell lines carrying one rat chromosome. When these ES cells were used to generate chimeric mice, the rat chromosome was retained and contributed to differentiation across multiple organs. To simulate the analysis of unknown genes from an extinct species, we applied the high-coverage gene expression profiling method²⁷ to chimeric organs and confirmed that rat genes were actively expressed in vivo.

Results

Production of cloned embryos or hybrid ES cell lines from rat somatic cell nuclei and mouse oocytes

Somatic cells were extracted from the blood of a GFP-transgenic (Tg) rat carcass that had been cryopreserved at − 30 °C for more than 1 years. The nuclei were then transferred into enucleated mouse oocytes to attempt to produce cloned rat embryos. Artificial activation of 195 reconstructed mouse oocytes with rat nuclei produced rat-derived pseudopronuclei in over half of the oocytes (58.6%) (Table S1). While most embryos (93.4%) underwent the first cell division to reach the two-cell stage, development ceased thereafter, confirming the difficulty of interspecies nuclear transfer between mouse and rat.

Next, GFP-Tg rat somatic cells were injected into intact mouse oocytes without enucleation and induced parthenogenetic activation in the presence of cytochalasin B (CB). Of 296 reconstructed hybrid oocytes, 218 (73.6%) formed pseudo-pronuclei, while 67 (30.7%) developed to the morula or blastocyst stage. All embryos exhibited either complete or partial GFP expression (Fig. S1a–S1c, Table S1). Attempts to establish ES cell lines from these embryos yielded 9 ES-like cell lines; however, none expressed rat-derived GFP. Two additional cell lines that exhibited cellular morphology distinct from ES cell characteristics and expressed GFP were established, but both disappeared after several passages. Although we cannot exclude the possibility that mouse ES cells retained some rat chromosomes other than those carrying the GFP transgene, no further analyses were performed. We hypothesized that introducing the entire complement of rat chromosomes into mouse oocytes impairs embryonic development. If so, transferring a single rat chromosome, rather than the whole chromosomes, into a mouse oocyte might allow normal development following fertilization with mouse sperm.

Development of a technique for isolating a single chromosome from metaphase II oocytes

To enable the selective isolation of a single rat chromosome, we exploited premature chromosome condensation (PCC) induced by metaphase II (MII) mouse oocytes. Injecting somatic cell nuclei into MII oocytes triggers PCC, irrespective of the donor cell cycle stage, and occurs concomitantly with spindle assembly. Building on this principle, we injected rat somatic cells into intact mouse oocytes to convert the rat nuclei into condensed chromosomes and optimized the extraction of a single rat chromosome from the condensed rat chromosome set.

We first attempted to dissociate chromosomes in MII mouse oocytes. To visualize the chromosomes, we microinjected mRNAs encoding α-tubulin-GFP and histone H2B–mRFP²⁸ or an antibody against histone H3S10ph labeled with phycoerythrin (Fig. 1a). Subsequently, MII oocytes were treated with microtubule polymerization inhibitors nocodazole or demecolcine, either individually or in combination²⁹.

Fig. 1.

Successful dispersion of mouse metaphase II (MII) chromosomes within the same oocytes. (a) Mouse oocytes injected with mRNAs encoding GFP-α-tubulin and histone H2B–mRFP, enabling visualization of the MII spindle and chromosomes by fluorescence microscopy. (b) Two hours after treatment with microtubule polymerization inhibitor nocodazole. A portion of the chromosomes had begun to disperse. (c) Five hours after nocodazole treatment. The oocyte chromosomes were fully dispersed throughout the cytoplasm. (d) Images of the same oocyte at different focal planes.

After 2 h of treatment, some oocytes exhibited partial chromosome dispersion, but none underwent complete dissociation (Fig. 1b). At 5 h post-treatment, 37 oocytes (62.7%) with chromosomes were fully and evenly dispersed throughout the oocyte cytoplasm following nocodazole treatment alone (Fig. 1c and d, Table S2). Some oocytes exhibited partial chromosome dispersion or appeared to have condensed chromatin. Twelve oocytes (20.3%) underwent spontaneous activation and formed pronuclei. Treatment with demecolcine alone resulted in only partial chromosome dispersion, whereas the combined treatment of the two produced fully dispersed chromosomes in 11 oocytes (36.7%), a lower frequency than nocodazole alone.

At 2 h after inhibitor treatment, the oocytes were transferred to a micromanipulation chamber containing cytochalasin B and equipped with a mercury lamp fluorescence unit. While simultaneously performing fluorescence and brightfield observation, we attempted to extract a single chromosome using a micromanipulator fitted with a glass pipette with an inner diameter of 4–5 μm. When a chromosome was aspirated into the pipette, we found that, despite appearing dispersed in the cytoplasm, the chromosome strands remained interconnected. As a result, even when attempts were made to isolate a single chromosome, it was often the case that five to ten chromosomes were simultaneously aspirated together in a sausage-like configuration (Fig. 2a and d).

Fig. 2.

Isolation of a single chromosome from a metaphase II (MII) oocyte and its injection into another oocyte. (a–d) Chromosome isolation attempted 2 h after nocodazole treatment. (a) Attempt to extract chromosomes from an MII oocyte. (b) Fluorescence microscopy image of the same oocyte. Multiple chromosomes were aspirated simultaneously. (c) Chromosomes released from the pipette. (d) Fluorescence microscopy image of the previous image showing several chromosomes connected in a sausage-like manner. (e–h) A single chromosome isolated from an oocyte 5 h after nocodazole treatment and subsequently injected into another intact MII oocyte. (e) Just before injection of the isolated single chromosome into an intact MII oocyte. (f) Fluorescence microscopy image of the same oocyte. (g) The extracted chromosome was successfully introduced into the oocyte. (h) The oocyte was released from the holding pipette, and successful injection was confirmed again from a different angle. Arrows indicate the isolated and injected single chromosomes.

