ILAR Journal Online, Volume 41(4) 2000: Cryobiology of Embryos, Germ Cells, and Ovaries

Online Issues

<< All Back-issues

<< This Issue's Table of Contents

ILAR Journal V41(4) 2000
Cryobiology of Embryos, Germ Cells, and Ovaries

Factors Affecting the Efficiency of Embryo Cryopreservation and Rederivation of Rat and Mouse Models
W. F. Rall, P. M. Schmidt, X. Lin, S. S. Brown, A. C. Ward, and C. T. Hansen

All authors are with the Veterinary Resources Program, Office of Research Services, National Institutes of Health (NIH), in Bethesda, Maryland. W.F. Rall, Ph.D., is a Physiologist and Manager of the Embryo Cryopreservation Program; P. M. Schmidt, M.S., is a Physiologist; X. Lin, D.D.S., is a Biologist; S. S. Brown is a Biological Laboratory Animal Technician; A. C. Ward, B.S., is a Biologist; and C. T. Hansen, Ph.D., is a Geneticist and Manager of the NIH Animal Genetic Resource.

Abstract

The efficiency of embryo banking for rat and mouse models of human disease and normal biological processes depends on the ease of obtaining embryos. Authors report on the effect of genotype on embryo production and rederivation. In an effort to establish banks of cryopreserved embryos, they provide two databases for comparing banking efficiency: one that contains the embryo collection results from approximately 11,000 rat embryo donors (111 models) and another that contains the embryo collection results from 4,023 mouse embryo donors (57 induced mutant models). The genotype of donor females affected the efficiency of embryo collection in two ways. First, the proportion of females yielding embryos varied markedly among genotypes (rats: 16-100%, mean=71%; mice: 24-95%, mean=65%). Second, the mean number of embryos recovered from females yielding embryos varied considerably (rats: 4-10.6, mean=7.8; mice 5.3-32.2, mean=13.7). Genotype also affected the efficiency of rederivation of banked rat and mouse embryos models by embryo transfer. For rats, thawed embryos (n=684) from 33 genotypes were transferred into 66 recipient females (pregnancy rate: 76%). The average rate of developing live newborns for individual rat genotypes was 30% with a range of 10 to 58%. For mice, thawed embryos (n=2,064) from 59 genotypes were transferred into 119 pseudopregnant females (pregnancy rate: 78%). The average rate of development of individual mouse genotypes was 33% with a range of 11 to 53%. This analysis demonstrates that genotype is an important consideration when planning embryo banking programs.

Keywords: cryopreservation; embryo banking; embryo transfer; inbred strains; mouse; rat

Introduction

The National Institutes of Health (NIH¹) established an embryo cryopreservation program in 1979 to assist the NIH Animal Genetic Resource (NIHAGR¹) in the development and management of a stock center of laboratory animal models of human disease and normal physiological processes. Currently, our embryo bank contains approximately 300,000 embryos from rat, mouse, and rabbit strains and stocks. Seventy rat, 92 mouse, and two rabbit strains and stocks exist solely as banked embryos. The NIHAGR consists of 350 strains and stocks of rats, mice, rabbits, and guinea pigs. Model types include inbred and congenic strains as well as outbred and mutant stocks. Embryos have been collected and banked from all NIHAGR rat and mouse strains and stocks.

Since 1996, embryo banking has expanded to include 83 mouse models and one rat model developed and maintained by investigators working at NIH Institutes on the Bethesda, Maryland, campus (NIH intramural investigators). Most of these models are induced mutant mice on standard or unique inbred or mixed genetic backgrounds produced using transgenic or site-directed mutagenesis technologies (knock out, knock in, Cre-loxP). The banking of cryopreserved preimplantation embryos assists the management of animal models by providing protection against loss from breeding failure, catastrophe, or genetic changes. The warehousing of infrequently requested models as banked embryos is an economical alternative to a continuously breeding colony. The cost of cryopreserving a rodent model is often less than the annual cost of maintaining a breeding colony (Mobraaten 1986).

