Mice cloned from skin cells

Abstract

Adult stem cells represent unique populations of undifferentiated cells with self-renewal capacity. In many tissues, stem cells divide less often than their progeny. It has been widely speculated, but largely untested, that their undifferentiated and quiescent state may make stem cells more efficient as donors for cloning by nuclear transfer (NT). Here, we report the use of nuclei from hair follicle stem cells and other skin keratinocytes as NT donors. When keratinocyte stem cells (KSCs) were used as NT donors, 19 liveborn mice were obtained, 9 of which survived to adulthood. Embryonic keratinocytes and cumulus cells also gave rise to cloned mice. Although cloning efficiencies were similar (<6% per transferred blastocyst), success rates were consistently higher for males than for females. Adult keratinocyte stem cells were better NT donors than so-called transit amplifying (TA) keratinocytes in both sexes (1.6% vs. 0% in females and 5.4% vs. 2.8% in males). Our findings reveal skin as a source of readily accessible stem cells, the nuclei of which can be reprogrammed to the pluripotent state by exposure to the cytoplasm of unfertilized oocytes.

Cloning of animals by nuclear transfer (NT) was attempted as early as 1928 (1), and the first mammal cloned from an adult cell was reported in 1997 (2). A convenient experimental model system for cloning is the mouse, for which a cloning protocol was described in 1998 (3). In all species that have been cloned to date, cloning efficiencies remain disappointingly low: at best, a few percent of manipulated oocytes (4). In mice, NT has been successful with cells at various stages of differentiation, such as fetal germ cells (5), immature Sertoli cells (6), fibroblasts (7), T and B lymphocytes (8), natural killer T cells (9) and olfactory sensory neurons (10, 11).

NT entails the transfer of a nucleus from a somatic cell into an enucleated, unfertilized oocyte. This reconstructed, diploid oocyte is then cultured in vitro to produce a preimplantation embryo, such as a morula or a blastocyst. Depending on the study, NT-derived blastocysts have either been cultured further to generate ES cells (ntES cells) (12) or used directly for uterine transfer to generate cloned mice (3). The success rate for establishment of ntES cell lines or live birth of cloned animals may be influenced by the proliferation or differentiation state as well as the sex of the donor nucleus. Such factors could contribute to the variable efficiency with which the donor nucleus can be “reprogrammed” to undergo the orchestrated sequence of chromatin changes that establish a permissive state for embryonic development.

Cumulus cells (CCs) surrounding the oocyte have often been used in NT research, primarily because they are easy to collect. A limitation is that CCs are only present in females. Some lines of ES cells, which are typically male, appear at first glance to be more efficient as nucleus donors, at least when the number of liveborn cloned mice is compared with the number of transferred blastocysts (13). Their performance is less unusual when the success rate is calculated as the number of liveborn cloned mice compared with the number of reconstructed oocytes. Nevertheless, the apparently higher cloning efficiency of certain lines of (male) ES cells compared with (female) CCs has raised the speculation that the nucleus of an (undifferentiated) ES cell may be reprogrammed more efficiently than the nucleus of a (differentiated) somatic cell. If so, stem cells from adult mice may be expected to be better NT donors than differentiated somatic cells such as CCs because their differentiation status has often been considered as intermediate between ES cells and differentiated somatic cells.

Skin epithelium is an accessible tissue that contains multipotent keratinocyte stem cells (KSCs) localized in a region of the hair follicle known as the bulge (14–16). KSCs within the bulge region are mobilized in the regeneration of skin epithelial tissues, including the normal cyclic regrowth of the hair follicle, and in epidermal repair after wounding (17–20). Highly enriched populations of bulge cells can be isolated from mice by FACS on the basis of cell-surface markers and/or expressed transgenes (18, 21–23). When passaged in culture, progeny from individual bulge cells generate epidermis, hair follicles, and sebaceous glands, thereby revealing the multipotency of these stem cells (20, 22).