In contrast, when chromosome extraction was performed 5 h after inhibitor treatment, isolating a single chromosome became easier. Injecting the isolated chromosome into another intact oocyte was also easier. Furthermore, due to the small pipette size, oocyte survival was higher than that observed with sperm injection. Fluorescence imaging confirmed the successful introduction of a single chromosome into the recipient oocyte (Fig. 2e and h).

Development of a technique for chromosome extraction from mouse cumulus cells using mouse oocytes and single-chromosome transfer into another oocyte

Next, mouse cumulus cell nuclei were injected into enucleated mouse oocytes, and the reconstructed oocytes were treated with nocodazole, demecolcine, or a combination of both. After 2–3 h of treatment, 26 oocytes exhibited complete chromosome dispersion throughout the oocyte cytoplasm (26.3%) following nocodazole treatment, 26 oocytes (21.7%) with demecolcine, and 27 oocytes (47.7%) with the combined treatment (Fig. 3a and b, Table S3). Extending the treatment to 5 h did not increase the proportion of oocytes with fully dispersed chromosomes, indicating that 2–3 h was sufficient. Confocal imaging confirmed chromosome dispersion (Fig. 3c), while live imaging demonstrated three-dimensional scattering (Fig. 3d, Supplementary Movie 1).

Fig. 3.

Successful dispersion and isolation of individual mouse or rat somatic chromosomes within mouse oocytes. (a) Reconstructed mouse oocytes generated by injecting the nucleus derived from mouse somatic (cumulus) cells, followed by a 2-hour nocodazole treatment. (b) Fluorescence microscopy image of the same oocytes showing the dispersion of somatic chromosomes within the ooplasm. Chromosomes were stained by H3S10ph antibody labeled with phycoerythrin. (c) Stacked image of nocodazole reconstructed mouse oocytes obtained by confocal laser microscopy. (d) Snapshot from a video demonstrating the three-dimensional dispersion of somatic chromosomes within the oocyte (see Supplementary Movie 1). Chromosomes were stained by α-tubulin-GFP and histone H2B–mRFP mRNA injection. (e) Karyoplasts containing a single chromosome extracted from the reconstructed oocyte using a micropipette. (f) Fluorescence microscopy image of the same karyoplasts. (g) Somatic cells recovered from peripheral blood in the tail of a frozen rat cadaver. (h) Fluorescence microscopy image of the same cells. Those nuclei were stained by Hoechst. (i) Several nuclei of those rat cells were aspirated into a micropipette. (j) Mouse oocytes injected with rat somatic cell nuclei. Blue indicate rat nucleus. As shown in Fig. 1g and i, rat nuclei are easy to handle, so Hoechst staining is not performed in actual nuclear transfer experiments. (k) Fluorescence microscopy image of rat chromosomes dispersed within the mouse oocyte. (l) Merged bright-field and fluorescence microscopy images of the same sample. This level of brightness is sufficient for micromanipulator operation.

We then attempted to extract chromosomes from these dispersed reconstructed oocytes using glass pipettes with an inner diameter of 4–5 μm. Even after 2 h of inhibitor treatment, individual chromosomes were easily extracted without interconnection, enabling the retrieval of multiple chromosomes from a single oocyte (Fig. 3e and f). The isolated somatic chromosomes were readily injected into another intact MII oocyte, followed by intracytoplasmic sperm injection (ICSI) using sperm from GFP-Tg mice (129/Sv) (Table S4). Although the developmental rate to the morula or blastocyst stage was slightly reduced (40%), ES cell lines were successfully established from some embryos. These results demonstrate the successful development of a single-chromosome transfer technique, albeit within the same species. The technique facilitates somatic chromosome formation within an oocyte, dispersion into individual chromosomes, and transfer of a single chromosome into another oocyte.

Single-chromosome transfer using somatic cell nuclei from frozen rats

After establishing a single-chromosome transfer technique within the same species, we next conducted experiments using frozen cadavers of a different species. Blood was collected from the tail or foot of GFP-Tg rats that had been stored at − 30 °C for more than 1 year. Somatic cells (presumably leukocytes) were isolated from the blood (Fig. 3g and h) and injected into enucleated mouse oocytes (Fig. 3i and j). At 2–5 h after injection, dispersion of rat chromosomes was observed in 15 oocytes (17.4%) following nocodazole treatment and in 23 oocytes (31.9%) following demecolcine treatment (Fig. 3k and l; Table 1). Notably, the dispersion of rat chromosomes was less uniform than that observed with live mouse cumulus cells. Extending the inhibitor treatment duration did not increase the number of oocytes with dispersed chromosomes. Nevertheless, as in experiments using mouse cumulus cell nuclei, individual rat chromosomes were successfully retrieved using a micropipette.

Table 1.

Development of a method to disassemble chromosomes of frozen rat somatic cell nuclei within enucleated mouse oocytes.

Type of inhibitor		No. of examined oocytes	Separated	Partially separated	Aggregated	Activated	No nucleus	No. of failed observation
Noco.	Deme.
+	−	86	15 (17.4) a	8 (9.3)	48 (55.8)	10 (11.6)	5 (5.8)	0
−	+	72	23 (31.9) b	2*	17*	1*	0*	29*

*Of the 72 reconstructed oocytes examined, 29 that were not classified as “Separated” failed during the process of staining them using a different method for detailed analysis, rendering them unobservable. While the remaining 20 oocytes were examined, no percentage is shown because of the differing sample sizes.

a vs. b: p = 0.0336.