We report here on the effect of genotype on the efficiency of embryo banking for diverse genotypes of rats and mice. Our past and ongoing efforts to establish banks of cryopreserved embryos provides two large databases for comparing embryo banking efficiency. One database contains the embryo collection results from approximately 11,000 rat embryo donors from 111 models maintained by the NIHAGR. The second contains the embryo collection results from 4,023 mouse embryo donors from 57 induced mutant models developed and maintained by NIH intramural investigators.

The physiological and behavioral responses of rats and mice to estrous cycle synchronization or superovulation treatments determine the effort required to collect the necessary number of embryos to establish an embryo bank. Many inbred strains and mutants exhibit impaired fertility and low embryonic viability that affect this effort. For example, in the mouse and rat, genotype is reported to influence ovulation rate in response to superovulatory treatments with gonadotropins (Bindon and Pennycuik 1974; Dorsch and Hedrich 1998; Durrant et al. 1980; Schmidt et al. 1989; Spearow and Barkley 1999), the ability of preimplantation mouse embryos to develop in vitro (Chatot et al. 1990; Dandekar and Glass 1987; Pomp and Eisen, 1991; Scott and Whittingham, 1996), and the rate of development and quality grades of preimplantation embryos (Al-Shorepy et al. 1992; Durrant et al. 1980; Goldbard et al. 1982; Wu et al. 1999).

Initial reports of mouse embryo banking noted differences in the postthaw survival of embryos from various inbred, outbred, hybrid, and mutant genotypes (Glenister and Lyon 1981; Whittingham et al. 1977a,b; Yokoyama et al. 1981). Subsequent studies designed to evaluate the potential effect of genotype on the efficiency of embryo banking have confirmed genetic variations in the postthaw survival of embryos (Dinnyés et al. 1995; Pomp and Eisen 1990, 1991; Schmidt et al. 1985, 1987; Vincente and Garcia-Ximénez 1993) but provided little insight as to the cryobiological factors that allowed embryos from one genotype to survive at rates higher than those of embryos from a second genotype. Dinnyés et al. (1995) reported high rates of postthaw, in vivo implantation of four inbred and one mixed genotype when optimized mouse embryo cryopreservation procedures were used. However, large differences in the efficiency of superovulation and postimplantation mortality were noted for these genotypes.

Our goal is to establish banks of 500 embryos for each mouse and rat genotype. This number is based on earlier reports of the effect of genotype on the rederivation of banked mouse embryos (Whittingham et al. 1977b), and the standard was based on an overall rederivation efficiency of not less than 10%. At this overall efficiency, a bank of 500 embryos carrying the gene of interest would be sufficient to yield approximately 50 pups for use as founders to reestablish a strain or stock. If five to 10 pups were rederived on each occasion, a 500-embryo bank would allow the genotype to be reestablished five to 10 times. Long-term viability issues require embryos from each model be divided between at least two storage refrigerators. In the event of the loss of embryos in one refrigerator, the remaining 250 embryos would allow the genotype to be reestablished two to five times.

We report first on the effect of genotype on the efficiency of embryo collection for diverse genotypes of rats using a database from 91 rat embryo banks established for the NIHAGR. A similar analysis is made for diverse genotypes of mice using a database from 57 mouse embryo banks established for induced mutant models developed and maintained by NIH intramural investigators. Embryo banking is ongoing for some of these rat and mouse genotypes. In those cases, we have extrapolated current data to estimate the number of embryo donor females required to establish a bank of 500 cryopreserved embryos. We then report on the effect of genotype on the efficiency of rederivation of these and other rat and mouse models by embryo transfer. All animals used in our studies were euthanized humanely in accordance with the Guide for the Care and Use of Laboratory Animals (NRC 1996).

