LAUSR

Dear Editor, Currently, wild-type (WT) hamsters, ferrets, cats, and non-human primates are being used as COVID-19 animal models. However, no severe clinical symptoms develop in these animals. Similarly, most of the human ACE2 (hACE2) transgenic mouse models develop only mild COVID-19 disease with only a few recent transgenic mouse models developing severe and fatal respiratory diseases, calling for a better large animal model that could mimic the full spectrum of COVID-19 symptoms. Although the pig is thought to be a better model for human diseases in general, due to its similarity to human anatomy, physiology, and immunology, previous studies have shown that WT pigs are not susceptible to SARS-CoV-2. Here we report our attempt to create the first humanized pig expressing the hACE2 receptor for COVID-19 research, speculating that humanization of the pig ACE2 receptor could make pigs susceptible to SARSCoV-2. To create a COVID-19 pig model with the targeted insertion of hACE2 at the pig ACE2 locus, we constructed a homologous recombination donor vector with homology arms of ~1 kb on each side. We used the CRISPR/ Cas9 system to increase the chance of homologous recombination. Single-guide RNAs (sgRNAs) closest to the start codon of exon 1 were selected for the construction of sgRNA-expressing vector pX459 (Fig. 1a). To optimize the efficiency of the sgRNA, 9 sgRNAs were synthesized and assembled. IBRS-2 porcine kidney cells were electroporated with plasmids of pACE2-sgRNAs and Cas9. Sanger sequencing was used to identify the indels and evaluate the targeting efficiency for these sgRNAs. The cleavage efficiency of sgRNA3 was higher than that of other sgRNAs. Therefore, we chose sgRNA3 for subsequent experiments. Porcine fetal fibroblasts (PFFs) isolated from embryos of Bama mini-pigs were electroporated with plasmids of pACE2-sgRNA3, Cas9, and donor linearized by in vitro cleavage of the donor vector by restriction enzymes (SpeI and NotI). We used puromycin selection (1 μg/mL for 2 days) to enrich positive cell colonies, and their genotype was identified using PCR and Sanger sequencing (Supplementary Fig. S1a, b). Among the 85 single cell colonies, three of them (3/85, 3.53%) were identified as positive for hACE2 insertion. Next, the verified colonies were used as donor cells for somatic cell nuclear transfer (SCNT) into three surrogates. After about four months of pregnancy, these surrogates gave birth to nine genetically modified piglets (Fig. 1b). Genomic DNA was extracted from a variety of tissues of these piglets one day after birth, and their genotype was identified by PCR. The results showed that all of the piglets were positive for hACE2 insertion (Supplementary Fig. S1a). The transcriptional level of hACE2 in several tissues of the knock-in piglets was quantified by qPCR, with GAPDH serving as the reference gene. The hACE2mRNA levels in the kidney, liver, small intestine, and lung of the knock-in pigs were significantly higher than their WT littermates. In contrast, only low mRNA levels of hACE2 were detected in the brains of knock-in pigs (Fig. 1c). To further determine whether the CRISPR/Cas9-mediated knock-in of hACE2 could increase the expression of hACE2, the protein was isolated from different organs in the WT and knock-in pigs. Western blot showed high