CRISPR/Cas9 microinjection in oocytes disables pancreas development in sheep

Abstract

One of the ultimate goals of regenerative medicine is the generation of patient-specific organs from pluripotent stem cells (PSCs). Sheep are potential hosts for growing human organs through the technique of blastocyst complementation. We report here the creation of pancreatogenesis-disabled sheep by oocyte microinjection of CRISPR/Cas9 targeting PDX1, a critical gene for pancreas development. We compared the efficiency of target mutations after microinjecting the CRISPR/Cas9 system in metaphase II (MII) oocytes and zygote stage embryos. MII oocyte microinjection reduced lysis, improved blastocyst rate, increased the number of targeted bi-allelic mutations, and resulted in similar degree of mosaicism when compared to zygote microinjection. While the use of a single sgRNA was efficient at inducing mutated fetuses, the lack of complete gene inactivation resulted in animals with an intact pancreas. When using a dual sgRNA system, we achieved complete PDX1 disruption. This PDX1^-/- fetus lacked a pancreas and provides the basis for the production of gene-edited sheep as a host for interspecies organ generation. In the future, combining gene editing with CRISPR/Cas9 and PSCs complementation could result in a powerful approach for human organ generation.

Figure 1

CRISPR/Cas9 microinjection of sheep oocytes and zygotes. (a) Schematic representation of CRISPR/Cas9 injection in sheep MII oocytes and zygotes. Presumptive embryos were in vitro cultured until the blastocyst stage for embryo genotyping by Sanger sequencing and mutations were determined using the TIDE bioinformatics package. (b) Lysis after microinjection of CRISPR/Cas9 was lower in MII oocytes than in zygotes. Development was higher after microinjection in MII oocytes and non-injected embryos compared to zygotes (*P < 0.05; **P < 0.05). (c) Sanger sequencing results from a bi-allelic and a mono-allelic mutant sheep blastocyst. The PAM sequence is underlined and the gRNA target region is shown in blue. Red dashes represent deletions. Mutation efficiency is presented in the pie chart. For MII injected oocytes they were 46% (6/13) bi-allelic mutated blastocyst and 20% (3/15) for zygote microinjected blastocysts.

Figure 2

Deep sequencing reveals different mutations in sheep blastocysts microinjected with CRISPR/Cas9. Alignment of next-generation sequencing data obtained from sheep blastocysts injected with CRISPR/Cas9 targeting PDX1 at MII oocytes (a) or at Zygotes (b). Mutations with frequencies higher than 12.5% were considered. Frequencies of alleles are shown in the HeatMap on the right panel and the top row refers to the total number of reads analyzed. Insertions with respect to the reference are presented underneath the alignments. (c) Number of alleles in mutant embryos after CRISPR/Cas9 microinjection in MII oocytes and Zygote.

Figure 3

Targeting PDX1 with a single sgRNA resulted in in-frame mutations. (a) Schematic representation of the sgRNA targeting PDX1 gene in sheep. Sheep oocytes were microinjected with Cas9 mRNA and PDX1 sgRNA, cultured in vitro, and 16 injected blastocyst and 4 un-injected blastocysts were collected and transferred to 5 recipient ewes. (b) Seven fetuses (3 controls and 4 microinjected) were collected at day 75 of development and 2 of the microinjected fetuses had mutations. (c) Sequencing results from 9-10 colonies of each of the mutant fetuses after PCR and sub-cloning of the target region. Red dashes represent deletions and red letters insertions; insertions (+) or deletions (−) are shown to the right of each allele. The PAM sequence is underlined and the target region is shown in red. (d) Protein sequences of the Mutant#2 indicate that one of the mutant alleles only disrupted a few amino acids indicated in red, and therefore potentially active PDX1 might be present. (e) The pancreas was present in both mutant fetuses.

Figure 4

CRISPR/Cas9 using dual sgRNAs can effectively knockout PDX1 in sheep. (a) Schematic representation of the two gRNAs designed to target PDX1 loci in sheep. Dual sgRNA microinjection can induce a bi-allelic deletion that can be identified by PCR amplification and gel electrophoresis of the target region. The full-length gel is presented in the Supplementary Information. Mutation efficiency of the dual sgRNA microinjection is presented in the pie chart. From 21 microinjected oocytes 4 had mono-allelic and 4 had bi-allelic deletions. (b) Two sgRNAs targeting PDX1 gene were microinjected into the MII II oocytes before IVF, cultured in vitro for 6 days and transfer to a recipient sheep. The fetus was collected at 4 months of gestation. (c) Genomic DNA was isolated and subjected to PCR, sub-cloning and Sanger sequencing. All of the sequenced colonies showed mutations with a 208 bp deletion. (d) Gel electrophoresis of PCR product -using specific primers for PDX1- from different tissues (liver, lung, heart, kidney, muscle and spleen) of the mutant fetus. Full-length gel is presented in the Supplementary information. (e) Protein sequence of the disrupted allele of the PDX1-KO sheep fetus is shown in red.

Figure 5

PDX1-KO phenotype in sheep. (a) Macroscopic appearance of the vestigial pancreas of a PDX1^−/− 4-month-old male fetus compared to a WT fetus of the same age. In the right panel the dashed lines indicate the pancreas (WT) and the vestigial pancreas (PDX1^−/−) that were isolated for histology. St.: stomach; D.: duodenum. (b) Histology of the pancreas. The left panel shows representative images (40X) of the pancreas and vestigial structure stained with hematoxylin and eosin. Insets are high magnification images (200X) to illustrate the lack of Langerhans islets in the PDX1^−/− vestigial tissue. Right panel shows pancreatic tissue sections immunostained for PDX1 (green) and Insulin (red). The withe arrow indicates a PDX1 positive cell. Bar indicates 50 µm.