Yonai et al. 2005

Study overview: In the most comprehensive study of reproductive function in cattle clones, Yonai et al. (2005) (previously mentioned in other Developmental Nodes) performed an extensive analysis of reproductive performance in Holstein and Jersey clones with shortened telomeres, including puberty onset, estrus behavior, hormone cycling, the appearance of follicular waves, fertility and birthing for three estrus cycles. Once puberty onset had been determined, ovulation and formation of corpora lutea were monitored thrice weekly, with plasma samples to monitor progesterone levels collected every three days. After puberty, the estrus behavior of the clones was monitored twice daily until the animals became pregnant, with the length of the estrous periods and occurrence of standing behavior recorded. Plasma samples and ultrasonography were used to identify follicular waves and monitor progesterone and 17-ß estradiol concentrations between day 18 of estrus and the day of ovulation over 17 estrus cycles in the Jersey clones and 28 estrus cycles in the Holstein clones. All clones were bred by artificial insemination using semen from the same lot of one bull (breed unspecified). Pregnancies were diagnosed by ultrasonography at 40 days after AI. For the first and second postpartum cycles, all clones were artificially inseminated at first estrus, which usually occurred 90 days after parturition. The length of gestation and resulting calves’ birth weights were recorded. Table VI-6 summarizes the data collected in this very detailed study.

Walker et al. (2002) have reported on the largest litters of piglets produced by SCNT. Donor cells were derived from Duroc fetal fibroblasts, and fused with in vitro matured oöcytes. A total of 511 embryos were transferred into five surrogate dams, with between 59 and 128 embryos per recipient. All five recipients were confirmed pregnant by ultrasound between days 28 and 40 post-implantation. Four of the five pregnancies went to term, and litters containing between 5 and 9 piglet clones (total of 28) were delivered. Three of the four surrogate dams were induced and delivered on gestational day 115. The fourth was allowed to deliver naturally, and produced her litter on gestational day 117. One of the 28 clones was stillborn, but no abnormalities were noted on necropsy. One of the live born clones presented with anal atresia (no anus or tail), and was the smallest of all of the clones (birth weight of 0.72 kg, and crown rump length of 23.5 cm). The authors noted that anal atresia is a developmental abnormality seen at a natural low frequency in conventional piglets. The question of whether this is a random event due to genetic or inappropriate reprogramming cannot be answered from this dataset.

The remaining piglets had birth weights that appear to be a little lower than conventional piglets of the same breed. The authors noted with explicit surprise that there was little correlation between litter size, placental weights, and fetal weights (Table VI-8). They predicted a correlation of 0.639 between placental and fetal weight, but noted that the lowest mean birth weights occurred in the litters with the smallest number of piglets. The authors asserted that without the appropriate controls for litter size, in vitro oöcyte maturation and other manipulations, it is inappropriate to assign the SCNT process as the cause of the difference in birth weights. Two of these litters subsequently served as the source of the clinical and behavioral studies of Archer et al. (2003 a, b).

Viagen Inc. provided birth weights of seven male swine clones as part of the data package presented to CVM. Clones were smaller at birth than AI comparators of similar genetic background (See Appendix G: Viagen Dataset). No detailed health data were available on these clones for this developmental node. All clones survived the neonatal period.

Additional data submitted to CVM included birth weight, average daily weight gain (ADG), body temperature, and pulse rates on another cohort of neonatal swine clones (see Chapter V). Birth weights for three clones ranged from 1.1 to 1.4 kg, and ADG ranged from 0.46 to 0.55 kg; however, because the breed of swine was not identified, it is not possible to determine whether these data are within normal ranges. The report indicated that two of the five piglets, both from the same litter and weighing 1.0 kg at birth, died within the first 48 hours. The cause of death was not reported, and no other details were provided. Body temperatures of the piglets were low (range 98.8 to 101.8°F) during the first 48 hours compared to reference body temperature for adult swine (102-103°F). This finding is not unusual, however, as neonatal swine generally have difficulty regulating body temperature, and require supplemental heat after birth (see Chapter V).

(b) Summary Statement on the Embryo/Fetal to Perinatal Developmental in Swine Clones (Developmental Nodes 1 and 2)

The production of swine clones differs from the other livestock species discussed in this risk assessment because of the requirement for a minimum number of viable fetuses to maintain the pregnancy. The gestational losses observed are a function of the combined low “success rate” for embryonic and fetal development for the individual clone and the requirement for a minimum number of growing fetuses to implant. Clone piglets do not appear to exhibit the overgrowth phenomena observed in cattle, and if anything, newborn swine clones may be smaller than their non-clone counterparts. Although there is one report of an anomaly at birth (e.g., anal atresia), piglet clones appear to be normal and healthy.

ii. Juvenile Development and Function in Swine Clones (Developmental Node 3)

(a) Peer-reviewed Publications

Archer et al. (2003 a,b) have investigated the degree of behavioral and physiological variability exhibited among litters of swine clones and their closely related conventional siblings. The derivation of these clones has been described in Walker et al. (2002). The clone cohorts consisted of two litters of 5 and 4 female swine derived from the same cell line born 6 weeks apart. The control groups consisted of a litter of four female full siblings (both parents in common) and a litter of four female half-siblings taken from three sows mated to the same boar. All animals were farrowed (born) in conventional farrowing crates, and weaned at 5-6 weeks of age when they were placed in adjacent identical pens and given continuous access to identical standard rations and water. Results in these studies were presented as means and ranges; individual animal data were not provided.

One study (Archer et al. 2003b) evaluated behavioral characteristics including food preference (for apples, bananas, saltine crackers, and carrots), temperament (as judged by time to remove a towel placed on the pig’s head, attempts to escape mild restraints, being placed on their backs, and being lifted off the ground), and time budgets (the amount of time spent engaged in a particular activity in their pens). The results of this study indicated that the behaviors of swine clones were no more homogenous than the behaviors of siblings and may be more variable than the comparator animals, although the statistical power to draw such a conclusion was limited. The authors conclude that “...using nuclear transfer to replicate animals to reproduce certain behavioral characteristics is an unrealistic expectation.” The relevance of the study to an evaluation of the health of swine clones, however, is that the animals behaved in much the same manner as conventional animals, and displayed no behavioral anomalies at the times tested (15-16 weeks of age for the food trials, 8-9 weeks and 14-15 weeks for the towel test, 7 weeks for the restraint tests, and 13-15 weeks for the time budget tests)

Another study performed by this group (Archer et al. 2003a) evaluated whether the SCNT process introduced epigenetic changes into animal clones that could be manifested at the genomic (e.g., methylation status) (See Chapter IV), physiological (e.g., blood chemistry), and anatomical (e.g., weight, size, coat) levels. Body weights of all the animals overlapped and were within the normal range for the age and breed, with the exception of a single clone that was small at birth, and never attained the size of its littermates. This is likely a case of “runting,” which is observed in conventional animals as well. Teat number was the same for all animals (6,6 distribution) except for one clone (6,7 distribution), within the normal variability in conventional pigs.

One of the clones also exhibited an unusual hair growth pattern (e.g., longer and sparser), which the authors state prompted an examination of the histology of the skin. Results of that investigation indicated that with one exception, skin morphology showed no unusual variations among the pigs. The exception was a clone that exhibited morphology indicative of hyperkeratosis.¹¹ Hyperkeratosis, also referred to as parakeratosis, also occurs in naturally bred and AI pigs between the ages of 6 and 16 weeks, and is generally associated with zinc and essential fatty acid deficiency or excess dietary calcium or phytates. Gastrointestinal disorders may also affect zinc absorption, and contribute to the development of this condition (Cameron 1999). Other possible causes of hyperkeratosis include heredity, and other non-specific causes of skin inflammation (Blood and Radostits 1989). Dermatitis vegetans is the inherited form of this disease in swine, and is a semi-lethal recessive gene (Blood and Radostits 1989). Although the phenotypic variation is interesting, it is of limited concern for food safety, as pork skin that exhibits severe hyperkeratosis would be condemned at slaughter.

