How Developments in Cryobiology, Reproductive Technologies and Conservation Genomics Could Shape Gene Banking Strategies for (Farm) Animals

Semen

For many animal species, there are adequate protocols for collection and freezing of semen, and for its use in artificial insemination (AI), and for some (farm) animal species, there is an existing ‘infrastructure’ for the collection and use of semen in AI. Semen seems very well suited for supporting breeding schemes of small breeds, especially in those species (e.g. cattle, horse) where AI is practised anyway and the use of the frozen semen does not incur extra costs. Semen can also be used to reconstruct a lost breed, but it will take 4–7 generations of backcrossing to restore the ‘original’ genotype, depending on the desired purity of the restored breed. This is time-consuming and costly, and it means that a large number of breeding doses need to be stored in the gene bank. The required number of doses is smaller if sexed semen is conserved, so that in each generation, predominantly female offspring is obtained that can in turn be inseminated with the sexed cryopreserved semen. Perhaps, this may only be possible in situations in which the sexing technology is (commercially) available, that is, in the case of semen of regular AI bulls. Another disadvantage of semen is that mitochondrial genes are not conserved, as sperm mitochondria do not contribute to the offspring’s mitochondrial population.

Embryos

Embryos are also used in gene banks for the conservation of genetic diversity, as they allow fast recovery of a lost breed in a single generation. Consequently, also fewer embryos than semen doses need to be collected, frozen and stored to allow recovery of a lost breed. Boettcher et al. (2005) showed that combined storage of semen and embryos could be most (cost)effective. Yet, easy and cost-effective use of embryos in gene banks may be limited to these species for which collection and transfer techniques are available and operational. This is certainly the case for cattle (Van Wagtendonk de Leeuw et al. 1997; Youngs 2011) and to a lesser extent also for sheep (Youngs 2011). Live offspring or pregnancies have also been reported for other farm animal species, including pigs (Li et al. 2009), horse (Ulrich and Nowshari 2002; Galli et al. 2007), goat (Rodríguez-Dorta et al. 2007) and rabbit (Naik et al. 2005), but in most of these species, the techniques are not routine.

Oocytes

Similar to embryos, restoration of a lost breed or genotype by using cryopreserved oocytes (plus semen) would not require backcrossing. In recent years, both slow freeze and vitrification protocols render excellent results in humans resulting in many healthy babies born (Porcu and Venturoli 2006; Cobo and Diaz 2011). Less data are available in animal species (see, e.g., Critser et al. 1997) than in humans. Live-born young produced from cryopreserved oocytes have been reported in cattle (Abe et al. 2005), mouse (Stachecki et al. 2002) and horse (Maclellan et al. 2002). The collection of oocytes through ovum pickup (ovum aspiration) procedures can be relatively efficient and could be a way to harvest germplasm from animals that are not capable or are not available for producing offspring or embryos. Yet, for many species, techniques of in vitro maturation, cryopreservation and IVF or ICSI of oocytes are currently not available or inefficient.

Ovaries

Cryopreservation of ovaries could be another way to conserve the genotype of animals. In the human, ovary banking can be used to prevent loss of fertility during cancer treatment of women. Cryopreserved ovaries or parts of ovaries may be used as a source of oocytes, or oocytes may be harvested from heterotopically grafted ovaries to be subsequently used in an IVF procedure to produce embryos (Terazono et al. 2012). In animals, an alternative would be to graft cryopreserved ovarian tissue or whole ovaries orthotopically in a recipient animal to restore fertility in that animal. This animal can then be mated and will produce offspring carrying the ovary donor genotype (see some examples in mice and in birds explained later). A similar approach is transplantation of cryopreserved testicular tissue (Song and Silversides 2007).

Stem cells, primordial germ cells and somatic cells

In theory, spermatogonial stem cells can be harvested from a recipient animal and lead to sperm production in a donor male (Izadyar et al. 2002, 2003; Trefil et al. 2010). Furthermore, this is not likely to be an easy and cost-efficient way to conserve genetic diversity. As explained in the following section, in birds, cryopreservation of PGCs may be an efficient way to preserve genotypes, breeds or breeding lines.

