Abstract
Surgical embryo transfer in mice is a key technique in assisted reproduction and applied for different purposes in biomedical research. Due to its frequent application in rodent facilities across the world, further improvement of the procedure can substantially contribute to fulfil the principles of the 3Rs. Here, we investigated the effect of bilateral and unilateral left- or right-sided oviduct transfers on the success of embryo transfers. In total, we performed 223 embryo transfers (56 unilateral left, 56 unilateral right, 111 bilateral), in which we transferred 10–14 two-cell embryos each. We found that the type of transfer significantly influenced both the pregnancy rate of recipients and the survival rate of transferred embryos. Bilateral transfers yielded higher pregnancy and survival rates than left-sided unilateral transfers. Right-sided unilateral transfers yielded higher pregnancy rates than left-sided unilateral transfers and did not differ in embryo survival rates from bilateral transfers. We found no evidence that the number of transferred embryos affected the pregnancy rate. However, the number of born pups increased with the number of transferred embryos. In conclusion, unilateral embryo transfers into the right reproductive tract yield equally high pregnancy and embryo survival rates as bilateral transfers. Given that a second abdominal incision can be prevented and the time of surgery can be reduced, we recommend applying unilateral right-sided transfers, as this would reduce postoperative pain and lower the impact on recipients.
Introduction
Embryo transfer is best practice for rederivating pathogen-contaminated laboratory mice. 1 Moreover, it is crucial for generating genetically modified mice2,3 and for revitalising cryopreserved strains, 4 and it can help to overcome fertility problems in subfertile mice. The procedure requires several steps, including the preparation of embryo donors and recipient females, embryo culture and the transfer itself. Considering the enormous number of embryo transfers that are conducted in research laboratories worldwide, each step has to be optimised in order to refine animal treatment maximally and to minimise the number of animals required for this procedure.
Based on important groundwork of several researchers, the efficacy of embryo transfer in mice has continuously improved over the last 70 years.5–8 Different factors can influence the outcome of embryo transfers, such as the genetic background of transferred embryos, 9 their developmental stage and culture conditions, 10 the origin of recipient females and the appropriate stage of pseudopregnancy,11,12 the number of transferred embryos9,13 and the transfer procedure, including necessary anaesthesia and pain relief.14–16 All developmental stages of preimplantation embryos, from zygotes and two-cell stages to blastocysts, can be transferred by either surgical or transcervical transfer into the oviduct or the uterine horn. Transcervical transfers, first reported by Beatty in 1951, 17 yield the best results with blastocysts. 18 In the best-practice non-surgical procedure, the transfer pipette is inserted vaginally through the cervix and enters one of the two uterine horns randomly, resulting in a unilateral embryo transfer. For the transfer of zygotes and early cleavage stages, surgical transfer into the oviduct is preferred. Surgical embryo transfer is currently the most commonly applied transfer technique in mice. In comparison to the transcervical technique, surgical transfer can be advantageous, as it is possible, for example, to confirm pseudopregnancy in surrogate mothers by observing the swollen ampulla at 0.5 days post coitus (dpc) or corpora lutea at the ovaries at 2.5–3.5 dpc. Furthermore, surgical transfer can be performed bilaterally or selectively unilateral into the oviduct or the uterine horn of a recipient.
In a previous study on surgical embryo transfer in mice, we showed that transuterine migration of fertilised ova is highly unlikely and cannot be induced by an excessive number of unilaterally transferred embryos. 12 The results from this study highlight the extreme capacity of a single uterine horn in implanting and nourishing embryos, as the number of transferred embryos exceeded the average litter size of 11.8 pups of a naturally breeding Crl:CD1 (ICR) outbreed stock. 19 Given the high capacity of a single uterine horn, unilateral embryo transfers could be routinely applied in mice to reduce the impact on the animals without significantly reducing the outcome of the procedure. However, the question arises if both uterine horns are comparable in their capacity to implant and nourish transferred embryos. Two recent studies have compared the results of uni- and bilateral embryo transfers in mice,20,21 and both used the left oviduct for unilateral transfers and considered only the number of born pups as the outcome. Thus, a comparison of embryo development between the left and right uterine horn was not possible. However, the results revealed that depending on the recipient genotype, a minimum number of embryos was required for the success of bilateral transfers and that unilateral transfers were advantageous when lower numbers of embryos were transferred.
