In the postimplantation stages, the embryo is surrounded by maternal tissues and becomes more dependent on maternal nutrition as it grows. During placenta formation after embryo implantation, the embryo is separated by a specific but still not fully mature barrier from the mother’s blood circulation. When the placenta is completely formed (which is considered at E14.5 [55]), it functions as the maternal–fetal barrier (also called placental barrier or biological barrier between the mother and fetus) mediating the maternal–fetal transfer of a large variety of substances (such as carbohydrates, fats, dietary fiber, minerals, protein, vitamins, and water). In some cases, cells, viruses (i.e., human immunodeficiency virus (HIV), rubella virus, human papillomavirus, and hepatitis C), and nucleic acids such as miRNA present in maternal circulation can be transferred to the fetus (reviewed by Tetro et al. [56] and Figueroa-Espada et al. [57]). Notably, Tüzel-Kox et al. [58] demonstrated that liposomes injected intravenously into pregnant rats could be trapped in the placenta. The entrapped materials are then transported to the fetus as free degraded molecules. These results suggest that nucleic acids, such as plasmid DNA encapsulated with liposomes, may be transferred via the maternal–fetal barrier to the fetuses when intravenously injected into a pregnant mouse.

Tsukamoto et al. [59] first showed evidence of transplacental delivery of nucleic acids to fetuses at early-to-mid gestational stages (E9 to 15) when a single tail-vein injection of a plasmid carrying the lacZ gene (encoding β-galactosidase) complexed with Lipofectamine (5-carboxyspermylglycine dioctadecylamide (DOGS)) was administered. Notably, fetuses treated at E9 contained at least 40 times more plasmid DNA than those treated at E12 or E15. No plasmid DNA was detected in fetuses treated at E3 or E6. Furthermore, fetuses recovered after transplacental gene delivery (TPGD) exhibited blue deposition due to gene expression of lacZ derived from the introduced plasmid when stained in the presence of X-Gal, a substrate for β-galactosidase. This means that plasmid DNA injected into the tail veins of pregnant female mice can be successfully delivered through the placental interface to the developing fetuses, and fetal cells are effectively transfected with the introduced DNA. Since the report by Tsukamoto et al. [59], several researchers have demonstrated the feasibility of this technology (TPGD) for RNA interference (RNAi)-based suppression of a target gene using short hairpin RNA (or small interfering RNA) and miRNA [15,60,61] and genetic immunization of fetuses [62]. The feasibility of using liposomal reagents and other reagents, such as cationic tetraamino fullerene, to deliver plasmid DNA through TPGD is also shown by several groups [63,64,65,66]. Since this technique does not require surgery, such as in utero gene delivery experiments, and, therefore, is noninvasive toward pregnant females, it can be a useful tool for studying the effects of genes on embryonic development (reviewed by Nakamura et al. [8]).

When fetuses with YS and placenta were dissected one day after dye injection and inspected for possible transplacental delivery of dye into the fetuses, no appreciable presence of dye was discernible. Almost all the dye was trapped in the placenta and YS. However, the presence of exogenous DNA was still discernible in some of the fetuses when a genomic PCR was carried out using the isolated fetuses two days after TPGD with a Cre expression plasmid DNA complexed with FuGENE6, a lipid specified to facilitate DNA delivery [64]. Furthermore, staining of the isolated fetuses in the presence of X-Gal revealed predominant staining in the fetal heart and weak staining in the head and peripheral portion of vertebrae.

According to Kikuchi et al. [64] and Nakamura et al. [8], in the early stages of postimplantation (E5.5 to E9.5), substances, including DNA/lipid complexes in maternal blood, are taken up by the visceral endoderm (VE) or YS and then transported to embryos by diffusion or vitelline circulation. During placental maturation, it may become a major tissue for controlling nutrient transfer in maternal blood. Most DNA/lipid complexes may be trapped in the VE/YS/placenta, and small amounts may be taken up and transported to the fetuses. Indeed, successful TPGD has been reported at E12.5 and E11.5 for delivering AAV particles [67] and plasmid DNA/liposome complexes [63], respectively. On the other hand, other groups successfully performed TPGD at E9.5 and E6.5 for delivering RNAi [15] and plasmid DNA/liposome complexes [59], respectively. 

As mentioned previously, CRISPR/Cas9-based genome-editing system uses only two components, namely, Cas9 endonuclease and gRNA. Transient expression of these components at the transfected cells or tissues is enough to induce indel-based mutations at a target locus. This plasmid, pCGSap1-EGFP, confers simultaneous Cas9 and gRNA expression targeted to EGFP cDNA upon transfection [7]. A solution containing pCGSap1-EGFP complexed with FuGENE6 was intravenously injected into the tail vein of pregnant wild-type female mice that had already been mated with male Tg mice carrying EGFP transgenes in a homozygous (Tg/Tg) state at E12.5. Without CRISPR/Cas9 system application, all the fetuses should express EGFP systemically because they carry the transgenes in a heterozygous (Tg/+) state. However, TPGD-based delivery of CRISPR reagents targeted to EGFP will likely reduce EGFP fluorescence levels in these fluorescent fetuses due to genome editing in one allele having the chromosomally integrated EGFP transgenes. When fluorescence was inspected for fetuses isolated two days after TPGD, three of 24 fetuses exhibited reduced fluorescence in their heart, consistent with the previous finding that the fetal heart is an organ preferentially transfected with TPGD [64]. Molecular biological analysis of these isolated fetuses demonstrated the presence of the transgene construct (Cas9 gene) and indels at the target EGFP sequence. Notably, these fetuses showing reduced fluorescence comprised genome-edited and unedited cells as mosaic mutations. These results suggest that this TPGD-based genome editing, called TPGD-GEF, is effective in causing mutations at a target locus in a specific part (embryonic heart, in this case) of a fetus. It can potentially produce cardiovascular disease models and aid in basic research on fetal gene therapy for congenital heart diseases.