Methods for Production, Cryopreservation, and Rederivation

Rat Embryos

Animals and embryo collection. Embryos were obtained from females from pedigree foundation colonies maintained by the NIHAGR. They represented 43 inbred strains, 35 con-genic strains, four outbred stocks, and nine mutant stocks. The estrous cycle of adult females (>7 wk old) was synchronized by intraperitoneal injection of 40 µg of an analog of gonadotropin-releasing hormone (GnRH-a; Sigma L-4513) (Sigma Chemical Company, St. Louis, Missouri) using a modification of methods described previously (Vanderhyden and Armstrong 1988). The day of treatment with GnRH-a was defined as Synchronization Day 1 (S-Day 1). Females were paired with either their father or brother (inbred strains) or a male appropriate for the breeding scheme of that colony (outbred stocks, mutant stocks) 2 days later. Females exhibiting a 4-day estrous cycle would be expected to mate on the evening of S-Day 5. (A small number of females exhibiting a 5-day estrous cycle would be expected to mate on the evening of S-Day 6.) Embryos were flushed from excised oviducts and uterine horns using phosphate-buffered medium PB1 medium (Whittingham 1974) on the morning of the 10th day after injection of GnRH-a (S-Day 10). Embryos were examined at 100 x magnification, and only those judged morphologically normal (8-cell to blastocyst stage with pigmentation appropriate for the strain, no lysed blastomeres, and a spherical zona) were selected for cryopreservation.

Cryopreservation. The cryopreservation procedure is described in detail elsewhere (Glenister and Rall 1999, protocols 4 and 6). Briefly, plastic insemination straws (0.25 ml, no. A001, IMV, L'Aigle, France) were prepared by pushing the cotton plug 0.9 cm into the straw. A column (6.8 cm) of 1 M sucrose in PB 1 medium was aspirated into the straw and followed sequentially with columns of air (0.8 cm), cryoprotectant solution (1 cm), air (0.8 cm), cryoprotectant solution (1 cm), and finally air until the first column wetted the cotton plug. Finally, the plug end of the straw was heat sealed and a handle was slip-fitted over the heat seal as described by Leibo (1984). Embryos were equilibrated in 1.5 M glycerol in PB1 medium (22°C, 15 min) and transferred into the third column aspirated into the straw using a mouth-operated glass capillary pipette. Straws were heat sealed, placed in -7°C ethanol in the chamber of a programmed freezing machine (BioCool II, FTS Systems Inc., Stone Ridge, New York), and seeded. After a holding period of 10 min at -7°C, the ethanol was cooled at 0.5°C/ min to -40°C. Finally, after a holding period of 10 min at 40°C, the straws were plunged and stored in liquid nitrogen.

Warming, cryoprotectant dilution, and embryo transfer. Straws were held in room temperature air for 10 sec and then warmed rapidly in 22°C water for 10 sec. Embryos were immediately diluted out of the cryoprotectant by a modification of the one-step sucrose dilution procedure (Leibo 1984). Briefly, the diluent and cryoprotectant columns in the straw were mixed by shaking the straw to dislodge the air bubbles. The straw was then incubated in 35°C water for 3 min and in 22°C water for 2 to 5 min. The contents of the straw were expelled into an empty Petri dish, and the embryos were transferred into PB 1 medium and incubated for at least 10 min. Survival was assessed by the in vivo development of thawed embryos after transfer into the oviducts of pseudopregnant or pregnant N:NIH recipient females. The estrous cycles of recipient females (7-12 wk old) were synchronized using the GnRH-a treatment procedure described above. Pseudopregnant recipients were obtained by pairing females with a vasectomized N:NIH male. Pregnant recipients were obtained by pairing females known to be homozygous for coat color pigmentation or the albino allele with an N:NIH male homozygous for the same coat color. Mating was confirmed by the presence of a copulation plug, cornified epithelial cells (pseudopregnant females), or spermatozoa (pregnant females) in vaginal lavage. An average of eight embryos (range, 3-11) was transferred into each oviduct on the afternoon of the day mating was confirmed (defined as Day 1 of pseudopregnancy or pregnancy). Females were allowed to litter.