Generation of mouse ES cell lines harboring a rat single chromosome

A single rat chromosome was isolated from reconstructed mouse oocytes and injected into 148 intact mouse oocytes, followed by the injection of sperm from GFP-Tg mice (129/Sv) (Table 2). All 84 surviving oocytes were fertilized, with 48 (57.1%) developing to the morula or blastocyst stage. Because the mouse sperm carried the GFP-Tg transgene, all blastocysts were GFP-positive. In total, 61 embryos, including morulae and blastocysts, and 13 morphologically normal 8-cell stage embryos were transferred to ES cell derivation medium. From these, 17 GFP-positive ES cell lines (27.9%) were successfully established (Fig. 4a and b; Table 2). Karyotype analysis revealed that two lines (SCT-R4 and SCT-R5) contained 41 chromosomes.

Table 2.

Establishment of rat chromosome-containing mouse ES cell lines using frozen rat cadavers.

No. survived oocytes after rat single chromosome injection	No. survived oocytes after ICSI	No. of embryos developed (%)				No. plated embryos*	No. established cell lines**	No. rat chromosome contained ES cell lines***
		Frag/1-cell	2-cell	4-8-cell	Mor/Bla
148	84	6 (7.1)	7 (8.3)	23 (27.4)	48 (57.1)	61	17 (27.9)	2 (11.8)

Frag, Fragment: Moru, Morula: Blast, Blastocymmst.

*. Morula, blastocyst and some relatively good 8-cell embryos were used for plating.

**. Calculated from no. of plated embryos.

***. Calculated from no. of established cell lines.

Fig. 4.

Establishment of ES cell lines from mouse oocytes carrying a single rat chromosome and identification of the rat chromosome number. (a) Bright-field image of the established ES cell line SCT-R4. (b) Fluorescence microscopy image of the same sample. Because the oocyte was fertilized with sperm from a green fluorescent protein (GFP)-transgenic mouse, all cells exhibited GFP fluorescence. (c) 4′,6-Diamidino-2-phenylindole (DAPI)-stained metaphase chromosomes of ES cells. In total, 41 chromosomes were observed. (d) Spectral karyotyping (SKY) analysis using mouse chromosome specific prove. Among the chromosomes observed with DAPI staining, only the chromosome indicated by an arrow was not hybridized by the mouse probe. (e) Mouse chromosome identification based on SKY color assignments. A total of 40 chromosomes were successfully detected using the mouse probe, and no abnormalities such as translocations or deletions were observed. (f) DAPI-stained metaphase chromosomes of ES cells derived from same ES cell line. (g) SKY analysis using rat chromosome specific prove showing only a single hybridized chromosome. (h) Rat chromosome identification based on SKY color assignments. The chromosome labeled by the rat probe corresponded to rat chromosome 9. (i) Gene expression of rat chromosome 9 in rct-mES cell lines. The rat chromosome 9-derived gene was examined by PCR in the established ES cell lines SCT-R1 to SCT-R7. Rat-specific PCR bands (arrows) were detected in lanes 6 and 7, corresponding to SCT-R4 and SCT-R5. M: Marker, 1: Rat tail, 2: Mouse EpiLC, 3: SCT-R1, 4: SCT-R2, 5: SCT-R3, 6: SCT-R4, 7: SCT-R5, 8: SCT-R6, 9: SCT-R7, 10: dH2O.

To identify the origin of the extra chromosome, spectral karyotyping (SKY) was performed using mouse-specific probes. Staining with 4′,6-diamidino-2-phenylindole (DAPI) confirmed the presence of 41 chromosomes. However, SKY analysis using mouse-specific probes identified one chromosome that did not hybridize with mouse probes (Fig. 4c and d; the chromosome indicated by the arrow), while no abnormalities were observed in the remaining 40 chromosomes (Fig. 4e). Next, we performed SKY analysis on the same cell line using a rat-specific probe, and found that only one chromosome hybridized with the rat probe (Fig. 4f and g). SKY analysis revealed that this chromosome was rat chromosome 9 (Fig. 4h). To confirm these results, DNA was extracted from the cell lines, and genotyping was performed at the mStat1 8–9 locus and the rStat1 8–9 locus—genes located on chromosome 9 that differ in size between mice and rats. Because rat chromosomes were introduced into mouse ES cells, genome samples containing either only mouse or rat DNA were prepared as controls. These corresponded to the mouse EpiLC and rat tail samples, respectively. The mouse PRDM14 locus was used as a control to demonstrate that a mouse genomic locus is not amplified from rat genomic DNA by PCR. The results confirmed that the rat chromosome 9 sequence was present only in Lines SCT-R4 and R5 (Fig. 4i). We thus defined these ES cells harboring a rat chromosome as rat chromosome-transferred mouse ES cells (rct-mES cells).

Differentiation potential of rct-mES cells harboring a rat single chromosome

The differentiation potential of rct-mES cells and the expression of rat-derived genes were assessed by injecting rct-mES cells into the perivitelline space of 8-cell stage embryos collected from wild-type ICR mice to produce chimeric mice. For SCT-R4 cell line, 46 pups were obtained from 80 chimeric embryos. Eight exhibited coat color chimerism, while two showed relatively high levels of chimerism (Fig. 5a and b, Table S5). For SCT-R5 cell line, 35 pups were obtained from 60 chimeric embryos, with 6 being chimeric. However, none exhibited high coat color chimerism. The top male and female highly chimeric mice from SCT-R4 (Fig. 5c) were mated for approximately 10 months, but no offspring were produced.

Fig. 5.