Somatic cells can be used to produce embryos, as was first demonstrated by the creation of the sheep ‘Dolly’. Live offspring have been obtained from cloned embryos in a large number of mammalian species, including the farm animal species cattle, sheep, goats, buffalo, pigs, horse and mule (Campbell et al. 2007). Several companies offer cloning as a commercial service. Collection of suitable somatic cells is straightforward. Cryopreservation protocols for somatic cells are relatively simple and do not require controlled-rate freezing equipment. This means that establishing the collection is easy and cheap. The opposite is true, however, for using the material when needed. Utilization involves culturing the cells after thawing, reprogramming of the nuclei, collection of oocytes by ovum pickup or from slaughtered animals, culture and in vitro maturation of the oocytes, enucleation of the oocytes, transfer of the somatic nucleus to an enucleated oocyte, culture of the resulting embryos and finally, embryo transfer (ET) in recipients of the same or a related species. In this procedure, the donor mitochondrial genes are not conserved. Because of the low upfront costs, a gene bank of somatic cells was established as a backup for endangered livestock breeds in conditions with little infrastructure in Vietnam (Groeneveld et al. 2008).

Farm animal gene bank collections in the Netherlands

Apart from minor collections of embryos and oocytes, the Dutch gene bank of the Centre for Genetic Resources, the Netherlands (CGN) currently contains relatively large collections of semen (Table 1). CGN covers only farm animal species and species or breeds within species that may have a farm-related history. In addition to the rare native breeds, CGN also preserves regular updates (‘snapshots’) of semen from the breeding bulls and boars of the Dutch commercial cattle and pig breeding organizations. The gene bank has an EU-certified storage site for cattle semen, and semen stocks of different species and of different veterinary health status are stored separately. In addition, for security reasons, we have duplicate collections in two separate locations. CGN resources are used in the field to a limited extent to support breeding schemes of rare breeds (mainly cattle) in particular to keep inbreeding levels at an acceptable level. Perhaps, for those cattle breeds, a minimum of 100 breeding units per male are retained as ‘core collection’ for long-term conservation of allelic variation. The aim is further to maintain stocks of at least 25 males per breed in long-term storage for any future use, for example, reconstruction of a lost breed.

Semen

Semen is a versatile and often low cost option for gene banking of (farm) animal genetic diversity. Yet, for a number of species, sperm freezing methods give unsatisfactory results and further method development is required. In some other species, not the freezing per se, but rather the collection of semen may be the problem. In cattle, one obstacle is that no or not enough bulls of rare breeds are offered by AI stations and semen from bulls must be collected on farm, which is officially prohibited under EU veterinary regulations. In very small and inbred populations, there is a strong possibility that the semen quality of many or all of the remaining males is very low, and so cryobanking may be even more limited. In particular situations, collection of semen may be too costly or fails altogether. For instance in more or less feral animals, like the sheep of various heath breeds, the males may be too ‘wild’ to be trained for the collection of semen. We found that the ability to train rams for semen collection varied strongly between breeds. Therefore, CGN and other groups have tried the collection and use of epididymal spermatozoa for sheep (Woelders et al. 2004, 2005; Ehling et al. 2006). In sheep, the yield of spermatozoa from the caudae epididymidis is very high, well above 20 billion spermatozoa per animal in mature rams, allowing the production of more than 100 insemination doses (200 million sperm/dose) per animal. In addition, the freezability (post-thaw per cent motile and live sperm) and fertilizing ability of frozen-thawed epididymal spermatozoa from Veluwe rams (a Dutch heath sheep breed) appeared to be at least as good as that of ejaculated semen from rams of the same breed (Table 2). In the Netherlands, since 2004, epididymal spermatozoa have been used predominantly to produce the gene bank collections for Dutch native sheep breeds (Table 3). Also, in breeds where ejaculated semen can be collected, we found that collection of epididymal spermatozoa is most cost-efficient. The samples can be used without requiring special facilities or expertise, and no housing and training of the rams is required. We collaborate with the breed organizations and are notified when genetically suitable rams are being culled after having served for natural inseminations in the herd, and testes are collected and transported from the abbattoir to the institute, to be processed the same day. The total costs involved are very limited in comparison with the collection of ejaculates to the same number of doses (Woelders and Kaal 2008). We use a simple rapid method of obtaining the spermatozoa in a semi-quantitative way (Woelders et al. 2005; Woelders and Kaal 2008), which involves cutting the cauda (submersed in Tris egg yolk freezing medium) with a scalpel many times so it becomes more or less minced and squashed (see, e.g., Soler et al. 2003). This allows virtually all spermatozoa to escape from the convoluted epididymal duct. Alternative methods are float-up or retrograde flushing (e.g. Cary et al. 2004; Martinez-Pastor et al. 2006; Santiago-Moreno et al. 2009).

