Results and discussion
Several papers described methods for the transcervical transfer of embryos into pseudopregnant female mice a number of years ago; these earlier methods generally used rigid (i.e., glass) tubes (8,9,10,11). We initially attempted similar approaches as a means of performing nonsurgical embryo transfer. Problems we encountered with similar glass tubes included volume control, locating and damaging the cervix, and inserting the device to a consistent depth without puncturing the uterus. Reasoning that flexible tubing would be more amenable to embryo transfer, we developed several prototype needles from catheters attached to a 2-µL pipet tip, which could then be used with a P-2 Pipetman pipet (Gilson, Middleton, WI, USA) to allow for the transcervical transfer of controlled volumes (Supplementary Figure 1). Experiments performed on euthanized female CD-1 female mice indicated that the 24-gauge, 3/4-in catheter (Becton Dickinson) (Supplementary Figure 1B) was best at penetrating the cervix without damaging the uterus.
To test whether the embryos transferred using the NSET device developed to term, a series of experiments were performed. Fertilized eggs were obtained from the mating of B6C3F1 mice and incubated at 37°C in KSOM media in the presence of 5% CO2 prior to NSET transfer into pseudopregnant CD-1 females, 2.5 days post coitus (dpc) with vasectomized males. Initially, embryonic day 0.5 (e0.5) B6C3F1 embryos were allowed to develop in vitro to the blastocyst stage (e3.5) before NSET-mediated transcervical transfer into 2.5 dpc females. Roughly 95% of the embryos developed to blastocysts when cultured in vitro. Twenty embryos were transferred to each of six pseudopregnant CD-1 females, and five of these had pups (83%). A total of 40 pups were born, which represents 33% of the transferred blastocysts. We also transferred embryos that had developed from e0.5 to the morula stage (e2.5) in vitro. When transferred to pseudopregnant 2.5 dpc females, ~15% (29 of 200) of these morula-stage embryos developed to term. When we transferred two-cell embryos that had developed overnight (24 h), <1% (1/166) developed to term, as expected, since other studies suggested that these embryos would not survive when transplanted to the uterus (11). Based on these results, all future studies were performed with e3.5 embryos transferred to 2.5 dpc pseudopregnant females.
The transcervical transfer itself is quite straightforward (Figure 2; see Supplementary Materials for a detailed protocol). Embryos are collected and transferred to a media drop (15 µL) placed in a plastic Petri dish [BD Falcon (Cat. no. 351029; Franklin Lakes, NJ, USA) or similar]. The NSET device, attached to a P-2 Pipetman, is then used to draw 1.8 µL of media containing the embryos (which can be readily visualized using a dissecting microscope). A speculum is inserted into the vagina and positioned around the cervix. The catheter is then inserted into the speculum and through the cervix. At this stage, the media containing the embryos is ejected from the catheter into one of the uterine horns. The entire procedure takes several seconds. The recipient pseudo-pregnant females are not anesthetized and exhibit no apparent discomfort. The procedure requires no special equipment and little training, and can be performed in a laminar flow hood if necessary. Getting through the cervix requires some practice, but does not require the level of expertise or equipment that is needed for surgical transfer techniques. Importantly, recipient females are not subjected to the stress and pain of surgery, anesthesia is not required, and post-operative complications such as infection are not a concern. The NSET device randomly enters one of the two uterine horns, since transferred embryos always develop in a single uterine horn. A video of this procedure, demonstrating its ease and speed, can be found online at http://www.uky.edu/~magree00/MousePipet.avi.
In order to develop the NSET technique as a potential replacement for embryo transfer surgery, we optimized several parameters and compared the success rate of NSET to standard surgical procedures. To optimize the number of pups born relative to the number of embryos transferred, we removed one-cell embryos from pregnant mice and allowed them to incubate for three days in KSOM media. The resulting blastocysts were transferred to 2.5 dpc pseudopregnant females. The data indicate the transfer of 12 or 24 blastocysts results in smaller litter sizes than the transfer of 16 or 20 blastocysts. Transfer of 16 or 20 blastocysts resulted in average litters of ~7 live pups in either case, although the percentage of live pups was higher with 16 blastocysts (28.3% versus 20.7%). Based on the percentage of live pups and the litter size, 20 blastocysts were transferred in subsequent experiments. In comparison, 27.3% of embryos became live pups when performing standard surgical transfer of embryos. We also evaluated embryo transfer success as a function of the weight and age of the recipient CD-1 females. We had optimal success with recipient females that weighed ≥26 g and were ≥60 days old.