Chimeric mice were generated from rct-mES cells carrying a single rat chromosome and verification of the rat chromosome by PCR. (a) A chimeric mouse produced by injecting the SCT-R4 ES cell line into ICR embryos. (b) Fluorescence microscopy image of tissues from the same mouse, showing GFP expression in various body parts, indicating the contribution of ES cells. (c) Chimeric mice at 10 months of age. Because the ES cells were derived from BDF1 × 129 embryos, brown coat color regions represent ES cell-derived tissues. (d) Head musculature of a chimeric mouse generated from the SCT-R4 ES cell line. (e) Fluorescence microscopy image of the same region. (f) Fluorescence microscopy image of the chimeric mouse large intestine. (g) Fluorescence microscopy image of a chimeric mouse heart. (h) The GFP-positive region of the same heart was dissected. These tissues were subsequently used for gene expression analyses. (i) Genotyping analysis of mouse and rat alleles. PCR genotyping was performed using genes located on rat chromosome 9 that differ in product size between mouse and rat. Chimeric mice generated from the SCT-R4 and SCT-R5 cell lines exhibited both mouse- and rat-specific bands. M1: mStat1 8–9 locus; R1: rStat1 exon 8–9; M2: mStat1 9–10; R2: rStat1 exon 9–10; R3rSmarcal1 exon 21–22; M3: mSmarcal1 exon 21–22; R4: rSmarcal1 exon 11–12; M4: mSmarcal1 11–12; R5: rEif4e2; M5: mEif4e2.

Next, the two highly chimeric mice from SCT-R4 and one low-chimeric mouse from SCT-R5 were dissected to examine tissue chimerism. Because these cell lines expressed GFP, the chimerism of all organs could be assessed by GFP fluorescence. GFP was expressed in the brain, heart, and muscle of all chimeras. In some chimeras, GFP was also detected in the lung, intestine, thymus, testis, and bone (Fig. 5d and h, Table S6). These findings confirm that rct-mES cells could differentiate into derivatives of all three germ layers. GFP-positive regions were isolated from organs (Fig. 5g and h) with high chimerism for genotyping, which confirmed the presence of genes derived from rat chromosome 9 (Fig. 5i).

Expression of rat genes in chimeric mouse tissues derived from rct-mES cells carrying rat chromosomes

Furthermore, we examined rat gene expression in GFP-positive tissues isolated from the chimeric mouse. The transcripts obtained from GFP-positive tissues is potentially an admixture of mouse and rat transcripts. Given the high degree of sequence conservation between mouse and rat genes, hybridization-based microarray analyses or even RNA-seq could lead to interspecies cross-hybridization or cross-mapping, potentially compromising the accuracy of expression profiling. Moreover, in this case, the possible expression of rat-derived transcripts at extremely low levels must be considered. To distinguish between mouse and rat transcripts under these conditions, we employed the high coverage gene expression profiling (HiCEP) method, which enables high-sensitivity, high-reproducibility quantification of gene expression, including transcripts from unknown genes, thereby overcoming both of the aforementioned issues²⁷. Three types of samples were used for the analysis: GFP-positive and GFP-negative regions dissected from the hearts of the chimeric mouse (Fig. 5g and h) and rat hearts as a reference.

Consequently, we successfully detected 23,469 peaks in the GFP-positive heart regions, of which 64 were exclusively present in the GFP-positive regions. These peaks were also expressed in the rat heart, which was used as the reference (Fig. 6). The HiCEP method was optimal for our purpose of distinguishing species-specific signals. However, its primary limitation is that the gene corresponding to each peak cannot be readily identified. Thus, further analysis is required to identify the corresponding gene for each detected peak. We performed gene prediction and found that one of the detected peaks was likely derived from rat Hsp90ab1 (Supplementary Data 1). Because this gene is located on chromosome 9, this assignment is fully consistent with those of the SKY analysis, which demonstrated the presence of rat chromosome 9 in our chimeric mouse tissue derived from rct-mES cells. The expression of Hsp90ab1 has been previously reported in rat hearts (https://www.ncbi.nlm.nih.gov/gene/301252)³⁰. To validate our prediction, we conducted a competitive inhibition assay using two independent Hsp90ab1-specific primers (primers 1 and 2; Fig. S2a). The assay revealed the disappearance of the peak upon competitive inhibition; thus, the detected peak originated from Hsp90ab1²⁷ (Fig. S2b).

Fig. 6.

Expression of rat transcripts in chimeric mouse tissues derived from rct-mES cells carrying rat chromosomes. Peak profiles of the HiCEP analysis. The x- and y-axes indicate fragment lengths and intensities of fluorescence, respectively. The two lines plot the results of two HiCEP reactions performed independently using the same RNA fraction, indicating high reproducibility. The results of three samples: green fluorescent protein (GFP)-negative region of the chimeric mouse heart (top), GFP-positive region of the chimeric mouse heart (middle), and rat heart (bottom). The length of each peak and the 2 bases + 2 bases used for the selective polymerase chain reaction (PCR) in our HiCEP analysis are shown. (a) A representative pattern of HiCEP peaks (PCR ID: AG-AG). (b) Five representative images of the 64 peaks exclusively detected in the GFP-positive, but not GFP-negative, regions. Red arrows indicate transcripts exclusively expressed in the GFP-positive region, of which expression was also observed in the reference, rat heart, as indicated by black arrows.

Chromosome dispersion of somatic cell nuclei from frozen porcine and bovine tissues

Finally, we examined whether our method could induce chromosomal dispersion within mouse oocytes using somatic cell nuclei from species other than rats. Nuclei were isolated from frozen porcine and bovine muscles, injected into enucleated mouse oocytes, and cultured in the presence of nocodazole. Although chromosome dispersion was less uniform than that observed with mouse or rat nuclei, several chromosomes were scattered within the reconstructed oocytes (Fig. S3, Table S7). These results indicate that the developed approach for retrieving chromosomes from frozen cadavers is applicable to mammalian species beyond rats.