Blood samples were taken from the animals for analysis at 15 and 27 weeks of age (Table VI-9) (Archer et al. 2003a). Although the hypothesis being testing in this study addressed the degree of variability among clones relative to the degree of variability among controls, these data are very instructive in that they provide the most extensive analysis of the physiological status of swine clones at two different times in development. (See previous discussion of Cyagra dataset). Unlike the Cyagra dataset, however, very few

animals were evaluated (nine clones and eight controls). Nonetheless, the data are compelling in that they demonstrate that the physiological parameters investigated do not indicate any material differences between the clones and controls. In addition, they provide confidence that these clones are responding appropriately to age-specific signals. Just as the Cyagra cattle clones, the piglet clones initially exhibit relatively high alkaline phosphatase levels: at week 15 both clones and controls have mean levels of 209 and 235 U/L, respectively, while 8 weeks later (at 27 weeks of age), the mean alkaline phosphatase levels have decreased to 101 and 118 U/L, respectively. (Alkaline phosphatase provides a measure of bone growth in young animals.) Phosphorus levels, also an indicator of bone growth, show similar age-related changes, as does T3.

Genomic methylation levels (also discussed in Chapter IV) were evaluated in two repeated sequences, one found at the centromere¹² and the other in the euchromatin¹³ regions of the chromosomes (Archer et al. 2003a). The investigators discovered that one euchromatin region of clones had a different degree of methylation from controls. They further observed that another region had an increase in the variability of the degree of methylation in clones relative to controls. The investigators stated that it was not possible “to prove cause and effect” between alteration in methylation patterns and any of the measurements that they had taken on these animals. Additionally, because all of the animals in this study (clones and controls) appear to be healthy, with the exception of the pig with hyperkeratosis/parakeratosis, the developmental relevance of these methylation changes are not clear. It may be that animals that do not survive have higher degrees of variability or derangement of methylation patterns, and that what is being observed in this study is a set of animals that has adapted to or compensated for differences in methylation, or the inherent tolerance of biological systems for changes in methylation status of genes.

The authors concluded that “while cloning creates animals within the normal phenotypic range, it does affect some traits by increasing variability associated with that phenotype.” Their final conclusion with respect to phenotypic variability among clones was that they were not necessarily less variable than their closely related, sexually reproduced half-siblings. For pigs, at least, this implies that although genetics may have a strong influence, various environmental influences, including intra-uterine environments, may play a significant role in eventual phenotype of the animal.

(b) Viagen Dataset

The data on which the following discussion is based are found in Appendix F, along with a more detailed description of the results of the study.

Two groups of swine clones were used for Viagen Study 1. In the first group, seven clones (1 Duroc and 6 Hamline) were evaluated for survival, health, growth, meat and carcass characteristics. Fifteen conventional barrows (young males) (all Hampshire) were selected as comparators. Because the study was initiated after the birth of the clones (delivered by C-section), the observation period did not begin until shortly after they were weaned. Clones were followed from 50 days after birth through slaughter at approximately 6 months of age. comparators¹⁴ were selected as age-matched pigs selected from litters sired by the Hampshire nuclear donor boar in a conventional breeding (AI) program.

Clones raised to slaughter weight (approximately 270 lbs) took on average 27 days longer to reach that approximate weight, and when finally slaughtered tended to weigh less than their comparators. These observations are likely due to the husbandry of the clones. Further, because these animals were delivered via C-section, they faced additional stress during the earliest stages of life. In addition, because of the late initiation of the study, these clones were raised under pathogen-free conditions until 50 days of age before being transferred to more conventional (pathogen containing) rearing conditions, while all of the comparators were raised under conventional conditions. Clones also did not receive colostrum, and were deprived of passively transferred maternal immunity. Combined with the change from pathogen-free rearing, the significant immune challenge that the clones experienced would have slowed growth as the animals adapted to their new environment, regardless of whether the animals were produced by sexual reproduction or nuclear transfer.

Four of the clones exhibited appropriate responses to the immune challenge and were able to adapt and grow, albeit at a lower rate than animals which had been raised under more conventional conditions from birth. Three of the clones were considered “poor-doers:” animals that exhibit slow growth rates and other health problems, such as chronic scouring. At slaughter, organ weights as a percentage of body weight were smaller for clones than for their comparators. The clones also had lower blood IGF-I and estradiol levels in their blood than comparators. It is unknown how the change from a pathogen free environment to a more conventional one may have impacted organ weights or hormone status, as none of the comparators were subjected to similar immune challenges because they were all raised under conventional conditions from time of birth (Appendix F). However, given the physiological and immunological stress that the animals experienced, in the opinion of CVM’s veterinarians, these animals performed as well as could possibly be expected.

The second group of swine clones was used to study reproductive function; that discussion is found in Section iii.

(c) Summary Statement for Juvenile Development and Function in Swine Clones (Developmental Node 3)

The dataset reported by Archer et al. (2003 a,b) in which both behavior and physiological variables were measured on an individual animal basis is the larger and more tightly controlled study of the two studying juvenile swine clones. Those studies indicate that swine clones overlap their conventional counterparts in behavior and health, and that there are no significant differences between the two groups. Measures of age-appropriate physiological responses (e.g., alkaline phosphatase, phosphorus, and serum protein indicative of increased globulins) indicate that clones are responding normally to growth signals. One case of parakeratosis was observed in this clone cohort. It is not possible to determine whether its incidence was due to SCNT, as it is a condition that is also present in conventional pigs. The Viagen dataset is less-well controlled, and its outcome confounded by the unconventional shift from a pathogen free environment to more conventional husbandry. Based on the lack of colostrums immediately after birth, and the transfer from pathogen free to conventional housing, it is not possible to ascribe any of the differences in growth to cloning. Further, most of the animals were able to respond appropriately to immune challenges. None of these outcomes were observed in the studies of Archer or Martin, again implying that the changes in husbandry were likely responsible for the outcomes. Finally, none of the swine clones exhibited any adverse outcomes that have not been observed in conventionally bred and reared swine.

iii. Reproductive Development and Function in Swine Clones (Developmental Node 4)

(a) Peer-reviewed Publications

Martin et al. (2004) described birth outcomes of clone females which were mated via artificial insemination to clone males as normal in duration and uneventful. The 62 live offspring of the clone X clone mating were reported to be normal at birth with the exception of one pig that had contracture of the flexor tendons of both hind limbs. The authors reported that the rate for this abnormality (1.6 percent) was similar to estimates of the frequency within the Australian swine industry (1.2 percent). The stillborn rate for the clone offspring litters was 4.5 percent while a comparator group had a stillborn rate of 8 percent. Evaluation of the semen from the boars, showed similar ejaculate volume, sperm concentration and motility between the clones and comparators. These investigators further reported that 100 percent of gilt clones (5) became pregnant following insemination at second estrus. Consequently, the limited data indicate that gilts and boars from cloning mature similar to non-clones.