Mouse Embryos

Animals and embryo collection. Embryos from females (4-6 wk old) were obtained from either the National Cancer Institute Animal Production contract or foundation colonies maintained by intramural investigators. Females were induced to superovulate with 5 IU of eCG (G4877, Sigma) followed 46 hr later by 5 IU of hCG (CG5, Sigma). Immediately after the hCG injection, females were individually paired with a mature male of the appropriate genotype (see below). Embryos were flushed from excised oviducts and uterine horns using PB1 medium 65-70 hr after hCG. Embryos judged morphologically normal (8-16 blastomeres) were selected.

Cryopreservation, warming, and cryoprotectant dilution. The cryopreservation, warming, and cryoprotectant dilution procedures for mouse embryos were identical to those described above for rat embryos.

Embryo transfer. Survival was assessed by the in vivo development of embryos after transfer into the oviducts of pseudopregnant recipient females as described by Glenister and Rall (1999). Recipients were naturally cycling B6D2F1 or FVC3F1 females (8-12 wk old) that were paired with a vasectomized B6D2F1 or FVC3F1 male. An average of eight embryos (3-11) were transferred into each oviduct on Day 1 of pseudopregnancy (day of finding of a copulatory plug). Females were allowed to litter.

Criteria for Establishment of a Bank of Cryopreserved Embryos

A model was considered "banked" when four conditions were met: (1) A total of 500 embryos carrying the gene of interest were cryopreserved, (2) pups had been rederived from banked embryos, (3) rederived pups exhibited the desired genotype, and (4) the rederived pups had bred and produced pups.

Efficiency of Embryo Collection

From Rat Donors after Estrous Cycle Synchronization

Banking of pedigree embryos from NIHAGR rat models was initiated in 1994. The embryo banking strategy was structured to complement current resources and procedures used to maintain a large number of small foundation colonies in a maximum barrier facility. Each foundation colony represents a closed breeding group of an inbred strain or outbred stock and is a single unique genotype (model). In a typical week, approximately 50 embryo donor females were assigned at weaning from 15 to 20 different foundation colonies. Donor females were synchronized, mated, and then transferred from the barrier facility to the embryo cryopreservation laboratory on the day of embryo collection. Information concerning the pedigree of each embryo donor and sire and results of embryo collection and cryopreservation were entered into a computer database.

The data presented here summarize the embryo production characteristics of all foundation colonies for which at least 20 donor females were processed, representing a total of 8,064 embryo donors from 91 foundation colonies. A total of 42,725 embryos was cryopreserved from these females. The effort required to establish an embryo bank for each genotype is directly proportional to the number of embryo donors that must be used to obtain 500 embryos. Analysis of our database allows prediction of the effort required to collect a sufficient number of embryos that varied markedly for each foundation colony. In the best case, 49 donor females would be required; in the worst case, 612. Most (60%) of the genotypes require 100 or fewer embryo donors to establish a bank of 500 embryos (median=83 donors).

The genotype of donor females was found to affect the efficiency of embryo collection in two ways (Figure 1). First, for rats, the proportion of females yielding embryos varied from 16 to 100% (mean=71%). For most (65%) genotypes examined in this study, more than two thirds of the synchronized embryo donors yielded embryos. Second, the mean number of embryos recovered varied considerably among genotypes. For rats, the mean number of embryos collected from females yielding embryos was 7.8, and, for individual genotypes, varied from 4 to 10.6.

Preliminary analysis suggests that some of the variation might reflect differences in the efficiency of estrous cycle synchronization, mating failure due to female and male infertility or behavioral incompatibilities, or anatomical abnormalities (e.g., ipsilateral gonadal agenesis of ACI/N). These differences presumably reflect the effects of inbreeding and deleterious mutations on reproduction (Silver 1995). For example, 14 genotypes require more than 150 embryo donors to establish an embryo bank. Eight of these genotypes are congenic strains carrying deleterious mutations, and the remainder are inbred strains with low reproductive efficiencies. Modifications of methods used to prepare donor females may increase the efficiency of embryo production from these genotypes.