A direct comparison between bilateral and left- or right-sided unilateral uterine transfers in mice has been reported by Wiebold and Becker. 22 They showed that pregnancy rates tended to be higher in bilateral compared to unilateral transfers, although there was no difference between left- or right-sided unilateral transfers. Additionally, they found that embryo survival until day 18 of pregnancy was significantly higher for embryos transferred to the right uterine horn in both the bilateral and unilateral transfers. In combination with other findings from their study, they concluded that the left and right uterine horns are not equal in function.
Here, we provide data of surgical embryo transfers conducted for routine rederivation in mouse lines with undefined microbiological status. We always transferred fresh two-cell stage embryos and compared pregnancy rates and embryo survival in bilateral oviduct transfers with unilateral transfers to either the left or right oviduct.
Methods
Animals
We performed a total of 60 rederivations within about 14 weeks in order to restart a barrier mouse facility completely. Donors for rederivation belong to 60 different genetically modified mouse lines, predominantly of B6.Cg congenic, C57BL/6N, C57BL/6J, C57BL/10 and undefined genetic background. Because of their unknown hygienic status, donor mice were kept and prepared for embryo production in a quarantine room. Embryo recipients and vasectomised males used to induce pseudopregnancy in recipients belonged mainly to the Him:OF1 outbred stock. In addition to these in-house produced recipients, we used Crl:MF1, CD2F1 and B6CBAF1 females purchased from Charles River Laboratories, Germany. All mice had an approved SPF status according to FELASA recommendations. 23 In-house bred and commercially purchased mice were between 7 and 14 weeks old when embryo transfers were conducted.
All experimental procedures were discussed and approved by the Ethics and Animal Welfare Committee of the University of Veterinary Medicine Vienna and the national authority (Austrian Federal Ministry of Education, Science and Culture) in accordance with good scientific practice guidelines and national legislation (licence number: BMBWK-68.205/0240-BrGT/2005).
Housing
Recipient females were housed in open-top type II and III Macrolon® cages (Tecniplast, Buguggiate, Italy) under barrier conditions equipped with bedding (LIGNOCEL® Select; Rettenmaier and Söhne, Rosenberg, Germany), nesting material (PurZellin; Hartmann, Wiener Neudorf, Austria) and a house (Des Res™ Mouse House; LBS, Horley, UK) for enrichment. Standard rodent diet (V1124-300; Ssniff, Soest, Germany) and tap water acidified to pH 3 were accessible ad libitum. Room temperature was 22 ± 1°C, relative humidity 50 ± 10% and the photoperiod 12 hours light/12 hours dark.