Reproduction was measured in four boar clones in the Viagen dataset (Appendix F). The four clones (three Hampshire and one Duroc) were compared to three genetically related boars derived by AI. No differences were observed between clones and comparators in semen quality. Farrowing rates were higher for swine clones than comparators (73.5 vs. 62.5 percent), although this difference was attributed to the fact that the Hampshire comparator was five years old, and may have been nearing the end of his reproductive life. Litter size was more variable for boar clones, and mean litter size was slightly smaller for clones vs. comparators (10.94 vs. 11.76 pigs/litter), but were similar to US commercial swine production (10.66 pigs/litter).

iv. Post-Pubertal Maturation in Swine Clones (Developmental Node 5)

CVM was not able to identify any peer-reviewed studies on non-reproductive post-pubertal studies in swine clones. The Viagen dataset (Appendix F) indicated that

No remarkable differences were observed between clones and comparators for any of the characteristics evaluated. The small differences in backfat thickness and marbling are likely due to the lighter weight of clones vs. comparators at slaughter.
 
v. Conclusions Regarding the Food Safety of Swine Clones

The conclusions regarding food consumption risks from swine clones are drawn largely from the animal health information presented by Walker et al. (2002), Archer et al. (2003b), and the Viagen dataset, supported by less detailed discussions of animal health in the other studies reviewed. These results indicate that there are no apparent anomalies present that would have a direct impact on the safety of food products derived from swine clones. The measurements taken at 27 weeks of age are approximately the age at which pigs are sent to slaughter in the US, and thus provide an appropriate age cohort for the evaluation of food safety. The identified abnormalities in the Archer et al. (2003a) (parakeratosis) and Viagen dataset (lung adhesion) are abnormalities normally seen in case noted does not pose a food consumption risk, as the affected skin from the carcass would be condemned at the slaughterhouse, and would not enter the food supply. The apparently normal status of the clinical measurements indicates that the clones in this study possess the same physiological functions and behaviors as their conventional counterparts, and thus are not likely to pose a greater food consumption risk than conventional swine.

c. Sheep Clones

i. Peer-reviewed Publications

As sheep were the first mammal to be cloned by SCNT, the relative paucity of papers on the developmental success of sheep clones is somewhat surprising. The seminal paper in the history of animal cloning is that prepared by Wilmut et al. (1997) in which they describe the generation of “Dolly,” the first mammal to be born (July 5, 1996) as the result of SCNT. (Gene, a bull clone was being gestated at the same time, but due to differences in the length of pregnancy between cattle and sheep, Dolly was born first). Dolly was derived from the mammary epithelium of a 6-year-old Finn Dorset ewe. The trial from which Dolly was derived included cells from two other sources besides the mammary epithelium, and included fetal fibroblast cells from a 26-day-old Black Welsh fetus, and cells derived from a nine-day-old Poll Dorset embryo.

Table VI-10 summarizes the outcomes Wilmut et al. 1997 paper. Pregnancy rate, as measured by detectable pregnancy at days 50-60 post transfer, ranged between ~ 8 percent to as high as ~50 percent. A total of 62 percent of the implanted fetuses were lost. Wilmut et al. (1997) reported that at approximately day 110 of the pregnancies, four dead fetuses derived from the embryo cell lines were detected. Their surrogate dams were euthanized, and post-mortem examination of the fetuses revealed two cases of abnormal liver development, but no other abnormalities or evidence of infection. A total of eight live lambs were born. One lamb, derived from fetal fibroblasts, died within a few minutes of birth. No abnormalities were noted at the post-mortem. Wilmut et al. cite the mortality rate of 12.5 percent (1 of 8) as similar to that observed in a large study of commercial sheep breeding, where 8 percent of the lambs died within 24 hours of birth. The birth weights of all of these sheep were within the range of single lambs born to the surrogate Blackface dams used at the Roslin farm (up to 6.6 kg), and were reported to be appropriate to the birth weights of the donor breeds.

The following year, Shiels et al. (1999) compared the telomere lengths of Dolly and one of each of the sheep derived from the different cell sources described in Table VI-10, with age-matched control sheep, donor mammary gland tissue, and donor cells. As expected, the mean size of telomere fragments in control animals decreased with increasing age. Mean telomere sizes were smaller in all three sheep clones than in age-matched controls. Dolly’s mean telomere size in particular, was smaller than other one-year-old age-matched sheep, and more consistent with the telomere fragment sizes derived from a 6-year-old sheep (the age of the animal from which the donor cells were derived). These observations led to speculation that clones would reflect the age of the donor cell, rather than effectively “resetting the biological clock” to their chronological age.

Dolly’s health was scrupulously observed over the course of her life. She developed arthritis at an early age, and was reported to have been overweight. Dolly was euthanized in early 2003 at approximately six and one half years of age having contracted a virulent form of lung disease that was endemic at the facility where she had been housed. It is not clear whether any of the abnormalities that were observed with Dolly were the result of SCNT, the conditions under which she was housed, or some combination of the two.

Most of the other papers in the literature refer to sheep clones generated from transgenic somatic cells to propagate animals with pharmaceutical potential, and data in those papers deal with expression of transgenes, molecular mechanisms that may be involved with fetal overgrowth syndromes (Young et al. 2001), or techniques to increase survival of nuclear transfer (Papadopoulos et al. 2002; Ptak et al. 2002) or other in vitro produced embryos. McCreath et al. (2000) reported on the post-mortem examination of transgenic lambs that died in utero or in the perinatal phase of development. These animals revealed a range of abnormalities including a high incidence of kidney defects, liver and brain pathology. This research group did not discuss the health of the transgenic lambs that survived.

Recently, Rhind et al. (2003) published a commentary on pathology findings from both transgenic (n=5) and non-transgenic (n=3) sheep clones that were not viable after birth. (The transgenes were intended to be targeted deletions of the ?-1,3 galactosyl transferase or prion protein genes). Of the eight animals evaluated, seven were euthanized at birth or shortly thereafter, the eighth survived but was euthanized after 14 days. The authors concluded that many of the defects (e.g., hepatobiliary changes, kidney structure changes, and pulmonary hyptertension) may not be contained within the “large offspring syndrome” (LOS) classification that may be common to other animal clones. Pulmonary hypertension has been observed in transgenic cattle clones (Hill et al. 1999), and in swine clones derived from transgenic “knock-out” piglets (Lai et al. 2002), suggesting that this syndrome may be common to a many species of animal clones, and that a common defect may be responsible. The authors call for additional research into the developmental mechanisms that may be responsible for the common defects in clones, although it is important to note that most of the animals with defects were derived from transgenic donor cells. The relevance of these observations to food consumption risks are limited, as clones that have died would not be used for food consumption purposes.

ii. Conclusions Regarding the Food Safety of Sheep Clones

Very few conclusions can be drawn about the health of sheep clones, due to the small database available for evaluation, as despite Dolly’s high public visibility, there are very few other reports of non-transgenic sheep clones. In the absence of specific information regarding the health of sheep clones, the only inferences that can be made would be drawn from interspecies extrapolation from other ruminant clones, i.e., cattle and goats.

d. Goat Clones

Relative to cattle, the database on goat clones is relatively small, but quite rich for its size. Much of the work that has been reported on non-transgenic goat cloning arises from data collected in an attempt to perfect systems by which SCNT can be harnessed to develop transgenic goats for commercial applications, and are effectively limited to publications from one group.

i. Perinatal Development and Function in Goat Clones (Developmental Node 2)

In 2002, Keefer et al. (2002) published a report on the birth of nine goat clones derived from two lines of adult granulosa cells and one line of fetal fibroblasts. Ninety-one female granulosa cell-derived embryo clones were transferred into eight surrogate dams. Four of those dams became pregnant, as confirmed by ultrasound on gestational day 30 and 60. All of these pregnancies went to term, and seven clones were born. Table VI-11 summarizes the outcomes of Keefer et al. 2002. One of the recipients delivered a single kid; the remaining three surrogates gestating granulosa cell-derived clones delivered twins. One of the female twins died at birth, but appeared to be normal.

In addition, 54 male fetal fibroblast-derived embryo clones were implanted into six surrogate dams. Only one of those dams had a confirmed pregnancy and delivered two male kids, one of which died during delivery. The authors state that this kid also appeared normal.

The birth weights of all of the kids were cited as being within the normal range for this breed (Nigerian Dwarf) at an average of 1.7 ± 0.13 kg versus 1.3 ± 0.06 kg for females resulting from natural breeding. The authors reported that the placentae of these kids appeared normal, and had cotyledon numbers that were comparable to those from the placentae of naturally bred Nigerian Dwarf goats. Suckling response was delayed in half of the granulosa cell-derived kids, and in the fibroblast-derived kid. These animals were fed colostrum by intubation, and good suckling was reported to occur by Day 2. The clones were otherwise reported as healthy, and having no apparent abnormalities.

ii. Juvenile Development in Goat Clones (Developmental Node 3)

In the Keefer et al. (2001a) study, the authors reported that blood profiles of the clones were monitored for one year, and showed no anomalous results. No data addressing this statement were presented in the Keefer et al. (2002) publication. There is, however, an abstract that was published in 2001 (Keefer et al. 2001b) in which some blood parameters are provided (Table VI 12: Selected Laboratory Parameters for Goat Clones). In this brief account, only a few measurements were reported; the duration of monitoring was six months. Alkaline phosphatase levels show the appropriate age-related changes to be expected for rapidly growing infants and very young animals, dropping to lower levels as the animals aged. Given the adult range for alkaline phosphatase in Nigerian Dwarf goats is reported as 16-33 U/L, these animals likely were still growing at 6 months of age.