Discussion

In this study, we successfully isolated a single chromosome from the carcass of a rat cryopreserved for more than 1 year and introduced it into mouse ES cells. The resulting ES cells contributed to multiple tissues within chimeric mice, and the rat-derived genes carried on the introduced chromosome were expressed across these tissues. Our findings demonstrate that even chromosomes retrieved from long-term frozen specimens can retain sufficient functional integrity to be maintained in pluripotent cells and support transcriptional activity in vivo. With further development, this approach may provide a powerful platform for recovering chromosomes from extinct species and assessing the functions of their genes within a living mammalian context.

Injection of mouse fresh cumulus cells into mouse oocytes and subsequent treatment with a tubulin polymerization inhibitor³¹ enabled efficient and rapid chromosome dispersion, more so than when MII oocyte chromosomes were treated directly. In MII oocytes, we occasionally observed multiple chromosomes remaining interconnected and being removed together, whereas this phenomenon did not occur when using cumulus cell chromosomes. One explanation is that MII oocytes naturally reside in metaphase due to their intrinsic cell cycle arrest, during which the meiotic spindle is already fully assembled. Under such conditions, complete spindle disruption by a tubulin polymerization inhibitor may be difficult to achieve. In contrast, when cumulus cells, mainly at G0 or G1^32,33, were injected into oocytes, their nuclei bypassed the S phase and directly entered mitosis, resulting in premature chromosome condensation and the initiation of spindle assembly^34,35. However, because we applied the polymerization inhibitor immediately after nuclear transfer, the cumulus cell nuclei were unable to form a mature spindle within the oocyte. Consequently, chromosomes derived from cumulus cells were more readily separated into single units than those of MII oocytes, whose spindle structure was already established. These observations suggest that when collecting donor cells from frozen carcasses, it is preferable to avoid harvesting cells in the M phase of the cell cycle, as these may be less amenable to complete spindle disassembly and efficient chromosome isolation.

Similarly, if the donor cell is in S phase, the donor nucleus transplanted into the oocyte may undergo PCC while DNA synthesis is still ongoing, which is likely to cause severe chromosomal abnormalities³⁶. These considerations indicate that donor nuclei in the M or S phase should be avoided in the present approach. However, in practice, there is no reliable method to accurately determine the cell cycle stage of cells recovered from cadavers. Furthermore, it is technically challenging to selectively identify and recover specific tissues or cell types—such as cumulus cells—from frozen cadavers. Given these constraints, nuclear size may serve as a practical surrogate marker for estimating the cell cycle stage during donor nucleus selection. Larger nuclei could represent G2-phase nuclei in which DNA replication has been completed; nevertheless, they are also expected to include a substantial proportion of nuclei in the M or S phase. In contrast, smaller nuclei are more likely to correspond to nuclei in the G0/G1 phase, prior to the onset of DNA synthesis. Therefore, within the limitations of this method, selecting the smallest nuclei among those recovered may represent the most reasonable strategy for minimizing the risk of chromosomal abnormalities.

This study aimed to model the partial revival of extinct animals discovered in permafrost using freezing rats at − 30 °C, a temperature comparable to natural permafrost. Somatic cells recovered from these frozen rat carcasses could induce PCC when injected into mouse oocytes, similar to freshly isolated mouse cumulus cells. We previously cloned mice from somatic cells preserved at − 20 °C for 16 years, demonstrating that nuclei can remain largely intact when bodies are stored under such conditions for extended periods¹⁵. In contrast, rat somatic cells preserved for only 1 to 1.5 years exhibited relatively lower chromosome-dispersal efficiency than mouse cumulus cells. If this reduced efficiency reflects damage accumulated during the 1 to 1.5 year freezing period, then the somatic cell nuclei of extinct animals that have remained frozen in permafrost for thousands to tens of thousands of years have suffered more severe damage¹⁷. However, Yamagata et al.² successfully assembled recognizable chromosomal structures using nuclei of somatic cells obtained from a 28,000-year-old wooly mammoth recovered from permafrost and injected into mouse oocytes. Therefore, although most mammoth somatic cells likely suffered extensive nuclear damage, a limited subset of cells may have retained relatively intact nuclei. Because the present method relies on mouse oocytes, it does not require collecting oocytes from species closely related to the donor animal, thereby enabling large-scale nuclear transfer assays. Moreover, injection of somatic cell nuclei into mouse oocytes is a well-established, technically accessible procedure^15,34, allowing large numbers of cells to be screened. Even if the proportion of minimally damaged cells is extremely low, high-throughput approaches should enable the identification of such rare cells and the induction of PCC from extinct animals within mouse oocytes. In addition, because oocytes possess a robust capacity for DNA repair³⁷, it is conceivable that DNA damage accumulated in donor nuclei during the period in which animals were cryopreserved and left unattended could be repaired by the oocyte after nuclear transfer. Indeed, Yamagata et al. demonstrated, using live-cell imaging with an antibody against the DNA damage marker γH2A.X, that DNA damage in a subset of transplanted mammoth nuclei may be repaired within mouse oocytes². Furthermore, it has been shown that DNA damage observed in nuclear transfer embryos can be repaired through a BRCA1-dependent mechanism³⁸. However, recent studies have shown that the mouse nuclear transfer techniques currently in use result in an approximately threefold higher mutation rate in cloned mice compared with those produced by natural mating, and that repeated rounds of serial nuclear transfer lead to chromosomal translocations and deletions³⁹. Therefore, further development of the present study will require parallel efforts to improve mouse nuclear transfer technologies.