Embryo collection and transfer
Two scientists, highly experienced in performing the procedure, conducted embryo transfers. Embryo donors (originating from different genetic backgrounds) were superovulated by an intraperitoneal (i.p.) injection of 5 IU PMSG (Folligon®; Intervet, Boxmeer, Netherlands) followed by 5 IU hCG (Chorulon®; Intervet, Boxmeer, Netherlands) 48 hours later before being paired 1:1 with respective males. A vaginal plug check of mutant lines, especially with C57BL/6 background, is unreliable because plugs of this strain are very small, sometimes fragmented and located deeply in the vagina. To avoid undetected mating, all superovulated females were euthanised at 1.5 dpc by cervical dislocation before their oviducts were dissected and flushed with M2 medium to collect two-cell embryos. Whenever possible, we used juvenile females, since premature C57BL/6 donor mice respond better to superovulation and produce more embryos than adults. 24
To induce pseudopregnancy, females were mated overnight 2:1 to vasectomised outbred or hybrid males. Immediately after vaginal plug control the next morning, plug-positive recipients received a new water bottle containing Mexalen children’s syrup (40 mg/ml paracetamol; ratiopharm GmbH, Bayern, Germany) for post-surgical pain relief during the next 48 hours. The amount of paracetamol was calculated to provide the mice per os with 200 mg/kg body weight over 24 hours. 16 The freshly harvested two-cell embryos were transferred into 0.5 dpc pseudopregnant recipient females as described in detail elsewhere. 25 Briefly, recipients were anaesthetised by an i.p. injection of 100 mg ketamine (Ketasol; Graeub Veterinary Products, Bern, Switzerland) and 4 mg xylazine (Rompun; Graeub Veterinary Products, Bern, Switzerland) per kilogram of body weight, and the animals’ eyes were covered with eye ointment (Oleovit; Fresenius Kabi, Bad Homburg vor der Höhe, Germany) to protect them from drying. After laparotomy, the reproductive tract of one side was gently extracted from the abdominal cavity, and the ovarian bursa was carefully opened. Embryos were transferred with a mouth-controlled glass capillary directly via the infundibulum into the oviduct. To prevent hypothermia during and after surgery, a warming plate was used until recipient mice recovered from anaesthesia. Afterwards, animals were individually placed in a new cage, and daily health checks were performed. We did not observe any indications for postoperative pain, which was assessed the morning after surgery and on the following days by scoring the nest-building performance 26 and by inspecting the groomed fur and posture of the animals.
Recipients were randomly assigned to bilateral or unilateral embryo transfer, and unilateral transfers were randomised for the left or right uterine horn using Microsoft Excel. The number of transferred embryos (unilateral or bilateral) ranged between 10 and 14 per recipient and was evenly distributed between transfer types. For bilateral transfers with an uneven embryo number, one side was randomly assigned to receive one embryo more than the other.
Statistical analyses
For statistical analyses and data visualisation, we used IBM SPSS Statistics for Windows v28 (IBM Corp., Armonk, NY), R v4.1.2 (The R Foundation for Statistical Computing, Vienna, Austria) and GraphPad Prism v9.3.1 (GraphPad Software, San Diego, CA). We performed two independent models to test how pregnancy rates and embryo survival were affected by our manipulations.
We ran a generalised linear mixed-effects model (GLMM) with a binomial distribution and included the incidence of pregnancy (no pups versus one or more pups) as the dependent variable, and recipient genotype, transfer type (bilateral, unilateral left, unilateral right), number of transferred embryos (10, 11, 12, 13 or 14) and the corresponding interaction as fixed factors. We included the surgeon as a random factor into the model in order to control for differences in embryo transfer outcome caused by the surgeon’s performance. We further ran a linear mixed-effects model (LMM) with survival rate (number of born pups divided by number of transferred embryos) as the dependent variable, and transfer type (bilateral, unilateral left, unilateral right), number of transferred embryos (10, 11, 12, 13 or 14) and the interaction as fixed factors. Again, the surgeon was included as a random factor.
We verified that model assumptions were fulfilled in testing the distribution of model residuals and used Fisher’s least significant difference as the post hoc test. We applied a backward stepwise removal procedure 27 to avoid problems due to the inclusion of non-significant terms. 28 Removed variables were re-entered one by one to the final model to obtain relevant statistics.