Behboodi et al. (2005) compared hematology and blood clinical chemistry of four transgenic goat clones with four age-matched comparators and a published range for goat blood values (Pugh 2002). Hematology values were similar between clones and comparators, and all hematology values fell within the published range (Pugh 2002). For clinical chemistry, 18/24 values were not significantly different between clones and their age-matched comparators. Of the 19 clinical chemistry values for which published ranges were available, 18 of the values for clones and comparators fell within the published range. The one value out of the published range, creatine kinase (244.6 vs. 204.4 IU/L for clones and comparators), was not different between clones and comparators.

iii. Reproductive Development and Function in Goat Clones (Developmental Node 4)

One paper compares the sexual maturation and fertility of male Nigerian Dwarf clones to conventional bucks (Gauthier et al. 2001). Three clones (Stewart, Clint, and Danny) and four conventional animals that served as controls (Blue, Star, Banzai, and Ed) were trained to serve an artificial vagina beginning at the age of one month. Average age at first semen collection for both clones and controls was approximately 20 weeks, although volumes were small at the initial collection (<0.1 ml). Subsequent collections showed increased volume and increased sperm count (see Table VI-13 for a summary of the reproductive function in goat clones, comparators, and clone progeny).

Semen collected from two goat clones, Clint and Danny, at seven months was used to impregnate six Nigerian Dwarf does (three does for each buck). Although not explicitly stated, the implication is that the does were not clones. Five of the six does became pregnant. Two does impregnated by Clint gave birth vaginally to two sets of twins. Two does impregnated by Danny gave birth to singletons, and one doe gave birth to triplets. Nine kids were produced, and all appeared to be normal and healthy. Average birth weights for the male and female clone progeny were 1.7 ± 0.2 kg and 1.66 ± 0.1 kg, respectively, which do not differ significantly from average birth weights for conventional animals of this breed (1.7 ± 0.07 kg (n = 41) for males and 1.3 ± 0.31 kg (n = 79) for females). Semen was first collected from one of the progeny males at 28.4 weeks (Table VI-13).

The authors concluded that male Nigerian Dwarf goat clones developed sexual maturity similarly to their conventional counterparts. Further, these goat clones are fertile, and their progeny appear to be fertile as well.

In their study of goat clones generated from transgenic fibroblasts, Reggio et al. (2001) were able to produce a total of five healthy kids. Twenty-three surrogate dams were each impregnated with an average of eight embryo clones. Five of the dams that were detected as pregnant at day 30 completed their pregnancies, and gave birth naturally, providing a 100 percent success rate based on detectable pregnancy. All of the kids appeared healthy and vigorous. Birth weights averaged 3.8 kg (normal for the Toggenberg breed that served as the donor cell), and weaning weights were also within normal range (19.1-24.5 kg) for the breed. Each of the kids exhibited estrus, and has been bred to a buck. No reports of progeny were provided. Although this study is based on transgenic clones, it reiterates the high success rate that is experienced by researchers producing goat clones.

iv. Post-Pubertal Maturation in Goat Clones (Developmental Node 5)

CVM was not able to identify any published reports of measures of post-pubertal non-reproductive maturation in goat clones. Further, in the course of several presentations at scientific meetings, the Center learned that that the cohort of clones studied in Keefer et al. 2001a and 2002 has been terminated for business reasons.

v. Summary Statement on Health Status of Goat Clones

Based on these data, goat clones appear to have the least difficulty of any of the livestock species with respect to the SCNT process. Successful pregnancy outcome (when confirmed by ultrasound detection) is very high, and clones appear to be born at birth weights within the appropriate breed- and species- range. Suckling response was weak in some of the goat clones immediately after birth, but they appear to have recovered within one day. Available information on physiological parameters indicates that these animals appear to be normal. Data on reproductive function in these animals indicates that they enter puberty at the normal age range, produce viable semen, and normal, live offspring. The minimal reporting on one progeny animal also indicates that progeny are fertile.

vi. Conclusions for Food Consumption Risks from Goat Clones

Based on the data reviewed, there do not appear to be any anomalies present in the goat clones that would have a direct impact on the safety of food products derived these animals. Goats appear to be relatively “cloning friendly” with a high degree of successful live births following confirmation of pregnancy. All reports of health of the goat clones seem to indicate that they are normal and healthy. The available data on the physiological parameters of goat clones indicate that these animals respond as their conventional counterparts to internal signals for growth. The apparently normal status of the clinical measurements indicates that the clones in this study possess the same physiological functions and behaviors as their conventional counterparts. Further, unlike the other livestock clones, data on the reproductive behavior of male goat clones indicate that reproductive function is normal. Finally, although cursory in mention, it appears that male progeny of clone bucks also reach puberty at the appropriate time. Thus, although the number of animals that has been evaluated is not as large as in the case of bovine clones, goat clones appear to be healthy, and do not appear to be materially different from conventional goats.

3. Compositional Analysis Method

a. Overview

The operating hypothesis of the second prong complement to the Critical Biological Systems Approach is that if food products from healthy animal clones and their progeny meet the local, state, and federal regulatory requirements set forth for those products (e.g., Pasteurized Milk Ordinance,  USDA inspection criteria, absence of drug residues), and are not materially different from products from conventionally bred animals, then they would pose no more food consumption risk(s) than corresponding products derived from conventional animals.

Information on the composition of meat or milk from animal clones has been limited for several reasons. Few of the cattle clones are old enough to have been bred, given birth, and begun lactating. In addition, there is uncertainty regarding the kinds of analyses that could or should be performed in order to determine whether milk from animal clones is materially different from milk from non-clone animals. The issues associated with the compositional analysis of meat are similar, but have additional practical and economic components. During the course of preparing this Risk Assessment, CVM has contacted several food testing laboratories to inquire about the minimum sample size that would be required in order to perform a compositional analysis of meat. The Center’s hope was that systems were sufficiently miniaturized to allow analysis of “punch biopsies” of a shoulder or rump, but were informed that the minimum sample size would require sacrificing an animal. Nonetheless, there are now several studies that have evaluated the composition of the milk and meat of both cattle and swine clones, and one large study that has evaluated the composition of the meat of the progeny of swine clones.

b. Nutritional Risk

The primary concern for milk and meat from animal clones is that inappropriate reprogramming of the nucleus of donor cells does not result in epigenetic changes creating subtle hazards that may pose food consumption risks (Chapter III). Because, as previously discussed, there is no a priori reason to expect that SCNT will introduce any new, potentially toxic substances into the milk or meat of otherwise healthy animals, the remaining food safety concerns addressed whether subtle changes have occurred that would alter the presence of important nutrients. The most likely dietary risk would then be the absence or significant decrease in levels of vitamins and minerals whose daily requirements are in large part met by milk or meat.

The overall strategy we used to determine which milk or meat components could characterize their respective nutritional “footprints” involved selecting certain key nutrients and compositional parameters, while at the same time allowing sufficient flexibility in the non-essential components that vary with the genetic make-up and husbandry of the production animal. Finally, evaluation of the levels of the results of complex biochemical pathways in clones (e.g., saturated fats, vitamins) can further ensure that the clone is functioning appropriately, and thus indirectly support the hypothesis that the clones are appropriately reprogrammed and not materially different from their conventionally bred counterparts.