For similar practical reasons, epididymal spermatozoa are used for the conservation of genetic diversity of rare wildlife species, for example, the European bison (Kozdrowski et al. 2011), Red deer (Martinez-Pastor et al. 2006) and Ibex (Santiago-Moreno et al. 2009).

There are few publications regarding collection of epididymal spermatozoa in avian species (Clulow and Jones 1988; Kwon et al. 1995; Orlu and Egbunike 2010). The morphology and function of the epididymis in birds are quite different than in mammals. In quail, it was shown that most ‘extragonadal’ sperm are found in the vas deferens (Clulow and Jones 1988) rather than in the epididymal duct. In chicken, interesting sperm numbers can be retrieved from the vasa deferentia (2–4 billion sperm per male, C. Zuidberg, CGN, personal communication), and experiments are currently underway at CGN to investigate collection, freezability and post-thaw fertilizing ability of vas deferens sperm in birds in the interest of finding efficient ways of collection and cryopreserving Dutch native breeds of ducks, geese, chicken and pigeon.

Sheep AI

In sheep and goats, AI with fresh semen is possible and is being used in practice. Methods for freezing ram or buck semen are available and give a fair rate of success (Salamon and Maxwell 2000; Leboeuf et al. 2008). With frozen-thawed ram semen, good fertility results are obtained with laparoscopic insemination (surgical semen deposition in the uterine horns), but not with ‘normal’ cervical insemination. The reason seems to be that frozen-thawed spermatozoa have a shorter lifespan and probably a less efficient motility, and therefore, sperm migration through the cervix and uterus to reach the site of fertilization is impaired. Non-surgical intrauterine insemination is not easy with conventional techniques and instrumentation, because of the complex anatomy of the cervix. Laparoscopic insemination is prohibited in the Netherlands, and in any case, it is more complicated and time-consuming than conventional AI. Therefore, development of a non-surgical method of cervical passage would be desired for ewes. Similarly, intrauterine insemination could also be beneficial in goats.

The anatomy of the sheep cervix is highly complex, which makes it difficult to pass the cervix with an implement (Halbert et al. 1990a; Kershaw et al. 2005). Three approaches have been tried to reduce the difficulty of passing through the cervix. One is by using an implement to retract the cervix into the vagina to identify the cervical opening and align the funnel-shaped rings of the cervix to some degree (Halbert et al. 1990a,b). A second approach is the use of an agent to relax the cervix (e.g. PGE-2 or PGE analogues, oxytocin, interleukin-8, misoprostol, or hyaluronan (see several references in Wulster-Radcliffe et al. 2004; Perry et al. 2010), or the beta-adrenergic blocking agent carazolol (Gündüz et al. 2010). A third approach is designing appropriate transcervical AI and ET equipment to overcome the physical difficulties associated with the ovine cervix (Wulster-Radcliffe et al. 2004). In what is referred to as the Guelph system, restraint of the ewe and instrumentation were adapted to allow transcervical deposition of the semen in at least a fraction of all tested ewes (Halbert et al. 1990b), which resulted in higher pregnancy rates compared with vaginal semen deposition in the vagina or cervical os. In later studies, both improved transcervical AI implement (Wulster-Radcliffe et al. 2004) and administration of PGE2 (Candappa et al. 2009) or PGE1 analogue (Leethongdee et al. 2007) allowed 100% penetration of the cervix, but lambing data are either missing or not dramatically improved. In the case of goats, excellent kidding rates were reported after transcervical AI with frozen-thawed semen (Sohnrey and Holtz 2005). Furthermore, it appears that in sheep, further research is needed.

Horse

The problems and potential solutions associated with cryobanking of horse germplasm are reviewed in full detail by Smits et al. (2012) and will be discussed here only briefly. In the horse, recovery of a lost breed using semen is much more expensive than in other farm animal species (Gandini et al. 2007). This may be improved when using low dose (1–10 million sperm) insemination techniques, for example, by hysteroscopic insemination (Morris and Allen 2002). At the same time, however, the possibilities of using embryos or oocytes appear to be poor in the horse. Cryopreservation of horse embryos appears to be very difficult indeed (Tharasanit et al. 2005), with relatively few reported pregnancies (see references in, e.g., Choi et al. 2011) and only a few live births reported after transfer of cryopreserved embryos (Ulrich and Nowshari 2002; Galli et al. 2007). This is an area where horse reproductive specialists and cryobiologists should work together to improve the methods for cryopreservation of horse embryos taking into account the special morphological and other embryo characteristics seen in this species (see, e.g., recent advances in Choi et al. 2011; Scherzer et al. 2008).