If clones could be generated from extinct animals, recovering their complete genomes would be possible, enabling the investigation of the functions of genes that have been lost from existing species^1,3. However, given current technological limitations^39,40, the extent of DNA damage in donor cells, and the limited availability of oocytes and recipient females, the full resurrection of extinct animals by cloning remains highly challenging². In contrast, our approach allows somatic cell nuclei, even those bearing extensive cryopreservation-induced damage, to reconstitute relatively normal chromosome structures within mouse oocytes. These chromosomes retrieved from extinct species could subsequently be revived within mouse ES cells. Using chimera mouse techniques, such ES cells could then be differentiated into many tissues, enabling analysis of the expression of previously uncharacterized genes. Although the rat used as a model organism in this study, whose genome has been fully annotated, microarray analysis cannot be applied to extinct species or other organisms for which the complete set of genes or genome sequence is unavailable. Similarly, RNA-seq analysis can become highly complex and challenging in the absence of a reference genome. Furthermore, in comparative studies between closely related species, such as mouse and rat in the present study, identifying gene expression derived specifically to the resurrected chromosome becomes even more difficult. In contrast, the HiCEP method does not require any prior genomic information, facilitating highly accurate and reproducible detection of differences in gene expression profiles between the chimeric recipient animal and donor-derived genetic material, even when the genome and genes of the donor species are entirely unknown.

Extensive progress has been made in generating animals carrying chromosomes from different species. For example, mice harboring artificial chromosomes that contain segments of human chromosomes have been established⁴¹, and rats carrying an entire human chromosome have also been generated as transchromosomic models⁴². However, all these studies relied on viable donor cells and introduced foreign chromosomes using cell-fusion-based methods. Such approaches cannot be applied to extinct animals, as their cells are no longer viable. In contrast, our approach employs nuclear transfer, in which nuclei are isolated and directly injected into oocytes^15,34; therefore, the method is not limited by the viability of donor cells and can be performed as long as nuclear material can be recovered. In addition, our preliminary experiments demonstrated that nuclei recovered from porcine and bovine somatic cells derived from commercially available frozen meat were capable of inducing PCC in mouse oocytes using this technique. While optimization of PCC-inducing conditions, single-chromosome isolation and injection, and subsequent ES cell derivation have yet to be achieved, these findings suggest the potential applicability of this method to animal species other than rats.

Unfortunately, in this study, we could only introduce rat chromosome 9 into mouse ES cells. This outcome suggests two possibilities: chromosomes other than rat chromosome 9 may negatively affect early mouse embryonic development when introduced into mouse oocytes, or alternatively, they may be unable to persist stably within mouse ES cells. In fact, it has been reported that hybrid cell lines produced by cell fusion between two different species frequently exhibit preferential loss of chromosomes derived from one species over time^24,25. If either is the case, the practical utility of the current method would remain limited, and their reproducibility is very low. Nevertheless, the present work represents the first demonstration of isolating a single chromosome from frozen carcasses of a heterologous species and incorporating it into mouse ES cells. If further improvements to this technique are successfully achieved, it may enable not only the functional characterization of previously unknown genes derived from extinct animals, but also the potential rescue of embryos or stem cells carrying chromosomal abnormalities, such as aneuploidy.

Future studies will need to address several technical challenges, including optimization of the timing of chromosome injection and fertilization, and the mitigation of phototoxic effects induced by ultraviolet light during micromanipulation⁴³. Moreover, considerable technical development will be required to determine which chromosomes can be introduced without compromising embryonic development, to suppress developmental delays after chromosome introduction, and to prevent loss of the introduced chromosomes from derived ES cells.

Materials and methods

Animals

Eight- to 12-week-old B6D2F1 (C57BL/6 N × DBA/2) female (23–28 g of body weight) and 4-week-old BN/ScN × SD-Tg(CAG-EGFP)-Tg female and male rats (65–76 g of body weight) were obtained from the Shizuoka Laboratory Animal Center (Hamamatsu, Japan). The eGFP-Tg ICR mouse strain (8–10 weeks of age and 31–34 g of body weight) was generated and maintained in our laboratory. On the day of the experiment or after having finished all experiments, mice were euthanized by cervical dislocation. All experiments were conducted according to the Guide for the Care and Use of Laboratory Animals and approved by the Institutional Committee of Laboratory Animal Experimentation of Yamanashi University (reference no. A4-10), or of the RIKEN Center for Developmental Biology (reference no. AH14-13-19). All experiments were performed in accordance with these regulations and guidelines, in accordance with the ARRIVE guidelines. All mice were kept under specific pathogen-free conditions, under a controlled temperature (25 °C), relative humidity (50%), and photoperiod (14 L-10D). Mice were fed a commercial diet and provided with distilled water ad libitum.

Media

HEPES-CZB medium⁴⁴ and CZB⁴⁵ medium were used for oocyte/embryo manipulation and incubation in 5% CO₂ at 37 °C, respectively. The NIM medium⁴⁶ was used for the collection of rat, porcine and bovine nuclei.

Collection of oocytes and cumulus cells

Female mice were superovulated by injecting 5 IU of equine chorionic gonadotropin and 5 IU of human chorionic gonadotropin 48 h apart. Cumulus-oocyte complexes (COCs) were collected from the oviducts of females 14–16 h later and moved to a Falcon dish containing HEPES-CZB medium. To disperse the cumulus, COCs were transferred into a 50 µl droplet of HEPES-CZB medium containing 0.1% bovine testicular hyaluronidase for 3 min. Cumulus-free oocytes were washed twice and placed in a 20 µL droplet of CZB for culture. At the same time, the remaining cumulus cells were introduced into PVP medium⁴⁴ on the manipulation chamber.