In vitro, keratinocytes isolated from the bulge region of adult mice display a higher tendency to generate large colonies than those cultured from the epidermis or follicle outer root sheath (20, 22, 24). Typical of stem cells, the large colonies can be passaged for longer times and give rise to additional large colonies. By contrast, the small colonies undergo senescence after a few passages and hence display features more commonly associated with transit amplifying (TA) cells (22, 24, 25). In humans, epidermal keratinocytes that generate large colonies in vitro have been used successfully in regenerative medicine to repair burned skin (26).

In the present study, we tested the performance of mouse bulge cells from adult hair follicles in cloning by NT. We compared them with CCs, the donor cell type that is most frequently used in mouse NT research. We also tested TA keratinocytes from both sexes. We asked whether embryonic keratinocytes are more efficient than adult keratinocytes for cloning by NT. Our comparative study reveals the highest cloning efficiency for bulge cells of follicles from adult males.

Results

Engineering Mice to Genetically Tag Bulge Cells.

To test the efficiency of follicle bulge cells as NT donors, we first engineered mice carrying a β-galactosidase gene that is expressed specifically in the epithelium of the skin. For this purpose, ROSA-floxed-lacZ mice (27) were crossed with mice carrying the Cre recombinase under the control of the keratinocyte-specific keratin 14 (K14) promoter (28) (Fig. 1A). Backskin sections of control ROSA-floxed-lacZ mice showed no expression of β-galactosidase by histochemical staining with X-Gal. In contrast, backskin sections from K14-Cre/ROSA-floxed-lacZ mice expressed β-galactosidase exclusively in the keratinocytes of the epidermis, hair follicle, and sebaceous glands (Fig. 1A). These test mice can thus be used to monitor present or past K14 expression by β-galactosidase histochemistry with X-Gal.

Fig. 1.

Isolation of cells used as donors for NT. (A) A cross of two strains of genetically modified mice was used to specifically label the keratinocytes from stratified squamous epithelial origin. The first strain (27) carries a targeted mutation in the ROSA26 locus and contains the lacZ reporter gene, the expression of which is blocked by a DNA segment that is floxed by loxP sites. The second strain (28) carries a transgene expressing the site-specific Cre recombinase under the control of the human K14 promoter, which is specific for the basal layer of the epidermis and outer root sheath of the hair follicle. This region includes the bulge, which is the putative stem-cell compartment of the follicle. (Lower Right) The skin of a mouse with only the ROSA reporter does not express β-galactosidase enzymatic activity. (Lower Left) When a mouse also contains the K14-Cre transgene, the floxed DNA segment is excised specifically in K14-expressing cells, and β-galactosidase enzymatic activity can be detected by the generation of a blue precipitate after X-Gal staining. (B Left) Cells were sorted by FACS using antibodies against surface α6 integrin-phycoerythrin (PE) and CD34-allophycocyanin (APC). (B Right) The purity of these cells was analyzed by X-Gal histochemistry of cells after cytospin onto glass slides.

Skins from adult K14-Cre/ROSA-floxed-lacZ mice were used for isolation of keratinocytes by FACS. To enrich for keratinocytes with proliferative capacity, we used conjugated antibodies against the extracellular domain of α6 integrin (Fig. 1B). The α6β4 integrin forms the transmembrane core of epithelial-specific hemidesmosomes, which adhere to the basement membrane underlying all basal keratinocytes. Another surface marker, CD34, was used to distinguish bulge from non-bulge keratinocytes within the α6⁺ population (22). Bulge keratinocytes were positive for both α6 integrin and CD34 (α6⁺CD34⁺), and cells within this population have been demonstrated to be multipotent KSCs. By contrast, keratinocytes that are integrin α6-positive but CD34-negative (α6⁺CD34⁻) include the basal layer of the epidermis and the upper outer root sheath of the hair follicle. These populations are enriched for keratinocytes that have traditionally been referred to as TA cells. X-Gal staining of cytospin preparations revealed that both of these sorted populations expressed β-galactosidase enzymatic activity (Fig. 1B).

