Morphological abnormalities, impaired fetal development and decrease in myostatin expression following somatic cell nuclear transfer in dogs

Abbreviations:

BWS, Beckwith–Wiedemann syndrome; IGF, insulin like growth factor; IVP, in vitro-produced; LOS, large offspring syndrome (also, abnormal offspring syndrome); MFD, mean lesser fiber diameter; MRF, myogenic transcriptional regulatory factor; MRF4, myogenic factor 6; Myf-5, myogenic factor 5; MyoD, myogenic differentiation 1; SCNT, somatic cell nuclear transfer.

INTRODUCTION

Mammals have been successfully cloned for several years by transferring nuclei from somatic cells into enucleated oocytes. Cloning dogs by somatic cell nuclear transfer (SCNT) has been well described since the successful birth of “Snuppy,” the first SCNT-cloned dog (Lee et al., 2005), but this technology has not been routinely successful with dogs due to the difficulty of maturing canine oocytes in vitro. Despite the success of cloning various mammals, SCNT-cloned animals have abnormalities in morphology, gene expression, and metabolism. These phenotypic instabilities include “large offspring syndrome” (LOS) or “abnormal offspring syndrome” (AOS) (Wilson et al., 1995). These syndromes are well recognized in cattle and sheep SCNT embryos, and in pregnancies from in vitro-produced (IVP) embryos (Wilson et al., 1995; Young et al., 1998; Farin et al., 2006). They include a high rate of early embryonic death and abortion, production of large-size fetuses, musculoskeletal deformities, disproportionate fetal growth, abnormal organ growth, failures in the development of the allantois, and placental vascular abnormalities and development, including hydrallantois (Farin et al., 2006). Thus, LOS represents a wide range of developmental abnormalities that may arise from different mechanisms (Farin et al., 2006). Although problems generated by a variety of SCNT and IVP procedures have been reported to be associated with LOS, the occurrence of these problems is not predictable or reproducible (Young et al., 1998). Moreover, phenotypic instabilities in SCNT-cloned dogs have not been reported. Therefore, in the present study, we examined phenotypic abnormalities, resulting from SCNT, including increased muscle mass and macroglossia, in stillborn or deceased neonatal SCNT-cloned dogs.

RESULTS

There Are Morphological Abnormalities in SCNT-Cloned Dogs

We necropsied 12 SCNT-cloned dogs, and examined their morphologies relative to normal dogs. Dogs that failed to survive from previous studies were used. Dogs No. 1 and 2 were from Kim et al. (2009), dogs No. 3 and 4 from Hossein et al. (2009a), and dogs No. 5–12 were from Hossein et al. (2009b). The SCNT methods used in these studies were slightly different, but the major characteristics of abnormal cloned dogs were similar. These dogs were stillborn just prior to delivery or died with dyspnea shortly after birth. Dog No. 2 had incurred severe postmortem autolysis so it was not possible to render necropsy and examination. It had been diagnosed as stillborn in its mother's uterus 1 day prior to delivery and was excluded from examination. Ten of the 12 dogs exhibited identical morphological lesions. Specifically, they showed macroglossia (Fig. 1B), marked increases in skeletal muscle mass (Fig. 1D), and anterior abdominal wall defects (Fig. 1E). Mild to moderate edema or atelectasis lesions were observed in their lungs (Fig. 1F). Their tongues were very thick and big, protruded out of the mouth, and obstructed most of the throat area. This anatomical abnormality may induce dyspnea, and subsequent death. Increased heart and liver size were also observed. The normal cardiac silhouette in dogs should be 3–3.5 times the width of the intercostals space on the lateral view (Love and Berry, 2002); the size of the hearts in the SCNT-cloned dogs was more than four times the width of the intercostals space (Fig. 1G). The caudal margin of the normal liver is enclosed within the rib cage or is very close to the costal margin (Love and Berry, 2002); the livers of SCNT-cloned dogs were large enough to occupy most of the abdominal cavity region (Fig. 1H). Dog No. 10 exhibited a normal sized skeletal muscle mass and tongue (Fig. 1A,C). However, its heart and liver size were also increased, and other lesions were similar to those of the other dogs, with the exception of failed medulla development in the cerebrum (inset of Fig. 1C).