KO MHCα mice exhibited heart failure due to acute cardiac hypertrophy [68]. A solution containing pCGSap1-MHC encapsulated with FuGENE6 was intravenously administered through the HGD approach to pregnant females at E9.5 or E12.5. When fetuses were inspected two days after TPGD-GEF, only one female was confirmed to have genome-edited fetuses out of four females treated at E9.5 with a ratio of 40%. However, none of the genome-edited fetuses were obtained from the other two pregnant females (22 fetuses tested). Notably, none of the genome-edited fetuses were obtained from three pregnant (E12.5) females (31 fetuses tested). Molecular analysis revealed that all genome-edited fetuses recovered comprised a mixture of genome- and non-genome-edited cells. This mosaicism was found in the fetal heart and other organs. These findings suggest the variability and feasibility of TPGD-GEF coupled with hydrodynamics-based gene delivery (HGD) in E9.5 fetuses for the possible production of individuals with heart failure as a disease model.

Despite the potential risk of frequent embryonic lethality, treatment in utero, also called in utero gene therapy, fetal gene therapy, or prenatal gene editing, offers several distinct advantages over postnatal treatment. Small fetus size allows the delivery of a higher effective dose of the gene therapy. Immune tolerance can be stimulated, and the phenotypic onset of genetic diseases that manifest at perinatal stages can be prevented. Table 1 summarizes previous genome-editing applications targeting developing murine fetuses. This table describes several parameters of each procedure, such as ease of the procedure, tissue type mainly transfected, efficacy, safety, and cost.

The past experiments using in utero genome-editing approach and TPGD-GEF have demonstrated the feasibility of CRISPR/Cas9-based genome editing in mice. Most studies rely on physical approaches, including glass micropipette-aided injection, EP, tail-vein injection, and HGD. In utero, genome editing is always associated with a surgical treatment requiring a microscopic-guided injection of the gene-editing cargo into the target cells. However, the latter is technically difficult to perform and often a cause for fetal death due to mechanical damages. EP after in utero gene delivery is also one of the risk factors leading to fetal death, which, therefore, requires careful optimization of EP conditions. HGD requires introducing a large amount of fluid at once, which often causes physiological damage to the liver and other organs. Furthermore, the reagents used for in vivo genome editing of fetuses are viral vectors (including rAAVs, Ad, and retroviral vectors) as well as nonviral vectors (including plasmid DNA). Safety is the primary concern with currently used viral vectors, especially Ad, which often elicits a potential immunological response when repeated infection trials are attempted [69].

Overall, nonviral nano-vectors, also called NP, can exhibit profound advantages over the abovementioned physical approaches and virus-based delivery. Particularly, tail-vein injection of nonviral vectors complexed with DNA delivery reagents, such as liposomes, would be an ideal approach to achieve genome editing in fetuses because it is safer and relatively noninvasive than in utero genome editing. However, low genome-editing frequency remains an issue to be dissolved in the future.

To date, several types of nano-vectors, including lipids (such as 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine [DOPE] and cholesterol), polymers (such as polyethyleneimine, poly-L-lysine, and chitosan), inorganic chemicals (such as cationic arginine gold nanoparticles and CRISPR-Gold), and exosomes (such as exosome-liposome hybrid and engineered exosomes: CD9-HuR exosomes and NanoMEDIC) have been developed [70,71,72]. Since these nano-vectors can prevent the degradation of the CRISPR/Cas9 genome-edition system and directly enter into the nucleus to perform genome editing in vivo, they can be ideal delivery platforms for the CRISPR/Cas9 system, enabling effective ex vivo or in vivo transfection of CRISPR reagents encoding DNA, RNA, or RNP in a highly efficient and safe way [73,74,75]. For example, Wei et al. [76] encapsulated RNPs with cationic lipids composed of 1,2-dioleoyl-3-(trimethylammonium) propane as permanent cationic supplements, DOPE as a helper lipid, and cholesterol as a sterol, postmodified with polyethylene glycol phospholipid as PEGylated lipids to yield nanoparticles with retained activity and redirect DNA editing to the target tissues with decreased clearance and immunogenicity. They demonstrated that low-dose intravenous injections could effectively target specific tissues, including the sphincter muscles, brain, liver, and lungs. Aside from the study of Wei et al. [76], there have been many successful reports for in vivo genome editing using lipids [77,78,79,80], polymers [81,82,83,84,85,86,87], inorganic nanomaterials [88,89,90,91,92,93], and exosomes [94,95,96,97].

These nano-vectors can be easily scaled up and modified chemically. They are cost-effective, have large packaging capacity, and lower immunogenicity, which will match the demand for in utero and tail-vein injection-based gene delivery to fetuses. Notably, Thermo Fisher Scientific has formulated a lipid-based chemical transfection reagent optimized to deliver Cas9 RNP complexes [98].