NT with CCs and with Bulge and Non-Bulge Keratinocytes of Female Mice.

To compare the NT efficiency of bulge and TA keratinocytes to CCs, we used superovulated female donor mice (K14-Cre/ROSA-floxed-lacZ) as a source of all three types of nuclei. CCs were isolated by light enzymatic treatment of oocyte–CC complexes (3). α6⁺CD34⁺ and α6⁺CD34⁻ keratinocytes were used directly after FACS. Cells were examined microscopically, and those chosen for NT were intact, with a small cytoplasm:nuclear ratio.

Superovulated wild-type mice provided the unfertilized oocytes. After enucleation of oocytes, donor nuclei from CC and keratinocyte populations were transferred by piezo-actuated microinjection (3, 11). The reconstructed, diploid oocytes were then cultured in vitro to the blastocyst stage (Fig. 2A). Reconstructed oocytes derived from CC nuclei were cultured in somatic cell-culture medium MEMα because cloned embryos from CCs thrive better in a more somatic cell-like environment (29). Reconstructed embryos from α6⁺CD34⁺ and α6⁺CD34⁻ keratinocyte nuclei were cultured in the standard embryo culture medium known as K⁺-modified simplex optimized medium. To minimize variations in experimental conditions, we performed NT with CCs on one day and with α6⁺CD34⁺ and α6⁺CD34⁻ keratinocytes from the skins of the same donor mice the following day.

The rate of in vitro development of reconstructed oocytes into blastocysts was higher for CCs (61.3%) than for α6⁺CD34⁺ keratinocytes (52.0%; P < 0.05) (Table 1). However, in vivo development to term was similar for CCs and bulge keratinocytes (1.2% vs. 1.6% per transferred blastocyst; 0.7% vs. 0.8% per reconstructed oocyte). Fig. 2B shows three pups cloned from nuclei of α6⁺CD34⁺ keratinocytes and CCs isolated from the same mouse. The rates of blastocyst development were comparable for α6⁺CD34⁺ bulge keratinocytes (52.0%) and α6⁺CD34⁻ basal keratinocytes (59.3%), but no cloned pups were obtained from NT involving α6⁺CD34⁻ keratinocytes from female donors.

Table 1.

Production of cloned mice from skin keratinocytes and other somatic cells

Cell type	Sex (no. of replicates)	No. of reconstructed oocytes	No. of two cell (% of reconstructed oocytes)	No. of blastocysts (% of reconstructed oocytes)	No. of transferred blastocysts (no. of recipients)	No. of cloned mice (% of transferred blastocysts)	No. of surviving to adult (% of oocytes)
CC	F (12)	909	860 (94.6)	557 (61.3)*†	498 (26)	6 (1.2)	3 (0.3)
KSC	F (13)	781	735 (94.1)	406 (52.0)*†	366 (18)	6 (1.6)‡	4 (0.5)
TA	F (5)	432	406 (94.0)	256 (59.3)†	240 (12)	0 (0)§	0 (0)
Subtotal	F (30)	2,122	2,001 (94.3)	1,219 (57.4)	1,104 (56)	12 (1.1)	7 (0.3)
KSC	M (10)	490	471 (96.1)	242 (49.4)†	241 (14)	13 (5.4)‡	5 (1)
TA	M (8)	461	421 (91.3)	253 (54.9)†	253 (13)	7 (2.8)§	1 (0.2)
Subtotal	M (18)	951	892 (93.8)	495 (52.1)	494 (27)	20 (4.0)	6 (0.6)
KSC (embryonic)	NA (6)	424	354 (83.5)	157 (37.0)†	157 (7)	2 (1.3)	1 (0.2)
Total	54	3,497	3,247 (92.9)	1,871 (53.5)	1,755 (90)	34 (1.9)	14 (0.4)

F, female; M, male; NA, not applicable; KSCs, α6⁺CD34⁺ bulge keratinocytes; TA cells, α6⁺CD34⁻ non-bulge keratinocytes.