Results
We performed 223 embryo transfers for the rederivation of 60 mutant lines, and we found that the pregnancy rate in recipients (likelihood of recipients to have at least one offspring after embryo transfer) was significantly influenced by the type of embryo transfer (GLMM: F = 3.92, p = 0.021; Table 1). Pregnancy rates were significantly higher when transfers were performed bilaterally (post hoc test: p = 0.007) or unilaterally on the right side (post hoc test: p = 0.029) compared to the left side. No difference was detected when transfers were performed bilaterally versus the right side (post hoc test: p = 0.835). We did not find any difference in pregnancy rates depending on the number of transferred embryos (GLMM: F = 0.94, p = 0.440; Figure 1), and there was no significant interaction effect between the type of embryo transfer and the number of embryos transferred (GLMM: F = 1.53, p = 0.149). Also, recipient strain had no significant effect on the likelihood of pregnancy (GLMM: F = 0.41, p = 0.744).
Type and number of performed embryo transfers and resulting pregnancy rates.
Pregnancy rate is defined as the percentage of females that gave birth to at least one offspring after embryo transfer.
a,bp ≤ 0.05.

Pregnancy rates (percentage of females that gave birth to at least one offspring after embryo transfer) after unilateral and bilateral embryo transfer into the oviduct of recipient females depending on the number of transferred two-cell embryos. Pregnancy rates were unaffected by the number of transferred embryos (GLMM: F = 0.94, p = 0.440). Depicted are means ± 95% confidence intervals.
Embryo survival rate, defined as the number of born pups divided by the number of transferred embryos, was also significantly affected by the type of embryo transfer (LMM: F = 3.10, p = 0.048; Figure 2). Survival rates were significantly higher when embryos were transferred bilaterally compared to the left side (post hoc test: p = 0.018). No differences in survival rates were detected when transfers were performed bilaterally compared to the right side (post hoc test: p = 0.191) or between unilateral transfers to the right or left oviduct (post hoc test: p = 0.288). We found no difference in survival rates depending on the number of transferred embryos (LMM: F = 0.36, p = 0.835; Figure 3), although the number of pups born tended to increase with the number of transferred embryos (Kendall’s tau: p = 0.054; Figure 4). Again, there was no significant interaction effect between the type of embryo transfer and the number of embryos transferred (LMM: F = 0.51, p = 0.848), and we found no difference in survival rates depending on the recipient strain (LMM: F = 1.28, p = 0.281).

Embryo survival rates (percentage of born pups divided by the number of transferred embryos) after the transfer of two-cell embryos into the oviduct of recipient females depending on the type of transfer. Embryo survival rates were affected by transfer type (LMM: F = 3.10, p = 0.048).

Embryo survival rates (percentage of born pups divided by the number of transferred embryos) after unilateral and bilateral embryo transfers into the oviduct depending on the number of transferred two-cell embryos. Embryo survival rates were unaffected by the number of transferred embryos (LMM: F = 0.36, p = 0.835).

Number of born pups born in relation to the number of transferred two-cell embryos in unilateral and bilateral oviduct transfers. The number of pups born tended to increase with the number of transferred embryos (Kendall’s tau: p = 0.054).
Discussion
In this study, we used the routine rederivation protocol of genetically modified mouse lines to investigate the practicality of unilateral surgical embryo transfers and to determine potential differences in embryo transfer outcomes between bilateral and unilateral transfers depending on the transfer side. We compared pregnancy rates in recipient females (Table 1) and survival rates of two-cell embryos (Figure 2) between unilaterally performed transfers to the left or to the right oviduct and bilaterally performed transfers. We cannot exclude that the genetic background of transferred embryos might have affected the outcome, as described previously by Byers et al. 29 However, as we used several donors for the rederivation of each line and distributed the harvested embryos to several randomly assigned recipients, we can largely exclude a systematic bias in our data.
The results presented here emphasise that bilateral embryo transfers can be substituted by unilateral transfers if right-sided transfers become standard practise. Both pregnancy rate and survival of embryos transferred into the right oviduct are comparable with bilaterally conducted transfers, which show superior outcomes compared to left-sided transfers. We assume that avoiding the second abdominal incision necessary for bilateral transfers and the shortened time under general anaesthesia will considerably reduce postoperative pain and the general burden to recipients undergoing embryo transfers.
