Pig Coat Color Manipulation by MC1R Gene Editing

Black coat color in pigs is determined by the dominant E allele at the MC1R locus. Through comparing MC1R gene sequences between recessive e and dominant ED1 alleles, we identified four missense mutations that could affect MC1R protein function for eumelanin synthesis. With the aim of devising a genetic modification method for pig coat color manipulation, we mutated the e allele in the Duroc breed to the dominant ED1 allele using CRISPR-mediated homologous recombination for the four mutation substitutions at the MC1R locus. The MC1R-modified Duroc pigs generated using the allele replacement strategy displayed uniform black coat color across the body. A genotyping assay showed that the MC1R-modified Duroc pigs had a heterozygous ED1/e allele at the MC1R locus; in addition, the pigs remained in the Duroc genetic background. Our work offers a gene editing method for pig coat color manipulation, which could value the culture of new pig varieties meeting the needs of diversified market.


Introduction
Domestic pigs have coat color diversity. The European domestic pig breeds usually have a white or red coat color, whereas most Chinese indigenous breeds have a black coat color. The differential coat color phenotypes are also a reflection of the different domestication origins between European and Asian pig breeds [1]. The domestication strategy of Chinese indigenous pig breeds confers a high fat accumulation in pigs, which improves the pork flavor [2][3][4]. Despite the smaller body size and weight as well as less meat production, the higher meat quality of Chinese indigenous breeds usually brings a better market value than the European lean breeds. Current Chinese customers favor consumption of meat of indigenous pigs that have a common representative trait of a black coat color.
Many genes have been identified that are involved in mammalian coat color development [5,6]. Among them, melanocortin receptor 1 (MC1R) encoded by Extension (E) coat color locus is a well-known regulator of melanogenesis, controlling the switch between two basic types of melanin: eumelanin (black/brown) and phaeomelanin (yellow/red) [7]. The active form of MC1R promotes eumelanin synthesis, producing a uniform black coat color, whereas inactive MC1R is associated with the eumelanin to phaeomelanin switch, resulting in red or yellow coat color [7,8]. In pigs, four MC1R variants corresponding to five distinct E alleles have been found to be associated with different coat color phenotypes ranging from recessive red to dominant black [9]. The E allele for black coat color is dominant over the e allele for red coat color [9]. The coat color genetics provides a molecular method to regulate coat color in mammals. One can readily alter coat colors of animals by using or disrupting the alleles controlling specific coat colors [10,11]. In this respect, black pigs can common variations in Chinese indigenous black pigs [12]. V95M and L102P substitutio in the E D1 allele may contribute to the active function of MC1R, thus accounting for dominant black phenotype [13]. A164V and A243T are the specific missense mutations the Duroc breed. A243T is believed to substantially disrupt MC1R function, resulting i red coat color phenotype of Duroc pigs [9,13].
We set out to mutate all four missense SNPs in the MC1R coding region of Du pigs to the corresponding nucleotides in MC1R of Chinese indigenous black pigs ( Figu  1A,B). We used a CRISPR-mediated homologous recombination (HR) approach to repla the Duroc MC1R allele with the donor plasmid harboring the four indicated SNPs (Figu 1B). As the black pig MC1R E D1 allele is dominant over the Duroc recessive e allele, bo heterozygous E D1 /e and homozygous E D1 /E D1 genotypes can realize a black coat color p notype. Through CRISPR and donor co-transfection into Duroc fetal fibroblasts, drug lection and enrichment, and cell colony genotyping, we obtained 40 cell colonies with p designed substitutions in all four SNPs (with only one e allele being replaced with allele of a homologous donor in most colonies), but no homozygous mutants at all fo SNPs were screened out. The HR efficiency of precise replacement was 14.3% (40/28 This result indicates HR remains a low-efficiency event in primary fibroblast, even w the aid of CRISPR. Previous reports showed that CRISPR-mediated HR occurred in efficiency range from 0 to 41.2% in pig fetal fibroblast with a very low rate in most cas far less than the CRISPR-mediated knockout efficiency [14][15][16]. of the MC1R gene distinct between Duroc breed (e) and Chinese indigenous black breed (E D1 ) are shown in red. Landrace and Large White breeds have a specific D124N substitution compared with other breeds, which is the key variation forming the E P allele. The E allele associated with each MC1R variant is indicated in parentheses. (B) CRISPR system and donor design for SNP substitutions in Duroc MC1R. A guide RNA (gRNA) was designed to recognize the cleavage target in the SNP regions. A 2757-bp DNA sequence was amplified from the MC1R genomic region in Chinese indigenous black pigs as the donor to replace the Duroc MC1R homologous region by HR. The gRNA-recognized sequence in the donor was synonymously mutated to avoid second CRISPR cleavage after HR. The genotyping primers amplifying a 1627-bp sequence covering homology arm and adjacent genome region after HR were used in PCR assay to detect the occurrence of gene replacement. AA, amino acid position.