Skeletal Muscle and Tongue Hypertrophy in SCNT-Cloned Dogs

The most remarkable morphological changes observed in SCNT-cloned dogs compared with dog No. 10 were large skeletal muscles and tongues. Therefore, we microscopically examined their tongues and skeletal muscles. The tongue sizes of ten SCNT-cloned dogs (No. 1, 3–9, 11, 12) were larger than that of dog No. 10, which had normal-sized tongue and muscles (Fig. 2A,B). The normal-sized tongue of dog No. 10 contained relatively uniform-sized myofibers (Fig. 2C), whereas the tongues of the ten other SCNT-cloned dogs (No. 1, 3–9, 11, and 12) exhibited irregularly sized myofibers, with small and large myofibers mixed in a muscle bundle (Fig. 2D). The mean lesser fiber diameter (MFD) in the tongues of the ten other SCNT-cloned dogs was significantly larger than that of myofibers in dog No. 10 (Fig. 2E, Table 1). The number of myofibers containing internal nuclei did not show significant changes in the tongues (Fig. 2F, Table 1). Skeletal muscles of the other ten SCNT-cloned dogs also displayed significant hypertrophy compared with skeletal muscle of dog No. 10 (Fig. 3A–D, Table 1). Normal skeletal muscle cells of newborn-to-1-week-old dogs contain relatively uniform populations of myofibers, and larger cells are occasionally observed having internal nuclei (Braund and Lincoln, 1981). MFD of most their fibers was reported as 7–9 µm (Braund and Lincoln, 1981), and also MFD of dog No. 10 was 7.44 ± 0.58 µm. However, the skeletal muscle cells of the ten other SCNT-cloned dogs exhibited irregularly sized myofibers, and their MFD was 13.15 ± 1.97 µm. Moreover, many internal nuclei were observed in their myofibers compared to those of dog No. 10 (Fig. 3D). Myofibers containing internal nuclei were usually larger than any other myofibers in the ten other SCNT-cloned dogs. Thus, the MFD and number of internal nuclei in skeletal muscle were significantly greater in the ten other SCNT-cloned dogs compared with dog No. 10 (Fig. 3E,F).

Table 1. Mean Lesser Fiber Diameters (MFD) and the Average Number of Fibers With Internal Nucleus

ID of cloned dog	MFD ± SD		Internal nuclei (mean ± SD)
	Tongue	Skeletal muscle	Tongue	Skeletal muscle
1	16.99 ± 1.28	12.28 ± 0.34	0.4 ± 0.5	8.0 ± 5.2
2	NA	NA	NA	NA
3	12.46 ± 0.52	9.23 ± 0.53	0.6 ± 0.5	2.8 ± 1.8
4	12.49 ± 0.47	NA	0.4 ± 0.5	NA
5	16.72 ± 2.10	16.64 ± 3.23	1.4 ± 1.5	7.2 ± 1.6
6	14.86 ± 2.68	12.81 ± 2.12	1.8 ± 1.6	6.6 ±2.1
7	21.04 ± 3.44	14.14 ± 1.29	1.2 ± 0.4	7.6 ± 3.9
8	13.62 ± 1.95	13.22 ± 1.67	2.8 ± 1.3	2.8 ± 0.4
9	NA	14.24 ± 1.34	NA	5.0 ± 2.3
10	10.21 ± 0.34	7.44 ± 0.58	0.8 ± 1.1	2.0 ± 1.2
11	14.57 ± 1.51	12.46 ± 1.44	1.8 ± 0.4	5.6 ± 1.9
12	15.47 ± 2.23	13.33 ± 0.77	1.2 ± 1.3	7.0 ± 5.6

• NA, not applicable.