*, P < 0.05 (CCs vs. female KSCs);

†, P < 0.05 (embryonic KSCs vs. other donor cells);

‡, P < 0.05 (male KSCs vs. female KSCs);

§, P < 0.05 (male TA cells vs. female TA cells).

Fig. 2.

Mice cloned from adult skin keratinocytes and other cells by NT. (A) Cloned blastocysts from oocytes reconstructed by NT of bulge follicle keratinocyte nuclei into unfertilized mouse oocytes. The α6⁺CD34⁺ bulge keratinocytes were isolated by FACS from skin of adult female K14-Cre/ROSA-floxed-lacZ mice and used directly in NT. (B) Three mice cloned from the same donor mouse, an adult female. The pup on the left is from an α6⁺CD34⁺ bulge follicle keratinocyte nucleus, and the two pups on the right are from CC nuclei. (C) Three mice cloned from the same donor mouse, an adult male. The pup on the left is from an α6⁺CD34⁺ bulge follicle keratinocyte nucleus, and the two pups on the right are from α6⁺CD34⁻ keratinocyte nuclei. (D) Six-week-old mice cloned by NT from α6⁺CD34⁺ bulge follicle keratinocytes isolated from an adult male.

NT with Bulge and Non-Bulge Basal Keratinocytes of Male Mice.

CCs are widely used in mammalian NT research but do not exist in males. By contrast, α6⁺CD34⁺ and α6⁺CD34⁻ keratinocytes can be isolated equally well and with high purity from the skins of male and female mice. To compare the efficiency of male vs. female skin stem cells as NT donors, bulge and non-bulge basal keratinocytes were isolated from adult males by FACS as outlined above and used directly for NT.

The in vitro development of reconstructed oocytes into blastocysts was similar between α6⁺CD34⁺ and α6⁺CD34⁻ keratinocytes from adult males (49.4% vs. 54.9%). However, the birth rate after uterine transfer of cultured blastocysts was 2-fold higher when α6⁺CD34⁺ keratinocytes were used compared with α6⁺CD34⁻ keratinocytes from adult males (5.4% vs. 2.8% per transferred blastocyst; 2.6% vs. 1.5% per reconstructed oocyte).

Although rates of in vitro development were similar between sexes (49.4% in males vs. 52.0% in females for α6⁺CD34⁺ keratinocytes; 54.9% vs. 59.3% for α6⁺CD34⁻ keratinocytes), birth rates were consistently higher for males than females (5.4% vs. 1.6% per transferred blastocyst for α6⁺CD34⁺ keratinocytes; 2.8% vs. 0% for α6⁺CD34⁻ keratinocytes; P < 0.05). Fig. 2C shows three pups cloned from the nuclei of α6⁺CD34⁺ keratinocytes and α6⁺CD34⁻ keratinocytes harvested from the same male mouse.

Cellular Origin of the Cloned Mice.

Using PCR with LacZ primers, we next verified that genomic DNA isolated from the placentas of the cloned mice contained the LacZ gene (Fig. 3A). Although both CC and skin keratinocyte nuclei from K14-Cre/ROSA-floxed-lacZ mice harbored this gene, it was only recombined in cells in which the K14 promoter had been or was active at the time of NT. Correspondingly, only the tissues of a mouse cloned from a skin keratinocyte exhibited the broad, ROSA locus-driven expression of β-galactosidase that is reflective of this K14-Cre recombination event. Fig. 3 (B and C) illustrates this feature: X-Gal exposure of olfactory bulb tissue revealed strong staining in a mouse cloned from a keratinocyte but not in a mouse cloned from a CC isolated from the same donor mouse. Similarly, X-Gal staining of skin tissue showed that staining was restricted to the epithelium of a mouse cloned from a CC of a K14-Cre/ROSA-floxed-lacZ mouse, whereas a mouse cloned from a keratinocyte of the same donor mouse displayed X-Gal staining in the cells of both dermis and epidermis (Fig. 3 D and E, respectively).