Collection of horse oocytes is in itself a practical possibility, as oocytes can be collected, without knowing the stage of the estrous cycle, by transvaginal ultrasound-guided ovum pickup, allowing the collection of more than 70 oocytes per year per mare. But cryopreservation of horse oocytes is very difficult (Tharasanit et al. 2006a,b), with only one report of live offspring after transfer of embryos produced from cryopreserved oocytes (Maclellan et al. 2002).

Although the horse has the disadvantage in the field of embryos and oocytes compared with some of the other farm animal species, it seems to have an advantage with regard to the production of offspring from embryos generated from (cryopreserved) somatic cells by nuclear transfer, with many healthy live offspring and high efficiencies reported. Possibilities and successes are reviewed by Smits et al. 2012 and in the FAO Draft Guidelines for the Cryoconservation of Animal Genetic Resources (FAO 2011). In the Netherlands, CGN has considered cryobanking of somatic cells as a means to preserve the genotypes of unique, old mares for which other approaches to preserve their germplasm are currently not possible or impractical. The ethical acceptability of nuclear transfer ‘cloning’ is, however, debated, partly because of concerns of animal health and welfare, as cloning in many species is still associated with higher levels of embryonic loss and malformed young (Campbell et al. 2007; Oback 2008), but also because of other ethical considerations. One could still consider cryobanking of somatic cells in case no other viable route is available, on the premise that the material will not be used unless future methods regarding nuclear reprogramming have evolved to a stage that offspring can be obtained without compromising health or welfare or the genetic makeup in the first offspring and in subsequent generations. An overview of possible routes and areas for future research are shown (Fig. 1).

Laboratory animals

At the turn of the millennium, more than 2500 mutant mouse lines existed (European-Commission 2001), and since then, the number of genetically modified mouse lines has rapidly expanded, and ambitions are to produce a comprehensive resource of mouse mutants in which each of the approximately 20 000 genes in the mouse genome has been knocked out (NIH 2006). Institutions and core facilities offering mouse archiving/retrieving services all offer embryo and approximately half offer sperm cryopreservation as well. Few institutions provide ovary cryopreservation and recovery services. A feasible cryopreservation strategy should enable reconstitution of the original genotype, including the congenic background, in one or two generations. Embryo banking is an option, but in mutant mouse lines, poor response to induced superovulation and poor oocyte or embryo quality may be encountered. Both the mutation and the genetic background can impair embryo development (Glenister and Thornton 2000). Regardless, the costs involved in embryo banking are high (Glenister and Thornton 2000; European-Commission 2001; Landel 2005), several thousands of Euros for the cryopreservation of a single strain.

Semen banking may seem an obvious choice, as cryopreservation of mouse semen is easy and quick. If the semen can be used to fertilize oocytes of females of the same inbred background, only one generation breeding with the F1 offspring would be sufficient to recover the homozygous mutant against the original genetic background. Yet, the same problem here is that in the inbred strains, IVF results may be insufficient (Glenister and Thornton 2000).

The cryopreservation and use of ovaries for conserving a mutant genotype at first glance seems a less obvious option, as success may depend on fine surgical skills and requires good post-thaw survival of the ovaries that are able to revascularize, and elicit no graft rejection, etc. Yet, recent progress in cryopreservation and use of mouse ovaries has shown that this approach has great promise and may have several advantages over the banking of semen and embryos. Cryopreservation methods seem to have improved and are in fact made simpler by using vitrification protocols very similar to those used for embryos and oocytes (Huang et al. 2008, 2010). More importantly, it was shown that grafting is a relatively simple and successful procedure, which was demonstrated by using cryopreserved ovaries of a large series (22) of mutant strains with various genetic backgrounds (C57BL/6, FVB, BALB/c) to recover the genotype by orthotopic grafting of the vitrified ovaries in recipient mice (Huang et al. 2010). The method appears to be relatively easy and remarkably efficient. Collection of ovaries and subsequent vitrification of the ovaries are quick and simple, taking approximately 2 h per session. It requires no hormonal treatments of the females. The vitrified ovaries can be orthotopically transplanted in a simple and quick surgical procedure. Again, no preparation of the recipient females is required. Using recipients of the same genetic background as the donors prevents graft rejection. Within 10–14 days of transplantation, recipients can be bred to congenic wild-type males to obtain live offspring that is heterozygous for the mutation of interest, but homozygous for genetic background, with a success rate of >40% offspring of the donor genotype. For recovery of the mutant strain, only 5–10 ovary recipients and 1–2 males (both with proper wild-type isogenetic background) are needed to reconstitute the strain. The total cost of archiving and retrieving a mutant mouse line by ovary tissue freezing may be cheaper than with semen or embryo freezing, considering the number of animals required, and labour and equipment cost (Huang et al. 2008).