We set out to mutate all four missense SNPs in the MC1R coding region of Duroc pigs to the corresponding nucleotides in MC1R of Chinese indigenous black pigs ( Figure 1A,B). We used a CRISPR-mediated homologous recombination (HR) approach to replace the Duroc MC1R allele with the donor plasmid harboring the four indicated SNPs ( Figure 1B). As the black pig MC1R E D1 allele is dominant over the Duroc recessive e allele, both heterozygous E D1 /e and homozygous E D1 /E D1 genotypes can realize a black coat color phenotype. Through CRISPR and donor co-transfection into Duroc fetal fibroblasts, drug selection and enrichment, and cell colony genotyping, we obtained 40 cell colonies with predesigned substitutions in all four SNPs (with only one e allele being replaced with E D1 allele of a homologous donor in most colonies), but no homozygous mutants at all four SNPs were screened out. The HR efficiency of precise replacement was 14.3% (40/280). This result indicates HR remains a low-efficiency event in primary fibroblast, even with the aid of CRISPR. Previous reports showed that CRISPR-mediated HR occurred in an efficiency range from 0 to 41.2% in pig fetal fibroblast with a very low rate in most cases, far less than the CRISPR-mediated knockout efficiency [14][15][16].
The positive fibroblast colonies were used as the nuclear donor for somatic cell nuclear transfer (SCNT). We totally constructed 1938 MC1R-modified embryos from 3 cell colonies and transferred them to 9 surrogates. Five surrogates were pregnant and maintained to term, and finally gave birth to 30 piglets on day 116 of gestation. Among them, 19 cloned piglets were born healthy and developed normally ( Table 1). Analysis of the coat colors found that 17 were uniform solid dark black, 2 were light black, and 11 were red (Figure 2A,B). Recipients 19-937900, 20-694608, and 20-006896 received embryos derived from only one colony (#2-24) but gave birth to cloned pigs with different coat colors, implying that the colony contained a minor part of cells with unexpected genotypes. The MC1R genes of all cloned piglets were sequenced to investigate the relationship between MC1R genotypes and coat colors. The E alleles after HR were genotyped by combinations of three key SNPs (V95M, L102P, and T243A) affecting MC1R activity. The A164V mutation in the Duroc breed is observed in the functionally normal horse MC1R, thus has no effect on MC1R activity [9]. We found that all dark black pigs harbored at least one E D1 allele with all four SNP substituted. Of the 17 dark black pigs, 8 had three SNPs substituted (V95M, L102P, and T243A) on one allele and one SNP substituted (V164A) on both alleles, and 9 had all four SNPs substituted on only one allele ( Figure 2C and Table 2). Because AA164 is the closest SNP to the CRISPR-cleavage site (approximately 80 bp) among the four SNPs, AA164 has a much high HR rate than the other three SNPs. Both genotypes in dark black pigs after gene editing can be classified to E D1 /e. The result clearly indicated that the E D1 allele is dominant over the e allele, and presence of only one E D1 allele in the Duroc genetic background can confer a uniform black coat color phenotype. In contrast, the cloned pigs with wild type red coat color all had none of substitutions in those SNP sites. Some red pigs harbored homozygous or heterozygous insertion or deletion (Indel) in the gRNA region that clearly had a knockout effect on the MC1R gene ( Table 2). The same red coat color formed by either MC1R e allele expression or MC1R knockout suggests that the Duroc e allele is totally inactive in eumelanin synthesis.