Collection of rat, porcine and bovine somatic cells

Whole rat bodies (euthanatized by vertebrae dislocation and put into plastic bags) were kept at -30 °C for up to 18 months without defrost. The blood cells were collected by squeezing 1–2 µL of blood from the tip of the tail or foot after thawing; these were simply washed and diluted in NIM. Cells were kept at 4 °C until use. Porcine and bovine skeletal muscle were purchased from a grocery store and preserved at -30 °C until use. On the day of the experiment, tissues fragments were thawed by adding 400 µL of NIM and then homogenized gently to collect denuded nuclei, as described previously¹⁵.

mRNA synthesis

α-Tubulin-EGFP and H2B-mRFP1 cloned into the pcDNA3.1-poly(A) vector were used as a template for in vitro transcription²⁸. The linearized plasmid was dissolved in nuclease-free water as template DNA for subsequent in vitro transcription using the RiboMax Large-scale RNA Production System T7 kit (p1320; Promega). To improve translational efficiency, Ribo m⁷G Cap Analog (P1711; Promega) was added to the in vitro transcription mixture. The synthesized mRNA was treated with DNase I to eliminate the template DNA, purified with phenol/chloroform, and precipitated with ethanol. For the removal of unreacted nucleotides, the mRNA dissolved in RNase-free water was passed through the MicroSpin G25 column. The purified mRNA was stored at − 80 °C until use.

Microinjection of mRNA or antibody into oocytes

α-Tubulin-EGFP and H2B-mRFP1 mRNA were diluted with nuclease-free water to 5–10 ng/µL before use. Both mRNAs were injected into the cytoplasm of the oocytes, as described previously^47,48. Alternatively, an antibody against histone H3S10ph labeled with phycoerythrin⁴⁹ was injected as above. Briefly, microinjection was performed in HEPES-CZB on an inverted microscope with a micromanipulator (Narishige, Tokyo, Japan). The zona pellucida and cytosolic membrane were penetrated using a piezo drive (PRIME Tech, Tokyo, Japan). Ten minutes after microinjection, oocytes were washed and cultured in CZB. The mRNA-injected oocytes were incubated at 37 °C under 5% CO₂ in air for at least 2 h after injection to allow time for protein production.

Enucleation of oocytes

Groups of oocytes were transferred into a droplet of H-CZB containing 5 mg/mL CB on the microscope stage for enucleation of the MII spindle. Oocytes undergoing microsurgery were held with a holding pipette, and a hole was made in the zona pellucida following the application of several piezo-pulses to an enucleation pipette (inner diameter [ID]: 7–9 μm). The MII chromosome–spindle complex was aspirated into the pipette with a minimal volume of ooplasm. After enucleation of all oocytes in one group, they were transferred into CZB medium.

Injection of rat and mouse somatic cell nuclei into mouse enucleated or intact oocytes

For nuclear injection, rat blood cells or mouse cumulus cells were gently aspirated into and out of the injection pipette until their nuclei were largely devoid of visible cytoplasmic membrane. Each nucleus was immediately injected into an enucleated or intact mouse oocyte. The reconstructed oocytes were kept in the incubator until inhibitor or activation treatment.

Activation of reconstructed oocytes and embryo transfer

For rat interspecies cloning experiments, reconstructed oocytes were activated using 5 mM SrCl₂ in Ca²⁺-free CZB medium in the presence of 50 nM TSA supplemented with 5 µM LatA for 10 h⁵⁰ and then cultured in CZB until embryo transfer.

Chromosome formation of donor somatic nucleus within mouse oocytes and extraction and transfer of single chromosome into another mouse oocyte

Reconstructed oocytes were transferred into a CZB medium containing with or without 0.05 µg/mL demecolcine, 3 µg/ml nocodazole, or both and cultured for up to 5 h until use. The reconstructed oocytes were observed using a fluorescence or a confocal microscope. To extract a single chromosome, the reconstructed oocytes were transferred into a droplet of H-CZB containing 5 µg/ml CB on the stage of the micromanipulator. The oocytes and chromosomes were observed under brightfield and fluorescence microscopy using the microscope’s halogen and UV lamps simultaneously, and a single chromosome was extracted using a glass pipette (ID: 4–5 μm). This method is similar to the technique we previously reported for extracting micronuclei from two-cell-stage embryos²⁹, but a finer glass needle was used. The extracted single chromosome was injected into another oocyte using the same method as sperm injection (described below).

Intracytoplasmic sperm injection (ICSI)

ICSI was performed as described previously^44,51. For microinjection of spermatozoa, 1 to 2 µL spermatozoa suspension was moved directly to the injection chamber. Several piezo pulses separated the spermatozoa head from the tail, and the head was injected into the oocyte. The oocytes that survived ICSI were incubated in a CZB medium at 37 °C with 5% CO₂. Pronucleus formation was checked at 6 h after ICSI.

In vitro development and establishment of ES cell lines

After ICSI, survived embryos were cultured individually for 3 days and examined for blastocyst development. When embryos developed into morulae or blastocysts, they were treated with acid Tyrode solution to remove the zonae pellucidae and used to establish ES cell lines, as described previously^52,53, with slight modifications. Some 8-cell stage embryos were also used in this experiment. Embryos were placed in 96-multiwell dishes precoated with mouse embryonic fibroblasts in 20% Knock-out Serum Replacement (Invitrogen, Carlsbad, CA, USA) and 0.1 mg/mL adrenocorticotropic hormone (fragments 1–24; American Peptide Company, Sunnyvale, CA, USA). Proliferating outgrowths were dissociated using trypsin and replated to fibroblasts until stable cell lines emerged.