From Mouse Donors after Superovulation

Banking of mouse embryos for intramural NIH investigators was initiated in 1996. Most of these models were induced mutations (transgenic, knock out, knock in, Cre-loxP) on standard or unique inbred and mixed backgrounds. The banking strategy varied depending on the needs of the investigator, the genetic and reproductive characteristics of the genotype, and resources available from the investigator. Some investigators supplied superovulated mated females on the day of embryo collection, and others requested various levels of technical support in producing the embryo donor females. For example, some investigators sent a group of stud males carrying the gene of interest (homozygous or heterozygous), and embryo donor females were prepared by embryo cryopreservation staff using females obtained from either the investigator's colony or a commercial breeder. Other investigators transferred breeding pairs (or the entire colony) and complete responsibility for breeding and preparing embryo donors to our laboratory. Embryo donors were processed at one of two satellite laboratories capable of receiving mice of any health status (with the exception of ectromelia, zoonotic, and human pathogens carried by mice).

The data presented here represent 4,023 embryo donors from a representative group of 57 induced mutant genotypes. A total of 31,104 embryos were collected and cryopreserved from these females. Three banking strategies were used to bank induced mutant mouse models. The banking strategy for about half of the genotypes (n=28) was to mate super-ovulated females that were homozygous for the desired induced mutant with males that were similarly homozygous. For the remaining genotypes, unaffected females (usually C57BL/6N, FVB/N, or BALB/cAnN) were purchased from a commercial breeder, superovulated, and mated with a male that was either homozygous (n= 19) or heterozygous (n= 10) for the desired induced mutation. The use of a homozygous male ensures that all of the banked embryos carry one copy of the desired gene. However, mating an unaffected female with a heterozygous male results, on average, in only half of the banked embryos carrying one copy of the desired gene. In the latter case, a total of 1,000 embryos must be collected to ensure that 500 embryos carry the desired gene.

The effort required to establish a bank of cryopreserved mouse embryos for each genotype varies and is directly proportional to the number of embryo donors that must be used to obtain 500 embryos. Analysis of our database allows assessment of the effort required to collect a sufficient number of embryos. This analysis indicates that the required number of embryo donors varied markedly for each genotype. In the best case, 23 embryo donor females were required; in the worst case, 256. Most (55%) of the models require 75 or fewer embryo donors to establish a bank of 500 embryos (median=70 donors).

The primary difficulty in establishing mouse embryo banks is, similar to the rat, the variable efficiency of obtaining embryos from donor females. The genotype of donor females was found to affect the efficiency of embryo collection in two ways (Figure 2). First, for mice, the proportion of females yielding embryos ranged from 24 to 95% (mean=65%). For most (61%) models examined in this study, more than 60% of the superovulated embryo donors yielded embryos. Second, the mean number of embryos varied considerably among genotypes. For mice, the mean number of embryos collected from females yielding embryos varied from 5.3 to 32.2 (mean=13.7).

The highest and lowest yielding genotypes were, respectively, a double knock out of glycocerebrosidase and metaxin genes on a mixed B6/129 background (69% of unaffected C57BL/6N donor females yielded embryos, 32.2 embryos/ female) and a transgenic on a hybrid background using unaffected B6D2F1 hybrid females and heterozygous males (49% yielded embryos, 8 embryos/female).

Some of the variation in embryo production between genotypes might reflect differences in the efficiency of super-ovulation. Generally, the highest producing genotypes were on a C57BL/6N or FVB/N inbred background. For many genotypes, the number of normal embryos collected was low and similar to the number expected from natural estrous cycles. In those cases, the superovulation treatment resulted in a synchronous estrous cycle. Another potential source of variation was mating failure due to female and male infertility or behavioral incompatibilities. These differences presumably reflected the effects of inbreeding and deleterious mutations on reproduction (Silver 1995). For example, 11 genotypes required more than 100 embryo donors to establish an embryo bank. Six of these genotypes exhibited low embryo yields even though unaffected C57BL/6N, FVB/N, or B6D2F1 hybrid females were used as the embryo donor female. Because these genotypes normally superovulate well, the difficulty was likely to be male fertility or related mating problems. Two genotypes were on a BALB/cAnN background that normally responds poorly to superovulation. Three genotypes were deleterious homozygous mutations on unique inbred backgrounds.