In order to identify the nutrients in milk or meat whose alterations would most likely affect the overall diet, even if all of the dairy and meat products from conventional animals in the daily diet were replaced by counterparts produced by clones, we first determined which nutrients made a “major” or “moderate” contribution to the total daily diet of milk or meat consumers. For the purposes of this Risk Assessment, a nutrient in meat or milk was considered a major dietary source if it provides 50 percent or more of its recommended dietary allowance (RDA) in that food.  Likewise, a nutrient in a food providing 10 to 50 percent of its RDA in that food is considered a moderate dietary source. For example, a single eight ounce serving of whole milk provides milk drinkers with between 10 to 50 percent of the RDA of vitamin B₂, riboflavin (B₂), pantothenic acid (B₅), calcium, phosphorous and selenium. Another example, a single serving of three ounces of roasted eye of round beef provides a meat consumer with a moderate source of zinc, niacin, vitamin B₆, phosphorus, iron, and riboflavin, and a major source of vitamin B₁₂ and selenium.

In order to determine typical meat and milk consumption in the adult US population, we consulted the one-day food survey conducted by the National Health and Nutrition Examination Survey (NHANES) of 2000-2001. According to NHANES, the mean daily consumption of milk among adult milk drinkers was 11.5 ounces. At this level of consumption, milk becomes a major source of vitamin B₁₂), and a moderate source of thiamin (B₁), zinc, and potassium, in addition to the nutrients previously listed as provided in moderate amounts. The same survey showed that the 90th percentile consumption of milk by users was 24.1 ounces per day, making milk a major source of calcium, phosphorus, riboflavin, and pantothenic acid, and adding magnesium and vitamin B6 to the list of nutrients provided in moderate amounts. Among subjects who consumed meat, the mean intake of meat was 4.2 ounces or 120.2 grams. Among the 90th percentile of meat eaters, consumption was 8.4 ounces or 239.4 grams. In order to determine whether evaluating the mean or 90th percentile consumption of milk and meat, changed the actual number of nutrients designated as moderate or major, we found that there were no nutrients were added or deleted, although with increased consumption rates, some of the nutrients changed from moderate to major contributors to the diet.

Proteins are of dietary importance because once they are digested, they provide the body with amino acids. In particular, some amino acids are of dietary concern because of the inability of mammals to synthesize them de novo in sufficient quantity to meet the body’s needs. For this reason they are designated as “essential.” Therefore, for purposes of assessing nutritional risk from food products from animal clones, the nature of the protein in its initial food matrix (e.g., casein or actin) is less important than whether it contains the same level of essential amino acids as its counterpart in foods derived from conventional animals. Finally, certain fatty acids such as linolenic (18:3) and linoleic (18:2) acid are essential components of the diets of mammals (including humans) and have also been selected as “key nutrients.”

Table VI-14 summarizes the analytes that we believe could be used to assess the composition of milk and meat from clones and comparators to demonstrate that there are no material difference between the two groups of animals with respect to key nutrients and overall nutritional characteristics. Included are key essential vitamins, minerals, and fatty acids. Other less essential constituents (e.g., vitamin A in milk is often supplemented, iron is not a key nutrient in milk) are also included to illustrate that checking on the levels of non-essential nutrients can also provide a useful tool to demonstrate the similarity of milk and meat from clones and their contemporary comparators.

i. Milk

For the purposes of this Risk Assessment, CVM uses the term “milk” to mean the “lacteal secretion, practically free from colostrum, obtained by the complete milking of one or more healthy cows,” and that “milk that is in final package form for beverage use shall have been pasteurized or ultra-pasteurized, and shall contain not less than 8 1/4 percent milk solids not fat and not less than 3 1/4 percent milkfat” (21 CFR 131.110(a)).

The Grade “A” Pasteurized Milk Ordinance (a model ordinance for adoption by states, counties and municipalities to regulate the production, collection, processing, sale and distribution of milk and certain milk products) echoes this definition, replacing “cows” with the term “hooved animals.” Therefore, in this Risk Assessment, unless otherwise specified, the term “milk” will refer to the lacteal secretions of cows, goats, or sheep. Although most of the discussion in this Risk Assessment refers to cow’s milk, similar arguments may be applied to the milk of goats or sheep. Other hooved animals whose milk is covered by the PMO include water buffalos, although they are not covered by this Risk Assessment.

The biological role of the milk of any mammal is to provide nutrition to its own newborn and young. In addition to mother’s milk, humans consume the milk of a few other species, principally from cows. Milk and milk products provide a considerable portion of the nutrition of other age groups, including growing children and adolescents, pregnant and lactating women, and the elderly. In 2001, per capita American consumption of milk among all age groups was approximately 23 gallons of fluid milk, 30 pounds of cheese, and 27 pounds of frozen dairy products (USDA ERS 2003).¹⁷ In particular, bovine milk and milk products (excluding butter) provided approximately nine percent of the energy, 19 percent of the proteins, 12 percent of the fats, and 4.5 percent of the carbohydrates consumed by milk drinkers in the US in 2001 ¹⁸. Ensuring that these dietary levels do not alter significantly is a key component of evaluating the potential nutritional risk from the milk of animal clones.

The degree to which individuals may experience risk from the consumption of milk appears to be a function of individual susceptibility, rather than the intrinsic toxicity of milk. For example, certain individuals suffer from Cow’s Milk Protein Allergy, which has an incidence of 2-6 percent among young infants (Exl and Fritsché 2001). Cow’s Milk Protein Allergy usually presents during the first year of life, and generally resolves by school age (Bernstein et al. 2003). Lactose intolerance is another milk-related condition found in adults and children (to a lesser degree) that is also a function of individual physiology (i.e., decreased expression of the enzyme lactase), particularly among certain ethnic groups. Excess consumption of saturated fats, including those from dairy products, can lead to atherosclerosis and its consequent morbidities; again, these harms are a function of individual behavior and susceptibility and not an intrinsic hazard of milk itself.

State regulatory agencies have managed the risk(s) posed by milk by adopting the PMO. It was first developed (1924) by what was then known as the Public Health Service, a precursor to today’s U.S. Food and Drug Administration. Now known as the Grade “A” Pasteurized Milk Ordinance, the PMO is revised biennially (most recently in 2003) by the Center for Food Safety and Applied Nutrition and other centers of the FDA, with input from industry and state regulatory agencies.

Table VI-15 lists the PMO requirements for Grade A milk (as adopted by state and local governments). As milk from dairy clones would be subject to the same requirements as that from conventional dairy cows, potential risks associated with subtle changes in immune function that might result in increased rates of mastitis, for example, would be controlled by the somatic cell and bacterial load requirements of the PMO. Likewise, even if clones suffered more bacterial infections, and required additional treatment with antibiotics, existing requirements restrict the presence of antibiotic residues in Grade A milk, thereby ensuring that milk from clones would pose no more bacteriological or drug residue risk than milk from non-clone cows.

 Is it possible to be reasonably certain that milk from animal clones and their progeny is indistinguishable from that now available in commerce? The complexity of milk itself is one of the primary difficulties in determining whether residual non-PMO managed hazards exist in the milk of animal clones. Milk from cows, sheep, and goats are mixtures that are estimated to be composed of more than 100,000 molecules (Jenness 1988), whose presence and proportion is a function of both the genetics of the animal and its environment. Not every component in milk has been identified and characterized; thus determining whether animal clones are producing a hazardous substance in their milk although theoretically possible, is highly impractical.

As for new components or changes in currently present but unknown and uncharacterized components of milk, it is unlikely that the cloning process would trigger expression of a novel substance that would not have independently arisen through random mutations in cow populations. In addition, it seems unlikely that a reprogramming error would lead to expression of an excess of a metabolically active protein with no adverse effects on the producing animal itself. This is especially true if the many nutrients that are monitored by the comparison scheme in Table VI-14 are within the ranges of contemporary comparators, and the physiological and biochemical parameters monitored in assessments of animal health are also within the ranges exhibited by contemporary comparators.