Karyotype analysis of ES cell lines and identification of transplanted rat chromosome numbers

To increase the metaphase stage of ES cells, 10 µL/mL demecolcine (045-18761; Wako, Japan) was added to the medium and the cells were cultured for 2 h. Cells were detached by trypsin, exposed to 0.075 M KCl solution for 20 min, fixed with Carnoy’s solution, and applied onto the glass slides. To count chromosomes, the glass slides were stained with Giemsa or DAPI and examined under a microscope. For SCT-R4 and SCT-R5 cell lines, chromosomes were stained using spectral karyotyping with fluorescent in situ hybridization (SKY-FISH) chromosome painting techniques (Spectral Imaging Ltd., Vista, CA, USA) using mouse and rat probes according to the manufacturer’s protocols.

Genotyping

The spaces-specific PCR markers (mStat1 8–9 locus, rStat1 exon 8–9, mStat1 9–10, rStat1 exon 9–10, rSmarcal1 exon 21–22, mSmarcal1 exon 21–22, rSmarcal1 exon 11–12, mSmarcal1 11–12, rEif4e2, mEif4e2) were amplified using primer pair sequences obtained from Ensembl (https://asia.ensembl.org/index.html) and Rat genome information. DNA was extracted from rct-mES cells, tissue from chimera mice, mouse epiblast-like cells, and mouse and rat tail tips. PCR amplification was performed, and the products were separated on 2% polyacrylamide gel before visualization.

Production of chimera mice

Embryos were obtained from ICR strain females mated with ICR males. To generate chimeras, rct-mES cells were injected into the perivitelline space of 8-cell-stage embryos. The next day, the chimeric blastocysts were transferred into a day-2.5 pseudopregnant ICR female mice that had mated with a vasectomized male the 3 days before the transfer. On the day of embryo transfer, recipients were anaesthetised by intraperitoneal injection of the anaesthetic agents’ medetomidine, midazolam and butorphanol. After the embryo transfer was completed, atipamezole was administered, and the mice were kept warm until they regained consciousness. Overall, 6–10 embryos were transferred into each uteri. Pups were delivered at 19.5 days of gestation, and chimerism was assessed by GFP fluorescence. After sexual maturation, males and females with high levels of chimerism were crossed. At 10 months of age, various organs were collected from three chimeric mice, and the degree of chimerism in each organ was evaluated. For some organs, only the GFP-positive regions were dissected and used for subsequent analyses.

Gene expression analysis of the heart prepared from chimera mice using high coverage gene expression profiling technology

HiCEP is a genome-wide gene expression profiling procedure based on amplified fragment length polymorphism (AFLP)²⁷. In this procedure, each transcript is identified by combining restriction enzyme recognition sequences (here, MspI and MseI) with the two bases adjacent to both ends of the digested fragments. First, GFP-positive and GFP-negative regions were dissected from our chimeric mouse hearts, and, as a reference, hearts from F1 rats (BN/SsN × SD-Tg (CAG-EGFP)) (Japan SLC) were used. Total RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany), and 1 µg of total RNA was converted to cDNA using the SuperScriptIII First Strand Synthesis system (ThermoFisher Scientific, Waltham, MA, USA) with 5′ biotinylated oligo (dT) primers, followed by double strand cDNA synthesis. Subsequently, the cDNAs were digested with the restriction enzyme MspI, and the synthetic MspI adapter was ligated and then trapped by streptavidin-coated magnetic beads. Next, the cDNA fragments on the magnetic beads were digested with the restriction enzyme MseI, and their supernatant, including the digested fragments, was collected and then ligated with synthetic MseI adapters. These products were used as templates for selective PCR using fluorescent primers. The amplified products were denatured and loaded on an ABI PRISM 3100 electrophoresis system (ThermoFisher Scienctific).

Peak prediction and gene identification

We performed peak prediction in silico using 115,364 cDNA sequences deposited in the Rat RefSeq database (https://ftp.ncbi.nlm.nih.gov/refseq/R_norvegicus/mRNA_Prot/) and determined the length of the MspI-MseI fragments along with the sequences adjacent to each adapter, which were utilized for selective PCR. As a result, our system predicted that the 332-base (CG-AT) HiCEP peak was generated from NM_001004082, Rattus norvegicus heat shock protein 90 alpha family class B member 1 (Hsp90ab1). Therefore, we conducted a competitive inhibition assay to verify our prediction that the intensity of the fluorescently labeled PCR peak decreases with the addition of non-labeled gene-specific primers, two independent primers.

Statistical analysis

The disassembled spindle was evaluated using the χ² test. P < 0.01 was considered to statistically significant.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

S.W. and T.W. conceived and designed the study. S.W., A.R., S.M., K.Y., M.T., G.N., M.A. and T.W. performed experiments, analysed the data, and interpreted the results. T.W. wrote the manuscript. All authors read and edited the manuscript.

Funding

This work was partially funded by the Research Fellowships of Japan Society for the Promotion of Science to S.W. (23K08843), to K.Y. (24H02325), to T.W. (23K18124 and 24K01779); the Naito Foundation and Takahashi Industrial and Economic Research Foundation (189) to S.W.; Asada Science Foundation to T.W. and K.Y.; the Canon Foundation (M20-0008) to T.W.; JST Development of Advanced Measurement and Analysis Systems (JPMJSN09D1 and JPMJSN12E1) to M.A.

Data availability

All data supporting the findings of this study are available within the paper and its Supplementary Information.

Declarations

Competing interests

The authors declare no competing interests.