Fig. 3.

Genetic evidence to document the cellular origin of the cloned mice. (A) LacZ-specific PCR analyses of genomic DNAs from mice cloned with nuclei of male bulge KSCs (lanes 1–4), a wild-type mouse (WT), and a positive control (ROSA). The cloned mice harbor the lacZ gene, indicative of the ROSA-floxed-lacZ background. (B–D) X-Gal immunohistochemistry of histological cryosections. (B) Left olfactory bulb of a 548-day-old mouse cloned from a CC. Counterstaining was with neutral red. There is no X-Gal precipitate, consistent with the origin of the nucleus from a K14-negative cell. (C) Right olfactory bulb of a 547-day-old mouse cloned from an α6⁺CD34⁺ bulge keratinocyte. The donor was the same female as in B; the mice differ 1 day in age because cloning was done on successive days. Counterstaining was with neutral red. The entire olfactory bulb shows an intense X-Gal precipitate, consistent with the origin of the nucleus from a K14-positive cell. (D) Tail skin of a mouse cloned from a CC. Only the K14-positive cells in the epidermis express β-galactosidase activity, as is the case in K14-Cre/ROSA-flox-lacZ mice. (E) Tail skin of a mouse cloned from an α6⁺CD34⁺ bulge keratinocyte. Because the mouse is cloned from a bulge nucleus in which the ROSA26 locus was recombined by Cre, all of the cells of this mouse now express lacZ from the ROSA26 promoter, resulting in a widespread X-Gal precipitate. The difference between staining patterns of D and E is indicated by arrows.

NT with Embryonic Keratinocytes.

In an effort to increase the efficiency of cloning from keratinocyte nuclei, we isolated GFP-tagged keratinocytes from skins of K14-H2BGFP mice at embryonic day 17.5 and used them directly for NT, without FACS. (Embryos were combined and not sexed.) The rate of in vitro development of embryonic keratinocyte donors was lower (37.0%) compared with adult α6⁺CD34⁺ and α6⁺CD34⁻ keratinocytes used after FACS (49.4–59.3%). Of the transferred blastocysts, two mice (1.3%) developed to term; this rate is comparable with the other cell types in this study. Thus, under the conditions used here, the use of embryonic instead of adult keratinocytes did not enhance the cloning efficiency.

Phenotype and Tissue Analyses of Cloned Mice.

A total of 32 liveborn and 2 stillborn cloned mice were produced in this study (Table 2). Of the liveborn pups, six died within 1 h. The other 26 pups rapidly gained active movement after birth, and 12 of these died within a few days, although they looked normal when they were presented to foster mothers. A total of 14 cloned offspring survived beyond weaning, of which 9 mice are from FACS-purified α6⁺CD34⁺ bulge keratinocytes of adult mice (Fig. 2D).

Table 2.

Survival of cloned mice

Donor (sex)	Date of birth	Age	Remarks
CC (F)	05/31/05	548 days	Killed for analysis, 41.8 g
	05/31/05	548 days	Died spontaneously, 45.9 g
	11/02/05	23 days	Died spontaneously
	11/16/05	9 days	Died spontaneously
	08/16/06	1 day	Died spontaneously
	11/16/05		Born dead
KSC (F)	06/01/05	547 days	Killed for analysis, 38.4 g
	11/03/05	>412 days	Alive and well, 31.8 g
	11/03/05	>412 days	Alive and well, 33.6 g
	10/27/06	144 days	Died spontaneously
	09/29/05	1 h
	08/09/05		Born dead
KSC (M)	04/14/05	>615 days	Alive and well, 45.4 g
	04/14/05	490 days	Killed, severe prolapsed rectum, 44.2 g
	04/15/05	267 days	Died, accidental water shortage
	07/08/05	203 days	Died, injured by fight, 28.8 g
	07/08/05	175 days	Died, injured by fight, 28.9 g
	04/19/06	1 day
	05/31/06	1 day
	05/31/06	1 day
	06/01/06	1 day
	06/01/06	1 day
	06/01/06	1 day
	06/08/06	1 day
	06/01/06	5 h
TA (M)	08/02/06	>139 days	Alive and well, 38.6 g
	05/31/06	1 day
	06/07/06	1 day
	06/08/06	1 h
	08/02/06	1 h
	08/02/06	1 h
	08/02/06	1 h
KSC (embryonic)	01/09/06	3 days
	01/23/06	1 h