Efficiency of Rederivation

From Banked Rat Embryos

The results of transfers of 684 thawed embryos from 33 rat genotypes into 66 recipient females are summarized in Table 1. Approximately 76% of recipient females established pregnancy and yielded an overall rate of development of 27%. Comparisons based on the averages for each genotype indicate an average of 30% with a range of 10 to 58%. This variability is consistent with previous reports for the transfer of thawed mouse embryos for four inbred strains and a mixed genotype (e.g., Dinnyés et al. 1995).

Four recent requests by investigators to reestablish rat models that existed only as frozen embryos demonstrate the usefulness of rat embryo banking. For three of the models, LER/N, LEW/N-op (osteopetrosis), and LEW/N-tl (toothless), the foundation colonies were disbanded after, respectively, 75, 128, and 179 embryos were banked. In the fourth case, M520/N-lr (leaner), a "minimal bank" of 367 embryos was established before the foundation colony was disbanded. Nevertheless, sufficient numbers of pups were obtained after thawing and transfer of, respectively, 23, 94, 55, and 76 embryos to recover breeders to reestablish foundation colonies. Although two of the foundation colonies are currently in the early stages of expansion, three of the criteria for establishing a bank have been met. Pups were rederived from banked rat embryos, rederived pups exhibited the desired genotype, and the rederived pups have bred and produced pups. Embryo banking of these four models to reach our final criteria, 500 banked embryos, will resume when surplus females are available.

It is important to note that the most rigorous measure of the efficiency of rederivation of a genotype is the "overall" percentage that survived. This result is based on the total number of normal pups divided by the number of thawed embryos. It is the only measure that accurately reveals real-world production of normal pups from banked embryos.

From Banked Mouse Embryos

The results of transfers of 2,064 thawed mouse embryos from 59 genotypes into 119 pseudopregnant females are also shown in Table 1. Approximately 78% of the recipient females established pregnancy and yielded an overall rate of development of 23%. The average rate of development of the individual genotypes is 33% with a range of 11 to 53%.

A subset of our results for models on three inbred backgrounds can be compared with similar studies of the same strains reported by Dinnyés et al. (1995). We observed overall survival rates of 30% (n=408 embryos), 24% (n= 132), and 11% (n=252) for induced mutant models on, respectively, C57BL/6N, C3H/HeN, and BALB/cAnN backgrounds. Dinnyés and his colleagues reported similar overall survival rates of, respectively, 26% (n=291), 20% (n=293), and 14% (n=303) for the same inbred strains.

Conclusion

Our analysis shows a great deal of variation in the efficiency of embryo production from diverse genotypes of rat and mouse models. Some genotypes are efficient and require as few as 25 to 50 embryo donors to establish a bank of 500 embryos, whereas other genotypes may require hundreds of females. These results underscore the importance of the banking strategy. Namely, one size does not fit all. We described three strategies for banking induced mutant mouse models. Two involve the use of unaffected donor females purchased from commercial breeders (standard inbred strains or hybrids, which usually superovulate well). Each seeks to optimize embryo production while maintaining the current breeding scheme for the model. When these strategies fail or are inappropriate due to a nonstandard inbred background, serious consideration must be given to alternatives that preserve the induced mutation but alter the genetic background (e.g., hybrid or mixed background).

The phenotype of some induced mutations may reduce embryo production to the extent that embryo banking is prohibitively expensive. Cryopreservation of gametes or ovarian tissue (Sztein et al. 1998; Thornton et al. 1999) or extinction (and future recreation) may be the only practical choices. These alternatives, however, are not available for standard inbred strains of rats or mice, which are too important to abandon. Some help may become available if ongoing research (Spearow et al. 1999) provides new strategies for optimizing superovulation or estrous cycle synchronization methods for difficult genotypes.

