Genome editing for common marmosets

Seven years ago (2009), I contributed an article to this newsletter reporting that by inducing the green fluorescent protein gene using a lentivirus vector in the embryo of a common marmoset (Callithrix jacchus), a small primate (hereinafter referred to as a marmoset), green transgenic marmosets were created [1].
Many useful genetically modified mouse models have been developed using target-gene knockout technology. Therefore, the development of a target-gene knockout model was considered necessary for marmosets. However, to create a model using target-gene knockout technology, embryonic-stem (ES) cells are essential, from which germ-line chimeras can be produced. Because ES cells with chimera-forming ability have not been established in animals other than mice and rats, target-gene knockout models for these animals could not be created. However, using genome-editing techniques, target-gene knockout models can be developed without using ES cells with chimera-forming ability.

For a description of the usefulness and characteristics of marmosets as experimental animals, please refer to my previous article [1]. Our laboratory has been developing techniques to produce marmosets that can be used as human disease models using gene modification techniques. After establishing techniques to produce transgenic marmosets, we used various research methods to produce target-gene knockout marmosets. However, no useful results were obtained. A recent study reported that genome-editing techniques were useful for mammalian fertilized eggs, and thus we began to develop a target-gene knockout model from marmoset fertilized eggs using genome-editing techniques.

To develop genome-editing techniques for marmosets, we selected the interleukin-2 receptor common gamma chain (IL2rg) gene as the target gene.
We chose this gene for the following reasons: (1) the IL2rg gene exists on the X-chromosome, and a male marmoset develops immunodeficiency using only a mono-allelic knockout; (2) when an immunodeficient marmoset is produced, the produced marmoset can be used as a model in various research fields, such as regenerative medicine, hematology, oncology, and transplantation immunology, and (3) the produced marmoset develops immunodeficiency from its birth, and whether genome-editing techniques are effective for marmosets can be clarified during the early stage of growth.

For primates with long life cycles, it is preferable that the intended phenotype is expressed in the founder generation^∗ 1 and that the modified gene is transmitted to the next generation. We aimed not only to knock out the target gene, but also to create an individual that was not in a mosaic state and in which the modified gene was uniformly distributed. If the modified gene in the individual is in a mosaic state and a wild-type gene remains, it is likely that the intended phenotype will not be expressed in the founder generation. Moreover, in experiments using primates, animal ethics must be taken into consideration and the development of animals in which modification of the target gene has failed should be prevented.

We then selected artificial nucleases with excellent cleavage efficiency using fibroblasts and established a method for estimating the acquisition ratio of knockout animals using marmoset fertilized eggs and in vitro genome editing to obtain the mosaic ratio of individuals (Fig. 1). The acquisition ratio and mosaic ratio estimated using this method were nearly correlated with or slightly lower than the actual creation ratio and the actual mosaic ratio of individuals, respectively. Currently, using this screening method, other genes can be successfully knocked out. Therefore, this screening method may be effective for genome editing in primates.

By injecting eHiFi-ZFN and platinum TALEN, which are artificial nucleases that target the IL2rg gene, into an in vitro marmoset fertilized egg in the one-cell stage, we created marmosets in which the IL2rg gene was knocked out and all cells possessed the same modified gene sequence (Fig. 2) [2]. In the cord blood of these individuals, T and NK cells were deleted, the number of B cells was normal and similar to that in human patients suffering from X-SCID^∗ 2 (in mice, the number of B cells also markedly decreased), and the thymus gland was not present according to autopsy of deceased individuals. Therefore, these individuals were confirmed to be immunodeficient marmosets. Initially, these individuals were difficult to breed. However, because of improvements in anesthesia during Caesarean operation and oxygen concentration immediately after birth, the individuals could be bred for more than two years. When the sperm of a sexually mature male individual was investigated, the modified gene sequence, which was the same as that of somatic cells, was confirmed in the sperm. Therefore, germ-line chimeric marmosets can be produced.

In contrast, T cells, which were not confirmed to be present at birth, appeared, and B cells were deleted in adult individuals. In T cells, the number of which was increased, it was confirmed that the IL2rg gene was non-functional. Therefore, a signal transmission mechanism not involving the IL2rg gene may be related to the increased number of T cells. A similar phenomenon was reported in human patients suffering from X-SCID. The elucidation of the signal transmission mechanism related to the increased number of T cells is an interesting research theme that should be examined in future studies. IL2rg-knockout marmosets are useful as experimental models.