All milk is subject to Nutrition Labeling Requirements promulgated by FDA’s Center for Food Safety and Applied Nutrition under 21 CFR 101.9. These requirements provide a good starting point for milk characteristics that could be used as a basis of comparison. Additionally, the USDA Nutrient Database for Standard Reference (http://www.ars.usda.gov/ba/bhnrc/ndl) compiles data from a range of scientific, technical, food industry, and government agency sources to arrive at “composite” values of key nutrients in milk from cows, sheep, and goats. Table VI-14 provides a compilation of the key constituents and nutrients of milk from FDA’s Nutritional Labeling Requirements and USDA’s Nutrient Database.

If milk from clones and conventional animals does not materially differ in these constituents, it is unlikely that individuals consuming milk from animal clones will face increased risk(s) relative to individuals consuming milk from conventionally bred animals.

ii. Meat

For purposes of this Risk Assessment, CVM uses the term “meat” to mean “(1) The part of the muscle of any cattle, sheep, swine, or goats, which is skeletal or...tongue,...diaphragm,...heart, or...esophagus, with or without the accompanying and overlying fat, and the portions of bone, skin, sinew, nerve, and blood vessels which normally accompany the muscle tissue... It does not include the muscle found in the lips, snout, or ears....” and “(2) The product derived from the mechanical separation of the skeletal muscle tissue from the bones of livestock using the advances in mechanical meat/bone separation machinery and meat recovery systems that do not crush, grind, or pulverize bones, and from which the bones emerge comparable to those resulting from hand-deboning....” (9 CFR 301.2)

Meat comprises a large proportion of the average American’s diet, for both cultural and economic reasons (meat is relatively inexpensive in the US). In 2001, total annual per capita consumption of beef, veal, pork, lamb and mutton on a retail weight basis was approximately 122 pounds, and is estimated to have been about the same for 2002. The species-specific breakdown is approximately 69 pounds from beef and veal, 52 pounds from pork, and a little over a pound for lamb and mutton (USDA-NASS Statistical Highlights of US Agriculture 2001 and 2002). Goat consumption tends to be centered in various ethnic groups, but when averaged over the US population is about a half a pound per capita per year (USDA).

Meats provide a substantial portion of the nutrition in a non-vegetarian American diet. For example, beef provides approximately 50 percent of the total protein in a 2,000 calorie American diet, as well as approximately a third of the daily requirement of zinc and vitamin B12. It provides about 20 percent of the daily requirement for selenium, phosphorus, and niacin, and lesser although substantial amounts (i.e., 10-15 percent) of daily requirements for vitamins B6, riboflavin, thiamin, and iron (USDA Nutrient Database for Standard Reference Release 15, 2002).

Similar to milk, consumption of meat for millennia has taught that there are no significant intrinsic toxicants in meat from cattle, swine, sheep, or goats. Examples of meat allergies are rather rare, although they do exist. Cases of human immune-mediated allergies to the cattle proteins bovine serum albumin and bovine gamma globulin have been reported (Wuthrich et al. 1995; Han et al. 2000; Fiocchi et al. 2000; Tanabe et al. 2002). Humans allergic to cat serum albumin may also exhibit cross-reactivity to swine serum albumin (Hilger et al. 1997), in a phenomenon referred to as the pork-cat syndrome (Drouet et al. 1994). Children exhibiting positive skin prick test to bovine serum albumin may also cross react with sheep serum albumin (Fiocchi et al. 2000). As is the case for all immune-mediated allergic response, the individual’s susceptibility is in large part the driver for the response, as allergies are examples of the dysfunction of the immune response.

Just as for milk, there are no chemical composition schemes that “define” beef, pork, mutton, or goat meat. Due to the physiological function of muscles, and their need for rapid perfusion and oxygenation, meat also reflects the materials circulating the blood of the animal prior to slaughter. Myoglobin, the major storage protein for oxygen, is found in high concentration in muscle tissues. Unknown numbers of other large and small molecules are also found in meat, whose origins can be environmental, dietary, or endogenous. Each of these contributes to the complex profile that is responsible for the distinctive tastes and smells of meats.

The muscle tissue that makes up meat is composed of two major protein types: myofibrillar proteins, actin and myosin, which make up the fibers in muscle bundles, and connective tissue, which primarily consists of collagen and elastin. Collagen is the major component of gelatin, which results from the melting of collagen in the presence of hot water. Elastin is not greatly affected by cooking.

Tenderness, one of the primary considerations in carcass merit, is affected by the interplay of the myofibrillar and connective tissue proteins, and changes over the age of the animal and the amount of time since slaughter. The more connective tissue there is in a piece of meat, the tougher it tends to be; cooking, by solubilizing the collagen, decreases meat toughness. Collagen levels and structure tend to change in animals as they age, with the amount in young animals considerably lower than in older animals. With age, collagen undergoes more cross-linking, rendering it more insoluble and less likely to dissolve during the cooking process. The amount and distribution of fat in a muscle also influences tenderness. Marbling, or the presence of fatty deposits within muscles, also affects tenderness by functioning as a “lubricant” on the teeth or in the mouth, and by leaving “pockets” between muscle bundles as it melts during cooking. Changes in the amount of collagen or fat in the animal may affect meat quality with respect to tenderness or other qualities, but these would not pose nutritional or other food consumption risks. Further, it is likely that beef cattle clones will have changes in the amount or nature of marbling relative to average conventional beef cattle, as breeders will select animals as nuclear donors that have carcass qualities producing more uniformly tender and tasty meat. Similar selection procedures are being applied to animals used in conventional animal breeding programs, so the effect of cloning would be to speed the rate at which these desirable traits are introduced into breeding and production herds.

When an animal is slaughtered, rigor mortis (muscle stiffening observed after death) causes stable cross-links to form in muscle fibers due to the free flow of calcium across the cell membranes. The carcass stiffens and lactic acid levels accumulate resulting in a decrease in pH. The net result is that muscle fibers contract, and the meat appears “tough.” As the meat ages, however, a set of enzymes called calpains (calcium activated proteases) break down some of the structural components of the muscle, relieving the contraction, and degrading the connective tissue proteins, also releasing the degree to which the muscles are held together. Calpains are thought to function in concert with their antagonistic regulators, calpastatins, such that if calpastatin levels are high, calpain activity will be inhibited and less post-mortem degradation will occur, resulting in tougher meat. Most processors age beef for a minimum of 14 days to allow sufficient time for the calpains to work. Changes in levels of either calpains or calpastatins may thus affect meat tenderness, but likely would not pose food consumption risks.
 
As is the case for milk, the question of the appropriate comparator for meats may be approached from two perspectives. In order to determine whether cloning results in potential food consumption hazards relative to close genetic relatives, comparisons could be made to animals that are matched as closely as possible by age, husbandry (including diet), and environment. The second approach compares meat samples from animal clones more broadly to the national herds by using composite data sources.

Unlike milk, however, meat consists of various cuts that although made up mostly of muscle, contain different minor tissues, and whose function may affect composition. For example, the muscle in loin cuts may differ in composition from the muscle used to make bacon (i.e., belly muscles). In order to provide the most useful data for purposes of determining similarity to conventionally bred animals, it would be useful to compare cuts from each species that have the following characteristics:

• High US consumption levels (e.g., loin, rib, shoulder roasts, pork bellies, lamb or mutton shoulder or leg), and 

• Cuts of different muscularity that may have different overall compositions (e.g., if one tissue is lean, another may be fatty).

USDA’s Nutrient Database (http://www.ars.usda.gov/ba/bhnrc/ndl) contains composite tables that provide chemical compositions of several cuts of beef, pork, and lamb meats. Goat meat composition is only available as a single source.

There are no full chemical characterizations for meats. Moreover, as the definition of meat actually contains several tissue types, and each varies according to the genetics, breed, species, and environment of the food animal, it is unlikely that “complete” characterizations will ever be developed. The USDA requires nutritional labeling on “mixed” pork and beef products, and allows the voluntary labeling of raw products (9 CFR 317.300). Included in the labeling are calories, calories from fat, total fat, saturated fat, cholesterol, sodium protein and iron. Because meats are declared not to be a significant source of total carbohydrate, dietary fiber, sugars, vitamins A and C, and calcium, USDA does not require labeling information on them.

c. Characterization of Milk from Cow Clones

i. Peer-reviewed Reports

Walsh et al. (2003) evaluated the milk produced from 15 dairy cow clones from five different donor cell lines and three different breeds. These animals were produced by Infigen, Inc., and they have been described by Forsberg et al. (2002), reviewed in the Critical Biological Systems Section earlier in this Chapter. Clones were bred by AI between 14 to 16 months of age; the paper does not specify whether all of the heifers were inseminated with semen from the same bull. Five different cell lines were used as donors to generate the clones, and the breeds represented by the cell lines included two Holsteins, and two cell lines derived from cows resulting from crossbreeding Jersey and Holstein cattle.
 