We also observed considerable variation in the rederivation of normal pups born from banked embryos from diverse genotypes of rat and mouse models. This result has important practical implications for ensuring that sufficient embryos are available for future rederivation. Our goal to establish banks of 500 embryos is based on earlier reports of the effect of genotype on the rederivation of banked mouse embryos (Whittingham et al. 1977b). Our results show that the lowest efficiency was 10% for both mouse and rat embryo rederivation. This outcome confirms previous estimates of practical rederivation limitations and reaffirms the accepted size of a mouse embryo bank. Finally, our results for rat embryo banking and rederivation provide the first evidence that this assumption applies also to diverse genotypes of rat models.

Acknowledgments

We thank Dr. James S. Crowell, Jr., for helpful comments and suggestions for improving the manuscript and NIH intramural investigators for providing diverse induced mutant mouse models.

References

AI-Shorepy SA, Clutter AC, Blair RM, Nielsen MK. 1992. Effects of three methods of selection for litter size in mice on pre-implantation embryonic development. Biol Reprod 46:958-963.

Bindon BM, Pennycuik PR. 1974. Differences in ovarian sensitivity of mice

selected for fecundity. J Reprod Fertil 36:221-224.

Chatot CL, Lewis JL, Torres I, Ziomek CA. 1990. Development of 1-cell embryos from different strains of mice in CZB medium. Biol Reprod 42:432-440.

Dandekar PV, Glass RH. 1987. Development of mouse embryos in vitro is affected by strain and culture medium. Gamete Res 17:279-285.

Dinnyés A, Wallace GA, Rall WF. 1995. Effect of genotype on the efficiency of mouse embryo cryopreservation by vitrification or slow freezing methods. Mol Reprod Dev 40:429-435.

Dorsch MM and Hedrich HJ. 1998. Effective superovulation for rat inbred strains and in vitro culture of preimplantation embryos: A retrospective study. J Exp Anim Sci 39:99-108.

Durrant BS, Eisen EJ, Ulberg LC. 1980. Ovulation rate, embryo survival and ovarian sensitivity to gonadotropins in mice selected for litter size and body weight. J Reprod Fertil 59:329-339.

Glenister PH, Lyon MF. 1981. Experience of banking mutant genes. In: Zeilmaker GH, ed. Frozen Storage of Laboratory Animals. New York: Verlag. p 157-163.

Glenister PM, Rail WF. 1999. Cryopreservation and rederivation of embryos and gametes. In: Jackson I, Abbott C, eds. Mouse Genetics & Transgenics: A Practical Approach. 2nd ed. Oxford: Oxford University Press. p 27-59.

Goldbard SB, Verbanac KM, Warner CM. 1982. Role of the H-2 complex in preimplantation mouse embryo development. Biol Reprod 26:591-596.

Leibo SP. 1984. A one-step method for direct nonsurgical transfer of frozen-thawed bovine embryos. Theriogenology 21:767-790.

Mobraaten LE. 1986. Mouse embryo cryobanking. J in Vitro Fertil Emb Transfer 3:28-32.

NRC [National Research Council]. 1996. Guide for the Care and Use of Laboratory Animals. Washington DC: National Academy Press.

Pomp D, Eisen EJ. 1990. Genetic control of survival of frozen mouse embryos. Biol Reprod 42:775-786.

Pomp D, Eisen EJ. 1991. Variation among donor females in mammalian preimplantation embryo research. Theriogenology 35:1209-1224.

Schmidt PM, Hansen CT, Wildt DE. 1985. Viability of frozen-thawed mouse embryos is affected by genotype. Biol Reprod 32:507-514.

Schmidt PM, Schiewe MC, Wildt DE. 1987. The genotypic response of mouse embryos to multiple freezing variables. Biol Reprod 37:1121-1128.

Schmidt PM, Monfort SL, Wildt DE. 1989. Pregnant mares serum gonadotropin source influences fertilization and fresh or thawed embryo development, but the effect is genotype specific. Gamete Res 23:11-20.

Scott L, Whittingham DG. 1996. Influence of genetic background and media components on the development of mouse embryos in vitro. Mol Reprod Dev 43:336-346.

Silver LM. 1995. Mouse Genetics: Concepts and Applications. New York: Oxford University Press.

Spearow JL, Barkley M. 1999. Genetic control of hormone-induced ovulation rate in mice. Biol Reprod 61:851-856.