The number of 10–14 embryos per surrogate mother used in our study seems to be appropriate for fresh embryos transferred bi- or unilaterally. There was no difference in embryo survival within this range (Figure 3), and the calculated survival rates of about 40% were comparable to previously reported results.6,13,20 The transfer of 10–14 embryos is within the physiological ovulation rate of outbred and hybrid mice with high fertility, which were used as recipients in our study. Importantly, transferring a higher number of embryos in this range will reduce the number of required recipients and should thus be considered. A substantial reduction or increase in the number of transferred embryos below 10 or above 14 could reveal more distinct differences between unilateral and bilateral transfers, as has previously been shown in a study where 16 embryos were transferred. 22
We found no difference in pregnancy rates depending on the number of transferred embryos (Figure 1), and the number of born pups increased in all three transfer types proportionally with the number of transferred embryos (Figure 4). This is in accordance with previously published results. 6 However, the authors transferred zygotes ranging between 1 and 25, and the reported increase in developing foetuses per ICR surrogate mother progressively declined when more than 12 embryos were transferred, suggesting a restricted ‘uterine capacity’.
Bilateral embryo transfers are better when higher numbers of embryos are transferred per recipient and a minimum number of transferred embryos is required for a transfer to be successful.5,20,21 A recently published study performed unilateral left-sided oviduct transfers with 5–25 embryos per recipient to determine when optimal pregnancy rates and survival rates occur. 13 In line with our findings, the authors concluded from their results that 8–12 two-cell embryos are optimal for surgical oviduct transfers. The lowest embryo survival rate (32%) was found in transfers with the highest category of transferred embryos (i.e. 21–25). An even more pronounced reduction in survival rates and also pregnancy rates has recently been shown when more than 21 two-cell embryos were transferred per recipient female. 9
Wiebold and Becker reported a trend for higher pregnancy rates in bilateral embryo transfers compared to unilateral transfers into the left or right uterine horn. 22 Furthermore, the embryo survival rate in their study was significantly higher in the right uterine horn at day 18 of pregnancy, independent of whether the transfer was bilateral or unilateral right sided. The authors used SWISS-Webster outbred recipients and transferred 16 embryos (morulae and blastocysts) equally divided into both or all into one uterine horn. Interestingly, they also counted more corpora lutea at the right ovary of nulliparous SWISS-Webster outbred females mated with fertile males, and less resorptions in the right uterine horn, which could explain their differences between the left- and right-sided embryo transfers. A higher number of corpora lutea at the right compared to the left ovary was also reported in mice by Ribeiro et al. after several generations of selection for fertility, 30 suggesting that the right reproductive tract can indeed show increased performance.
In summary, we recommend the right-sided unilateral embryo transfer in mice as a standard practice, at least when transferring a moderate number of embryos, as performed in this study. According to our results and the reports of others, we suggest using the right side of the reproductive tract for oviduct transfer and also for transfer into the uterine horn, even though this was not tested here. Due to the limited uterine capacity and the fact that transmigration is a very rare event in mice, the optimal number of about 12 fresh embryos per recipient should not be excessively exceeded. Unilateral embryo transfer conducted with an appropriate number of embryos into the right side of the reproductive tract can contribute to the 3Rs without diminishing reproductive outcome.
Supplemental Material
sj-xlsx-1-lan-10.1177_00236772221149844 - Supplemental material for Comparison of unilateral and bilateral embryo transfer in mice
Supplemental material, sj-xlsx-1-lan-10.1177_00236772221149844 for Comparison of unilateral and bilateral embryo transfer in mice by Kerstin E Auer, Thomas Kolbe, Claudia Laschalt and Thomas Rülicke in Laboratory Animals
Footnotes
Acknowledgements
We gratefully acknowledge the excellent technical assistance of D. Klein and I. Mujic for animal work and Sabine Kreidl for lab work.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship and/or publication of this article.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