Comparator cows were housed at different farms from the clones, but were approximately age and lactation-stage matched. They consisted of five Holsteins living on one farm, and one Brown Swiss cow raised at a farm different from the clones or the comparator Holsteins. Because of the different rearing sites, clones and their comparators were fed different rations, and for the clones, the ration was changed during the course of the lactation. Each cow was lactating for at least 30 days prior to sample collection, and samples were collected at approximately two month intervals over the entire lactation cycle.

Cows were milked into individual buckets, the contents of the bucket mixed and distributed into various vessels appropriate to each analysis. Samples were coded at the collection site, although the coding was broken approximately half-way through the study for unspecified reasons. The milk components that were analyzed are found in Table VI-16.

Mastitis is an infection of the udder, and a common problem in dairy cattle, that characteristically causes an increase in the number of somatic cells (cells from the circulation) in the milk of affected cows. The somatic cell count of the milk from both clones and non-clones indicated that none of the milk being sampled came from cows with mastitis. This implies that the immune function of the clones was sufficient to ward off infection under the husbandry conditions that the cows experienced. (A similar lack of impact on somatic cell count was reported by Heyman et al (2004) for milk from 50 clone cows compared to milk from 68 contemporary non-clone controls). The pH of the milk from the clones was within the range of healthy cows (~6.5-6.8). Acid degree values, which indicate rancidity or off-flavors, were also within the normal range for fresh milk.

No significant differences were noted between clones and non-clones with respect to the concentrations of the individual milk proteins that were sampled. No significant differences (p> 0.05) were observed when the gross composition of milk from Holstein clones and Holstein non-clones was compared over the course of the entire lactation cycle (Table VI-17).

No significant differences were reported between milk from clones and sexually-reproduced cows with respect to the individual milk proteins that were assayed, although difference in the concentrations of as-casein, k-casein, and a-lactalbumin were noted over the course of the lactation.

For 12 of the 14 fatty acids analyzed, no significant differences were noted between clone and non-clone milk. Significant differences (p<0.05) were noted in the amount of palmitic acid (C16:0) and linolenic acid (C18:3) between clones and non-clones. The authors noted that the palmitic and linolenic acid levels for both clone and non-clone milk fall within published references for that substance, and speculated that difference between the levels in clones and non-clones could be attributed to diet. Differences were observed in the fatty acid profiles of the milks over the course of the lactation cycle. These were noted as being consistent with published accounts of lactation cycle differences, diet, and seasonality. The greatest variability was observed in the mineral content of milk from clones and non-clones, with significant differences noted for potassium, zinc, strontium, and phosphorus levels. The authors attribute these differences to the different diets that clones and non-clones were fed. (Clones and comparators were housed at different farms, and fed different rations.) The authors’ overall conclusion was that there were “no obvious differences between milk from clones and non-clones.”

In an abstract, Aoki et al. (2003) described the generation of two clones from cells derived from the colostrum of a Holstein cow, as well as providing summary comments regarding milk characteristics and milking behavior. According to the abstract, milk yield (measured in kilograms per week) was measured every four weeks over a 16 week period. They noted that significant differences were observed between milk yield at weeks 1, 9, 11, and 13, but in the other weeks, “they shared similar lactation curves.” Milk composition was apparently measured as milk fat, protein, lactose, solids-non-fat, and total solid percentages. The authors reported that there were “considerable resemblance[s]” between the milk of clones and non-clones. It should be noted, however, that the measurements in this study were made between clones in their first lactation, and comparators in their second lactation. There are often differences in milk yield and composition between successive lactations (Vasconcelos et al. 2004; Flis and Wattiaux 2005).

The laboratories at the University of Connecticut continued their surveillance of a set of Holstein clones (see CBSA section) by analyzing the composition of milk from clones (the composition of meat from Japanese Black clones is discussed in the Meat Composition section) (Tian et al. 2005). Ten dairy clones were produced through SCNT using skin fibroblast (n=4) or cumulus cells (n=6) of a 13 year old Holstein cow. Four of the surviving cumulus cell derived clones were compared with four age- and parity-matched comparator heifers. All animals were raised at the same facility from 2 months of age, with the same management and feeding. Both groups were bred by artificial insemination using semen from the same bull at 14-15 months of age.

Milk production was monitored starting immediately after calving through 305 days of lactation; milk samples were collected three times daily. Two milk samples were collected from each of three milkings on a given day of each week throughout the entire first lactation. One of the collected samples was used for the analysis of total protein, total fat, lactose, total solids, milk urea nitrogen, and somatic cell counts as routinely monitored by the Dairy Herd Improvement Association (DHIA) at a DHIA designated laboratory. Individual fatty acids that were measured included C4:0, 6:0, 8:0, 10:0, 12:0, 14:0, 14:1, 16:0, 16:1, 18:0, 18:1, 18:2, 18:3, and 20:0. The second collected sample was analyzed for protein profiles using denaturing SDS/PAGE (sodium dodecyl (lauryl) sulfate-polyacrylamide gel electrophoresis) stained with Comaisse blue. Relative quantities of each band were determined. Antibody concentrations (IgM, IgA, and IgG) were determined in colostrums from the first milking with a commercial assay.

The investigators report that there were no significant differences between the composition of milk from clone and age-matched, closely related comparators, or breed comparators. Clones and comparators also showed comparable lactation curves, with milk production increasing during the first month, and decreasing thereafter during the course of the lactation. The exception was one clone that birthed prematurely, and produced 30 percent less milk, as would be expected. Analysis of key milk proteins indicated that there were no differences among major or minor bands as analyzed by SDS/PAGE. The four major bands representing a-caseins, b-caseins, k-caseins, and b-lactoglobulins were consistent in all milk samples whether from clones or their comparators. Similarly, there was no difference between groups for minor protein bands. Antibody concentrations in colostrums were also similar between clone and non-clone cows, and reported to be in the typical range for colostrum antibody composition.

Yonai et al. (2005) (see previous discussions of these animals in the Developmental Nodes) presented milk composition data for six Holstein and four Jersey clone cows. Overall milk yield, fat, protein, and other solids not fat (SNF) were considered to be normal by the authors, with the observed inter-clone differences and differences from the donor animals attributed to diet and environmental conditions. The authors note that the heritability of milk production is approximately 30 percent and, considering the impact of environmental conditions on milk production, suggest that standardizing individual feeding conditions may be helpful for future comparisons.

ii. The Report of the Japanese Research Institute for Animal Science in Biochemistry and Toxicology

The Japanese Research Institute for Animal Science in Biochemistry and Toxicology provided a report entitled “Investigation on the Attributes of Cloned Bovine Products” published by the Japan Livestock Technology Association (Japan 2002). CVM was able to obtain a seven page English-language summary translation of the original 489 page Japanese report. Only the English-language summary is reviewed in this risk assessment.

The study investigated blood, milk, and meat constituents in blastomere nuclear transfer clones (BNT) and SCNT cattle clones. In addition, the results of rodent feeding studies conducted with edible products derived from the cattle clones are reported. The results for milk are discussed in this section; results for meat are discussed in the section Compositional Data on Meat from Clones. No information was provided on the production of the BNT or SCNT clone cattle in the English translation, and the comparator group was identified as “ordinary cattle.”
Milk constituents were compared between ordinary cattle, BNT clones, and SCNT clones. The results are reported as the mean of samples obtained three and six weeks after parturition and provided in Table VI-17. No biologically significant differences were observed between any of the groups for the parameters tested.