Spearow JL, Nutson PA, Mailliard WS, Porter M, Barkley M. 1999. Mapping genes that control hormone-induced ovulation rate in mice. Blot Reprod 61:857-872.

Sztein J, Sweet H, Farley J, Mobraaten L. 1998. Cryopmservation and ortho-topic transplantation of mouse ovaries: New approach in gamete banking. Biol Reprod 58:1071-1074.

Thornton CE, Brown SDM, Glenister PH. 1999. Large numbers of mice established by in vitro fertilization with cryopreserved spermatozoa: Implications and applications for genetic resource banks, mutagenesis screens, and mouse backcrosses. Mamm Genome 10:987-992.

Vanderhyden BC, Armstrong DT. 1988. Decreased embryonic survival of in vitro fertilized oocytes in rats is due to retardation of preimplantation development. J Reprod Fertil 83:851-857.

Vincente JS, Garcia-Ximénez F. 1993. Effects of strain and embryo transfer model (embryos from one versus two donor does/recipient) on results of cryopreservation in rabbit. Reprod Nutr Dev 33:5-13.

Whittingham DG. 1974. Embryo banks in the future of developmental genetics. Genetics 78:395-402.

Whittingham DG, Lyon MF, Glenister PH. 1977a. Long-term storage of mouse embryos at-196°C: The effect of background radiation. Genetical Res (Cambridge) 29:171-181.

Whittingham DG, Lyon MF, Glenister PH. 1977b. Re-establishment of breeding stocks of mutant and inbred strains of mice from embryos stored at -196°C for prolonged periods. Genetical Res (Cambridge) 30:287-299.

Wu LZ, Feng HH, Warner CM. 1999. Identification of two major histocompatibility complex class Ib genes, Q7 and Q9, as the Ped gene in the mouse. Biol Reprod 60:1114-1119.

Yokoyama M, Wakasugi N, Nomura T. 1981. An attempt to store inbred mouse strains. In: Zeilmaker GH, ed. Frozen Storage of Laboratory Animals. New York: Verlag. p 113-117.

¹Abbreviations used in this article: NIH, National Institutes of Health; NIHAGR, NIH Animal Genetic Resource.

Figure 1 Efficiency of embryo collection from 91 rat models (genotypes) maintained by the NIH Animal Genetic Resource from which at least 20 embryo donor females were processed. Each symbol represents a different genotype. The highest and lowest yielding genotypes were, respectively, a mutant stock designated rt3 (all rt3 donor females yielded embryos with an average of 10.2 embryos/donor female) and LEW/ S sN (16.7% of donor females yielded embryos with an average of 4.9 embryos/female). The number of donor females required to obtain 500 embryos can be calculated as 500 divided by the product of the fraction of treated donors yielding embryos and the average number of embryos per female. A list of genotypes used in this analysis can be found at the NIH Animal Genetic Resource Homepage on the Intemet (http://dirs.info.nih.gov/intramur/vrp/open/nihagr.htm).

Figure 2 Efficiency of embryo collection from 57 induced mutant mouse models (genotypes) banked for NIH intramural investigators. The different symbols indicate the embryo banking strategy (black symbols: homozygous females x homozygous males; open symbols: unaffected females x homozygous males; gray symbols: unaffected females x heterozygous males). See text for details.

Table 1 Summary of rederivation of thawed rat and mouse embryos by embryo transfer (1998 and 1999)

Species	Genotypes	Embryos thawed	Embryos transferred (no. of recipients)	Recipients establishing pregnancy (%)	Embryos transferred into pregnant recipients	Normal pups (%^a)	Estimated overall survival^b (average and range for genotypes^c)
Rat	33	684	673 (66)	50 (76%)	505	180 (36%)	27% (30%, 10-58%(
Mouse	59	2,064	1,781 (119)	93 (78%)	1406	476 (34%)	23% (33%, 11-53%)

^abased on total number of embryos transferred into pregnant recipients.
^bBased on total number of embryos thawed.
^cBased on individual overall survival averages of each genotype.