Using these optimized parameters, we then tested whether NSET could be used for producing transgenic mice by DNA microinjection. We directly compared the surgical and nonsurgical transfer of embryos that had been injected with five different transgenes that were purified using cesium chloride (CsCl, Cat. no. C-3032; Sigma, St. Louis, MO, USA) ultra-centrifugation or Qiagen kits (QIAquick Gel Extraction Kit, Cat. no. 28706; Qiagen, Gaithersburg, MD, USA). For each DNA, injections were performed in one-cell B6C3F1 embryos, which were then split into two groups. The first group of embryos was surgically transferred to the oviducts of 0.5 dpc recipient females on the day of injection; 30 embryos were transferred per mouse (into one oviduct, which was our standard procedure). The second group of embryos was incubated to the blastocyst stage (e3.5) before nonsurgical transfer to 2.5 dpc recipient females (20 embryos were transferred per mouse). This data indicates that the surgical and nonsurgical procedures are roughly equivalent with regard to generating transgenic offspring (Figure 3). Of the surgically transferred embryos, 25% became live pups. Of these, the number of pups that were transgenic ranged 13–25%, with an overall rate of ~19%. These numbers are consistent with the historical success rate of our transgenic facility (data not shown). While 25% of the embryos transferred using NSET became live pups, this represented 17% of the one-cell embryos. This means that ~1/3 of the injected embryos did not survive the in vitro culturing prior to NSET transfer. However, we did find that 20% of the pups born using NSET were transgenic, indicating no substantial difference in the success of generating transgenic founder mice between surgical and NSET-mediated embryo transfer.
We also determined whether NSET technology could be used to generate chimeric mice from embryonic stem (ES) cell–containing embryos. Chimeric aggregates were generated using the techniques described by Nagy and colleagues (6) using morulae from C57Bl/6 or B6C3F1 mice. These were transferred using NSET to nine recipient CD-1 pseudopregnant females in four separate experiments. Pups were born to six recipients, and the percentage that developed to term in five of these was quite reasonable (50–65%). Chimerism of the pups was evident based on coat color variegation (C57Bl/6 and B6C3F1 mice are black and agouti, respectively, whereas CD-1 are albino). Based on these results, we tested whether ES cell–containing chimeric mice could be generated using the NSET device. Using the R1 ES cells in which the enhancer region of the α-fetoprotein gene had been deleted using standard technology, we obtained chimeric mice, as judged by coat color. We were able to obtain germline transmission of the enhancer-deleted allele from one of the ES cell–derived chimeric males (L. Jin. L. Long, M.A.G. and B.T.S., unpublished data). Furthermore, cryopreserved embryos have been successfully transferred using the NSET device (data not shown).
Based on the results with our handmade NSET devices, we have manufactured NSET devices under current good manufacturing practices (cGMP) guidelines, with the catheter tip extruded from Teflon-FEP (DuPont) and the hub component molded using polyethylene (Figure 1). The speculums are extruded from polyethylene. Components are packaged under clean room conditions and sterilized. The manufactured NSET devices are sterile, more uniform in size, and more stable than the handmade devices. The manufactured NSET device has been tested in several experiments to generate transgenic mice and our preliminary data indicates that this works as well as the handmade devices (data not shown).
In summary, we have shown that NSET-mediated embryo transfer is equally efficient to standard oviduct transfer for the production of transgenic mice via microinjection. We have also successfully used this technology to transfer ES cell–containing blastocysts and cryopreserved embryos. One potential drawback with nonsurgical embryo transfer is that the status of the ovaries and uteri cannot be evaluated for inflammation, so embryos may occasionally be transferred into females where embryo development is unlikely to occur. NSET technology removes the pain and distress associated with the surgical transfer of embryos; eliminates the need for anesthesia, surgery, and post-operative recovery (and their potential complications, such as infections); and reduces the time required to monitor animals following surgery. The elimination of a surgical procedure represents a substantial refinement in embryo transfer [one of the three ‘Rs’ of animal research (reduction, refinement, and replacement) with the stated goal of improving humane treatment of experimental animals (12)]. NSET technology represents a substantial savings in time and eliminates the need for sterilization of specialized surgical equipment. Finally, NSET technology is substantially easier than uterine and oviduct surgery, thus requiring less training, and should facilitate efforts to manage and transport lines by embryo and sperm cryopreservation (13,14,15,16). This reduction in the level of expertise and specialized facilities required for surgical transfer could allow a greater number of institutions to participate in work involving genetically modified mice.