F, female; M, male.

Most of the cloned pups were not oversized at birth. However, their placentae weighed 2- to 3-fold more than normal placentae, and placentae of clones derived from male α6⁺CD34⁺ and α6⁺CD34⁻ keratinocytes had a lower weight than those of females (Table 3).

Table 3.

Weight of cloned mice at birth and weight of their placentas

Donor cell	No. of pups born	Weight of pup, g (mean ± SD)	Placental weight, g (mean ± SD)
CC	6	1.77 ± 0.22	0.27 ± 0.04
KSC (F)	6	1.71 ± 0.32	0.21 ± 0.07
KSC (M)	13	1.42 ± 0.23	0.17 ± 0.07
TA (M)	7	1.52 ± 0.24	0.15 ± 0.01
KSC (embryonic)	2	1.54 ± 0.58	0.24 ± 0.05
Subtotal	34	1.59 ± 0.29	0.22 ± 0.06
Control mice (ICR)	14	1.33 ± 0.06	0.08 ± 0.01

F, female; M, male; KSC, α6⁺CD34⁺ bulge keratinocytes; TA cells, α6⁺CD34⁻ non-bulge keratinocytes.

In the neonatal mice that died within a few days after birth, stomachs were empty of milk, suggesting that the pups had not been nursed. Of these mice, six cloned from male α6⁺CD34⁺ keratinocytes and two cloned from male α6⁺CD34⁻ keratinocytes were subjected to necropsy. In all cases, tissues appeared to have undergone normal development. As judged by hematoxylin and eosin staining of fixed tissue sections, no signs of inflammation or necrosis were evident in any of the tissues examined. The majority of control and cloned lungs appeared to be well aerated, but some contained proteinaceous material, suggesting that amniotic fluid had not been completely cleared from airways. That said, foster mothers had vigorously rubbed and dried these pups after removing them from the uterus, and most pups could take at least a few breaths after being born. In only one case did we see a partial collapse in the bronchioles and alveoli of a lung, in a pup cloned from an α6⁺CD34⁺ bulge keratinocyte.

Discussion

Production of viable cloned mice from somatic cell donors of nuclei remains an inefficient process, despite major efforts in the last decade to improve efficiency above the single-digit percentage range (30). The efficiency of NT has remained particularly low when adult somatic cells have been used as donors. The differentiated state of the donor cell cannot automatically be deemed incompatible with the reprogramming of the nucleus and development to term of a cloned mouse.

One possibility that may account generally for the poor cloning efficiency of adult somatic cells is that only the rare adult stem cell present within donor populations may possess the requisite developmental plasticity to permit chromatin remodeling (13, 31, 32). When coupled with the apparent higher NT success rate of certain ES cell lines (if expressed per transferred blastocyst), an unresolved question is whether the nuclei of stem cells may be easier to reprogram than nuclei from more differentiated cells. Although it is an attractive hypothesis, recent studies with mesenchymal stem cells from adult cows (33), hematopoietic stem cells from adult mice (34, 35), and neural stem cells from cultured neurospheres of neonatal mice (36) did not show increased success rates of NT.