Birds

A reliable and efficient strategy for gene banking germplasm of birds would be important for gene banking of rare species of wild birds and of rare breeds of poultry, but would also be important as an alternative to holding and propagating live birds to maintain breeding lines in research institutes and poultry breeding organizations. In birds and other macrolecithal taxa, methods for cryopreservation of the ova or embryos are not available. Cryopreservation of semen is possible in a number of bird species, but may not be possible or efficient in specific cases. For the chicken, adequate methods for freezing semen are available with good fertility being reported in a number of studies using cryopreserved semen (Tselutin et al. 1999; Woelders et al. 2006). CGN recently created a semen gene bank collection of native Dutch chicken breeds and currently has a stock of semen from 20 breeds with 5–19 males per breed (Table 1). Furthermore, validation of the semen freezing method we use at CGN (Woelders et al. 2006) was performed only using semen from a commercial chicken line, and post-thaw fertility of frozen-thawed semen of the rare breeds may be lower than that of the commercial breed. Variation between breeds was in fact reported by Blesbois et al. (2007). In turkey, a recent study also indicated large differences in success of cryobanking of semen for various breeding lines, with no success (no hatched poults) in one of the lines studied (J. A. Long, H. Phillip, P. A. Purdy, C. A. Zuidberg, S. J. Hiemstra, S. G. Velleman, H. Woelders, unpublished data).

Potential alternative types of germplasm for gene banking in bird species are ovaries and PGCs. Recent studies suggested that cryobanking of ovaries is a viable and efficient option in birds. Live offspring carrying the donor genotype was obtained after transplantation of cryopreserved half-ovaries in quail (Liu et al. 2010). Both slow freeze and vitrification protocols were used in that study, but vitrification gave best result. Like in the mouse, the procedures used for collection, vitrification and grafting of the ovaries appear to be remarkably easy to perform and successful. Graft rejection was prevented by oral administration of an immunosuppressant. This method would be ideal if it can be complemented with cryopreserved semen from the same breed, so that the original breed or breeding line can be recovered in a single generation. In that respect, it is interesting that a similar approach of grafting cryopreserved testicular tissue from roosters rendered males that produced (donor testis derived) sperm sufficient for successful intramagmal insemination of hens (Song and Silversides 2007). If no semen of the breed or line is available, several generations of backcrossing would be needed and this would imply that larger stocks of ovaries need to be banked, as would be the case when only semen banking would be used to secure a breed or line.

Another possibility in birds is cryobanking of PGCs. Primordial germ cells can be simply collected from blood from day 3, Hamburger Hamilton (HH) stage 14–16 chicken embryos. (Naito et al. 1999; Van de Lavoir et al. 2006; Macdonald et al. 2010). The PGCs can be cultured and propagated as clonal cell lines and frozen. Frozen-thawed PGCs can then be used to produce chimaeric embryos by injection of the PGCs into the cardiac tract of HH stage 14–16 embryos, which then are transferred into phase III host shells and cultured to hatching in a specialized procedure (Perry 1988). Primordial germ cells only seem to give germline transmission when transferred to a same sex embryo. Male hatchlings can be raised to sexual maturity and may then carry the donor genotype in a fraction (2–16%; Macdonald et al. 2010) of their sperm cells. Likewise, female PGCs in female embryos can generate female chimaeras with the donor genotype in a fraction (0–12%; Van de Lavoir et al. 2006) of the produced ova. Currently, more research is needed to make the procedures more easy and consistent, but in principle, the cryobanking of PGCs would have the advantage of allowing recovery of a lost breed or breeding line without the need of backcrossing, provided that also female PGCs can be cloned and propagated (c.f. Van de Lavoir et al. 2006; Macdonald et al. 2010).