Expression of Myogenic Regulatory Factors and Myostatin Genes Is Altered in SCNT-Cloned Dogs

We speculated that abnormal myogenesis may possibly occur in SCNT dogs. To test this hypothesis, we examined the mRNA expression level and DNA content of myogenic-related genes using RT-PCR and PCR analyses. Embryogenic myogenesis in vertebrates is controlled by a series of complex myogenic transcriptional regulatory factor (MRF) networks including myogenic factor 5 (Myf-5), myogenic differentiation 1 (MyoD), myogenin and myogenic factor 6 (MRF4), and myogenin (Bryson-Richardson and Currie, 2008). In addition, myostatin has been reported to be a negative regulator of muscle mass in skeletal muscle cells (McPherron et al., 1997). Therefore, we examined mRNA expression and DNA content of MRFs and myostatin in the tongues and skeletal muscles of four representative SCNT-cloned dogs (Fig. 4). The No. 10 SCNT-cloned dog was used as a normal control, and the biopsied triceps brachial muscle of a normal adult (4 years old) mixed breed dog was used as a standard control. There were no significant differences in the DNA content level of MRFs and myostatin among the animals. At the mRNA level, all factors were highly expressed in skeletal muscle of the adult dog. Dog No. 10 also showed a high level of mRNA expression of MRFs and myostatin, but not myogenin. However, expression of the MRFs and myostatin was down-regulated or absent in the tongue and skeletal muscle of the other SCNT-cloned dogs compared with dog No. 10 and the adult dog.

DISCUSSION

Since the first report of a clone produced by SCNT (Wilmut et al., 1997), several other species have been cloned. However, SCNT currently has many limitations, including a low success rate. The limited success of SCNT is often related to incorrect nuclear reprogramming of donor cells. The underlying potential mechanisms that may affect reprogramming efficiency include epigenetic modifications such as DNA methylation and histone remodeling (Young and Fairburn, 2000; Dean et al., 2001; Farin et al., 2004). Different cells and tissue types acquire different programs of gene expression and epigenetic modification during development (Bird, 2002; Li, 2002). It has been thought that the epigenetic dysregulation caused by SCNT or in vitro production may lead to inappropriate expression of key developmental genes that may contribute to lethality of cloned embryos (Farin et al., 2006).

Large Offspring Syndrome (LOS), characterized by excess fetal size, enlarged organs, fetal hydrops, respiratory distress, etc., has been observed during late gestation and has been shown to cause perinatal deaths in cloned cattle, sheep, and mice (Wilson et al., 1995; Wakayama et al., 1998; Young et al., 1998; Eggan et al., 2001; Chavatte-Palmer et al., 2002; Constant et al., 2006; Farin et al., 2006; Everts et al., 2008). Offspring with LOS are frequently born following SCNT or from embryos exposed to in vitro culture environments (Wilson et al., 1995; Farin et al., 2004; Young et al., 1998; Eggan et al., 2001). An enlarged and edematous placenta is an important abnormal morphology that is present in all cases of LOS (Constant et al., 2006; Everts et al., 2008). Altered expression of mRNA or protein expression of genes such as vascular endothelial growth factor (VEGF), peroxisome proliferators-activated receptor γ (PPARγ), leptin, bovine placental lactogen, transforming growth factor (TGF)-β, and TGF-β receptors have been found in the placentas from IVP or SCNT pregnancies (Farin et al., 2006). However, there are few reports regarding altered candidate genes in cloned animals caused by reprogramming errors.

LOS is clinically correlated in humans to Beckwith–Wiedemann syndrome (BWS), which is associated with imprinting defects (DeBaun et al., 2003). BWS is a rare congenital overgrowth syndrome with considerable phenotypic heterogeneity. The critical, causative region for BWS is located on chromosome 11p15, and the major features are anterior abdominal defects, macroglossia, and pre-or postnatal overgrowth (Elliott et al., 1994; Whisson et al., 1994). However, patients with BWS are not phenotypically identical to SCNT-cloned dogs. BWS patients display other complex features such as a cleft palate, ear abnormalities, and embryonal tumors, which was not evident in SCNT dogs (Elliott et al., 1994; Whisson et al., 1994).