We have now performed NT with well defined populations of adult skin keratinocytes, enriched for either KSCs (α6⁺CD34⁺ keratinocytes from the hair follicle bulge) or TA cells (α6⁺CD34⁻ keratinocytes from the epidermis and follicle outer root sheath). By using adult females as a source of donor nuclei, we compared the cloning efficiency of CCs, KSCs, and TA cells from the same donor individuals. We identified adult bulge keratinocytes from male mice as the cell type with the highest cloning efficiency in our study: 5.4% of transferred blastocysts developed to term, or 2.6% when expressed per reconstructed oocyte. Our results provide compelling evidence that adult skin stem cells of mice can be used as NT donors.

Cloned animals have been produced previously from female and male cells (2, 3, 37–41). However, few studies have directly compared the cloning efficiency of donor cells between sexes. Although we found no difference in blastocyst rate between female and male bulge keratinocytes, the rate of cloned offspring was >3-fold higher for males compared with females. We were able to obtain cloned mice from male but not from female TA keratinocytes.

One sex-related difference was introduced in our protocol, where we used females as dual donors of CCs and keratinocytes on successive days. This process required injecting exogenous hormones to superovulate the females, which could have influenced subsequent cloning efficiency with their keratinocytes. That said, previous studies also suggest that the success rate may be higher when cloning with nuclei from male vs. female donor cells. In one study, only male clones developed to term when hematopoietic stem cells were used as donors, and the best rate of cloned offspring was achieved with immature Sertoli cells (34). Similarly, cells from tail tips of male mice performed better in NT than those of female mice (42). A 2-fold advantage in using male clones for NT has also been noted for other mammalian species (43, 44). Given these collective findings, it seems likely that the difference in cloning efficiency reflects an inherent difference in female cells, which face an additional demand of reprogramming that is imposed by X-chromosome inactivation (45).

It remains unresolved whether the overall inefficiency of NT is reflective of imperfect erasure of the epigenetic signature of the donor nucleus, trauma attributable to trypsinization and FACS purification of the skin keratinocytes, or technical difficulties in nuclear isolation and transfer. Likewise, it remains difficult to distinguish between biological and methodological factors to explain differences in NT efficiency.

Although most of our studies were conducted on populations of adult skin keratinocytes purified by FACS, we also examined the cloning competence of single-cell suspensions of embryonic skin keratinocytes. The in vitro development rate was lower. We speculate that the reduced efficiency is attributable to the proliferative nature of the embryonic keratinocytes, a feature that may be less conducive to NT than the more quiescent state of bulge cells. By contrast, bulge keratinocytes in their native niche are growth- and differentiation-inhibited, features that could more closely resemble those of the nucleus within an unfertilized oocyte.

Our studies suggest that adult bulge KSCs (α6⁺CD34⁺ keratinocytes from the hair follicle bulge) are a promising new model system for mouse NT research, particularly when nuclei from male bulge follicles are used. There are several advantages of adult KSCs as a preferred choice of adult cells for NT: (i) Bulge KSCs are readily accessible and can be isolated from both females and males; (ii) bulge KSCs reside in a protected niche and cycle less frequently than their TA counterparts; (iii) bulge KSCs reside in the G₀/G₁ stage, which may be beneficial to the development of reconstructed embryos; and (iv) nuclei from bulge KSCs can be injected relatively easily into enucleated oocytes because of their small size. When coupled with the fact that bulge KSCs have undergone fewer divisions than most cell types within tissues, their accessibility and relative quiescence may make them more promising than adult mesenchymal stem cells (33), adult hematopoietic stem cells (34, 35), or neonatal neural stem cells (36).

Materials and Methods

Donor Mice.

Donor mice were a cross of Gt(ROSA26^tm1Sor) (27) and K14-Cre transgenic mice (28). The GtROSA26^tm1Sor mice contain a LacZ gene flanked by loxP sites and were in a mixed background containing 129S4/SvJaeSor and 129S1/SvImJ. The K14-Cre transgenic mice (28) were in an outbred CD1 background. Seven- to 8-week-old mice were used as donors for NT and are referred to as “adult mice” in this paper. Embryos from K14-H2BGFP mice (46) were in an outbred CD1 background.