Milk from these cow clones was tested for allergenic potential. The ability to digest a protein is one index of potential allergenicity; a protein that is less able to be digested may be more likely to provide an allergenic response. The protein digestion rate of freeze-dried milk combined in feed consumed by rats is reported below for milk obtained from ordinary cattle, BNT clone cattle, and SCNT clone cattle. The authors report that there was no biological difference among the groups tested.

In a separate study, mice were sensitized by intraperitoneal injection to extracts of milk from clone and non-clone cows. Fourteen days later, the abdominal wall of the mice was surgically retracted and an allergic reaction induced by re-injection of the freeze-dried milk extract into the abdominal wall. Control mice did not receive the second injection of milk extract. Allergenic response was assessed based on vascular permeability as measured by the diameter of dye leakage from the site of injection. No statistically significant differences in allergenic activity were reported between groups. The data are presented in Table VI-21.

Based on these two studies, the authors conclude that there were no biologically or statistically significant differences in the allergenic potential of milk from ordinary cattle or BNT or SCNT clones.

In addition to the composition and allergenicity studies, the Japanese Research Institute for Animal Science in Biochemistry and Toxicology performed a 28-day rodent feeding study.²² Rats were fed diets containing freeze dried milk from clones and ordinary cattle at concentrations of 0, 2.5, 5, or 10 percent of the diet for 14 weeks. General signs, body weight, food consumption, urinalysis, sensory and reflex function, spontaneous movement frequency, general function, reproductive cycle, hematology at autopsy, blood chemistry, organ weights, pathology and histopathology were compared between groups. English-language summary tables were provided in the original Japanese-language report; the summary tables have been provided in Appendix G. No biologically significant differences were reported in rats fed milk from clones compared to rats fed milk from ordinary cattle. In addition, it is noted that 10 cattle fed clone milk powder at 2.5, 5, or 10 percent of the diet showed no significant differences in body weight increase, indicating that the milk did not contain anti-nutrients or other toxicants to cattle. The duration of exposure is not reported.

Finally, the potential for milk from BNT and SCNT clone cattle to cause clastogenic²³ (DNA breaking) events was assessed using an in vivo mouse micronucleus assay. Mice were fed milk from ordinary cattle, BNT clone cattle, or SCNT clone cattle at 0, 2.5, 5, or 10 percent of the diet for 14 days. In addition, a positive control group received a single intraperitoneal injection of 2 mg/kg mitomycin C, a known clastogen. The positive control group showed a statistically significant increase in the incidence of micronuclei appearance and polychromatic erythrocyte rate, and was considered a positive test. No milk-fed group, whether derived from ordinary or clone cattle, was positive in this assay.

The report concludes that there were no biologically significant differences in the component analysis or the results of feeding milk from ordinary cattle, BNT clones, and SCNT clones.

iii. Additional Data
 
In a preliminary report of the nutritional contribution of food from cattle clones, Tome et al. (2004) found no differences in the response of rats receiving meat and milk from clones or non-clone cattle. Rats were fed milk or meat from clones or controls for three weeks. Outcomes evaluated included food intake, body weight gain, body composition, and fasting insulin levels.  In addition, no differences were detected for IgG, IgA, and IgM subtypes for rats receiving clone or non-clone derived diets. Further, no specific anti-milk or meat protein IgE responses were detected in rat sera. The authors conclude that there are no major differences in the nutritional value of milk or meat derived from clone or non-clone animals, and suggest that this study be confirmed in longer term exposure studies.

Wells et al (2004) briefly reported on the composition of milk from six 2-year old Friesian cow clones in their first lactation. The milk composition was compared to that of the single donor cow in her third lactation. All animals were managed together as part of a single dairy herd. Variables measured are presented in Table VI-22.

The comparison was made based on a single milk sample take at mid lactation. Although one of the protein levels (bovine serum albumin (BSA (4.52 ± 0.10 vs. 4.48)) and two of the fatty acids in the clones (C18:2 (3.76 ± 0.06 vs. 3.00), C18:3 (1.18 ± 0.07 vs. 0.90) were found to be statistically different (p>0.05) from the donor cow’s milk, they were reported to be within normal limits for this breed of cow, and not considered by the authors to be biologically significant. The authors conclude that overall milk composition of the clones was what might be expected for healthy cows.

In a presentation at the January 2005 31st IETS Annual Meeting, the same group (Reproductive Technologies Group, AgResearch Ltd, New Zealand (Lee and Wells 2005) presented data on AgResearch’s experience with cattle cloning from 1997 onwards. Milk composition was evaluated in three clone cows each from three clonal families (a clonal family is derived from the same source animal). Milk from a total of six SCNT clones (selection not specified) was compared to milk from a single donor cow (also not specified) (Table VI-23). In this preliminary communication of data, the investigators noted statistically significant differences for only BSA (162 ± 6 vs. 105 mg/L), and two fatty acids C18:2 (3.76 ± 0.06 vs. 3.00 percent of total) and C18:3 (1.19 ± 0.03 vs. 0.90 percent of total). They concluded that the composition of milk from clones was normal.

In 2004 through 2005, milk vitamin composition was compared for 3 clone cows each from 3 clonal families (n=9) to control non-clone cows (Table VI-24). Subsequently, Wells published a summary of a slightly expanded version of these data (Wells 2005) in a scientific journal (Tables VI-25 and VI-26). No details were provided for the comparator animals. No differences were reported in selected vitamins in milk. Wells concluded that the composition from these clones was within the normal range for milk.

iv. Summary Statement on Composition of Milk from Clones

Based on the available data, milk from cow clones does not appear to differ significantly in composition from milk from non-clones. Small differences have been noted between clones and comparators, but given the different diets and husbandry conditions of these animals, it is difficult to determine with certainty whether the small changes seen in some components were a function of the diet, handling, or related to cloning. In summary, none of the small reported differences in any of the studies indicate any concern for food safety.

d. Characterization of Meat from Clones and Their Progeny

i. Cattle

Two linked reports on carcass merit²⁴ (e.g., dressing percentage, fat depth, rib-eye area, yield and quality grade) of cattle produced via BNT have been published (Diles et al. 1996a,b). Neither paper addresses food safety issues. Both papers evaluate the degree to which body measurements are heritable (Diles et al. 1996a), and the degree to which there is phenotypic variability among clones and closely related siblings. The studies conclude that animals derived from BNT provide good models for determining which traits have strong genetic correlations.

As discussed in the section on milk composition, the Japanese Research Institute for Animal Science in Biochemistry and Toxicology provided an unpublished bound report “Investigation on the Attributes of Cloned Bovine Products” by the Japan Livestock Technology Association (Japan 2002).²⁵ The results for meat are discussed in this section. Takahashi and Ito (2004) have published a summary of these data, including some information characterizing the clones and their comparators. SCNT and BNT clones were derived from Japanese Black cattle at the Para Prefectural Animal Research Center. comparator animals were selected as conventionally bred Japanese Black cattle. All animals used for compositional analysis were sacrificed between 27 and 28 months of age, after fattening. For the in vitro digestion test, samples were taken from a one-day old conventional calf and a four day old clone.

Meat constituents were compared between ordinary cattle, BNT clone cattle, and SCNT clone cattle. The results are reported as the mean of analytical samples obtained from 9 sites; shoulder, chuck loin, rib loin, loin end, brisket, round, silver side, rump, and tender loin, and are provided in Table VI-27:

No biologically significant differences were observed between any of the groups of cattle (ordinary cattle, BNT clone cattle, and SCNT clone cattle) for the parameters tested

Meat from clone cows was tested for allergenic potential by comparing protein digestion rates with artificial digestive juice and in a rat model, and by looking for an allergenic response following direct challenge in rats. The rates of digestion by artificial digestive juices (artificial gastric juice and artificial intestinal juice) were compared for freeze dried meat derived from ordinary cattle, BNT clone cattle, or SCNT clone cattle (Table VI-28). No information is provided in the translation regarding the artificial digestive material. The results are presented below as the rate of protein digestion.