Myostatin-deficient animals exhibit similar disorders to LOS animals (Lee, 2004; Mosher et al., 2007). Animals with mutations in myostatin are characterized by increased muscle mass resulting from a combination of an increased number of muscle fibers and increased fiber size (Lee, 2004). The myostatin sequence is highly conserved across species, and mutations in the myostatin gene have been observed in mice, cattle, sheep, dogs, and humans. These mutations frequently result in a large and widespread increase in skeletal muscle mass (Lee, 2004; Schuelke et al., 2004; Mosher et al., 2007). Interestingly, the increased muscle phenotype in myostatin-mutated animals appears to be similar to the phenotype we observed in our SCNT-cloned dogs except for macroglossia.

The MRFs including MyoD, Myf-5, myogenin and MRF4 coordinate differentiation of skeletal muscle fibers during development (Gerrard and Grant, 1994). MyoD and Myf-5 are expressed in proliferating myoblasts before the onset of muscle differentiation (Emerson, 1990; Cornelison et al., 2000). Myogenin and MRF4 regulate terminal differentiation of myoblasts after the establishment of myotubes (Cornelison et al., 2000; Bryson-Richardson and Currie, 2008). Although myostatin is not a MRF, it regulates the final number of muscle fibers that are formed during embryogenesis (Lee, 2004). Manceau et al. (2008) revealed that myostatin acts in vivo to regulate the balance between proliferation and differentiation of embryonic muscle progenitors by promoting terminal differentiation. Moreover, Crosier et al. (2002) reported that there were no differences in expression of MRFs, although the expression of myostatin mRNA was down-regulated in skeletal muscles of IVP-derived fetuses compared with fetuses from embryos produced in vivo. In our case, we assumed that expression of MRFs might be increased in cloned dogs because we observed an increase in muscle fibers with internal nuclei, which is a characteristic of muscle regeneration. However, we observed undetectable or low mRNA expression levels of MRFs.

Embryogenesis and development are regulated by various genes. Hyperplasia and hypertrophy of the muscle are regulated by various growth factors, including hepatocyte growth factor, fibroblast growth factor 2, TGF-β, and insulin like growth factor (IGF) (Allen and Goll, 2003). IGF is a growth hormone that has been shown to play an important role in the development of mammals (Cingel-Ristic et al., 2004; Walenkamp and Wit, 2006). IGF-I has been reported to play a major role in the development, growth, differentiation, and maintenance of skeletal muscles (Florini et al., 1996). IGF-II is essential for normal embryonic and early fetal growth, and it has been suggested to play a role in the pathogenesis of BWS and in myostatin-mutated cattle (Gerrard and Grant, 1994; Best et al., 2006). Interestingly, myostatin is a member of the TGF-β family, and IGFs may regulate the expression and secretion of TGF-β during cell proliferation and initiation of differentiation (Florini et al., 1996). Therefore, we hypothesized that a disruption in myostatin gene expression may contribute to abnormalities such as increased muscle mass and macroglossia in SCNT-cloned dogs. However, this mechanism is not clear in our SCNT-cloned dogs. Therefore, a further investigation is required to determine the relationship between reduced myostatin gene expression and epigenetic modification during cloning. In addition, it is possible that other genes, which were not investigated in this study, may be responsible for the morphological abnormalities observed in SCNT-cloned dogs.

In summary, we observed the macroglossia, increased muscle mass, increased heart and liver sizes, and anterior abdominal wall defects in ten SCNT-cloned dogs. The reduced expression of myostatin, which is a negative regulator of skeletal muscle mass, was particularly observed in their large muscles and tongues. The underlying mechanisms of this decrease in myostatin expression and epigenetic dysregulation in SCNT-cloned dogs are not clear. However, the present study demonstrates that myostatin is a candidate gene that contributes to phenotypic developmental abnormalities in SCNT-cloned dogs.

MATERIALS AND METHODS