Preparation of Donor Cells.

CCs and keratinocytes were collected from the same donor mice. Females were superovulated with 5 units of pregnant mare serum gonadotropin (Sigma, St. Louis, MO) and 5 units of human choriogonadotropin (Sigma) 48 h later. CC collection was performed 14 h after human choriogonadotropin injection, and skins were processed overnight. FACS analyses and purification of bulge cells from adult mouse back skins have been described (21, 22). Cell isolations were performed on a FACSVantage SE system equipped with FACS DiVa software (BD Biosciences, Franklin Lakes, NJ). Cells were gated for single events and viability based on propidium iodide staining and were then sorted according their expression of α6 integrin and CD34. Purity of sorted cells was determined by postsort FACS analysis and by immunofluorescence analyses of cytospin FACS population. The purity of the sorted cells was determined by the analysis of epidermal markers. Sorted cells were cytospun and immunostained for the epidermal markers keratin 1, keratin 5, and β4 integrin (22); purity, which was >98%, was calculated based on keratin 5 and β4 integrin positivity. All cells were washed two times with Hepes–Chatot, Ziomet, and Bavister (CZB) medium and suspended in Hepes–CZB medium containing 3% (wt/vol) polyvinylpyrolidone.

Skin from K14-H2BGFP mice at embryonic day 17.5 was removed and incubated in dispase for 1 h at 37°C. The epidermis was separated from the dermis and treated with trypsin for 15 min at 37°C to obtain a single-cell suspension. To verify the epithelial nature of these cells, GFP fluorescence was evaluated on cytospin preparations. GFP-positive cells were used directly for NT.

NT.

NTs were performed as described previously (3, 11). Briefly, mature oocytes were collected from superovulated B6D2F1 females (8–10 weeks), and CCs were removed by hyaluronidase. Enucleation was performed in a droplet of Hepes–CZB medium containing 5 μg/ml cytochalasin B (CB) by using a blunt piezo-driven pipette. After enucleation, the spindle-free oocytes were washed extensively and kept in CZB medium up to 2 h before nucleus injection. Nuclei were injected into enucleated oocytes directly by using a piezo-driven pipette. The reconstructed oocytes were cultured in CZB medium for 1–3 h and then activated for 5–6 h in calcium-free CZB medium containing 10 mM Sr²⁺ and 5 μg/ml CB. After activation, reconstructed embryos from CCs were cultured in somatic cell medium MEMα with 1 mg/ml BSA, and reconstructed embryos from keratinocytes were cultured in K⁺-modified simplex optimized medium with amino acids at 37°C under 5% CO₂ in air. Reconstructed embryos that reached the morula or blastocyst stage by 3.5 days in culture were transferred into the uterine horns of 2.5-day-postcoitum pseudopregnant ICR females mated with vasectomized males. On day 20, the recipient females were subjected to cesarean section, and live pups were nursed by lactating ICR females.

Histochemistry.

β-Galactosidase activity assay was performed as described (47).

Histopathology.

Mice were killed by CO₂ asphyxiation. Complete postmortem evaluations were performed on all mice. All tissues were fixed in 10% neutral buffered formalin. All tissues were processed by routine methods and embedded in paraffin wax. Sections (5 μm) were stained with hematoxylin and eosin (HE) and evaluated with an Olympus BX45 light microscope (New York/New Jersey Scientific, Middlebush, NJ).

Statistical Analysis.

Development of reconstructed embryos at different stages was compared between groups by using Fisher's exact probability test.

Abbreviations

CC

cumulus cell

CZB

Chatot Ziomet and Bavister

K14

keratin 14

KSC

keratinocyte stem cell

NT or nt

nuclear transfer

TA

transit amplifying.