It is worth noting that two cloned pigs (#8 and #27) displayed a light black coat color ( Figures 2B and 3A). MC1R genotyping showed that they both had at least one e allele replaced with an E D1 allele with all four SNP substitutions. For the other e allele, #8 had two substituted SNPs (V164A and T243A, these mutations formed an E + allele) and #27 had no substitutions. They both had an intact E D1 allele (Table 2 and Figure 3C). However, their black coat color seemed not completely dominant. We sequenced the whole MC1R coding and flanking non-coding regions of the two pigs and detected no extra mutations. We then isolated the mRNA from ear skin and measured the MC1R expression status. A qPCR analysis of MC1R mRNA level detected that #8 and #27 had a similar MC1R expression level with the wild-type red Durocs, whereas the dark black Durocs harboring the E D1 allele had 3-5 times higher MC1R expression compared with the wild-type red Durocs ( Figure 3B). These results imply that full dominance of E D1 depends on its higher expression than that of e. An intact E D1 but without sufficient expression level could not support a completely dominant black coat color. We further sequenced the MC1R cDNA of cloned pigs and found the light black pigs had an equal expression level for E D1 and e, but the dark black pigs had a major expression for only E D1 , although they are heterozygous for E D1 and e in genotypes ( Figure 3C). For the reason why the E D1 are not expressed sufficiently in the two light black pigs, we suspected that the recombinant E D1 allele has distal mutations or structural variations caused by the HR process, which impaired E D1 expression. The detailed mechanism remains to be investigated further. To investigate the genetic background of the clone piglets, we carried out an admixture analysis to estimate the proportion of ancestries using genome-wide SNP data of cloned pigs. The results showed that the cloned pigs with the dark black coat color had a nearly 100% Duroc genetic structure ( Figure 4).
The solid black hair across the body of cloned pigs demonstrates that only one E D1 allele introduction can fully confer the black coat color trait into the Duroc genetic background, further confirming that the red coat color in Duroc is a recessive trait compared to the black coat color. In this regard, transgenic expression of foreign MC1R with the E D1 genotype can also achieve such coat color switching purposes. However, the transgenesis strategy could bring an inconsistent gene expression level because of the different transgene copies and integrated locations in the host genome, which may influence the coat color switching effect in transgenic pigs [17,18]. This point has been demonstrated in the light black pigs generated in our work. Although an intact E D1 is incorporated into the genome in the light black pigs, a low expression level of E D1 still cannot support a fully dominant black coat color. In addition, the transgenesis strategy is disadvantageous for subsequent pig breeding programs to culture the blackish lean pig breeds, as the transgene integration cannot be controlled to keep an identical genotype in different cells and pigs used for transgenesis [17,18]. Our HR-mediated point mutation strategy facilitates breeding of a new pig variety with black coat color and high meat yield.    Considering that multiple genetic loci can control coat color trait, the present MC1R manipulation strategy may not work well in some pig breeds with complex coat colors or patterns. For example, many European white pigs such as Large White and Landrace harbor the dominant white (I) locus, the molecular basis of which is duplication of the KIT gene [19]. Dominant white can mask other coat colors by allele dominance. The dominance relationship among these coat color gene loci can be reflected in pig crossbreeding practice. Black pig and Duroc crossbred usually has the uniform black coat color, implying a "black dominance" hereditary pattern in the Duroc genetic background. In contrast, Black pig and European white pig crossbred displays large variations in coat colors [20]. Therefore, MC1R manipulation in the presence of the dominant white locus cannot cause a full switch of coat color to black. For coat color manipulation in white pigs, inactivation of the dominant white locus is required prior to conferring other desired coat colors by specific gene mutations.
In conclusion, we here present a gene editing approach to realize the coat color switch without affecting gross phenotypes in Duroc or other red pig breeds. This approach can be used to decipher animal coat color genetics precisely and cultivate new animal breeds with desired coat color meeting specific demands in a cost-effective and timesaving way. To investigate the genetic background of the clone piglets, we carried out an admixture analysis to estimate the proportion of ancestries using genome-wide SNP data of cloned pigs. The results showed that the cloned pigs with the dark black coat color had a nearly 100% Duroc genetic structure (Figure 4). The solid black hair across the body of cloned pigs demonstrates that only on allele introduction can fully confer the black coat color trait into the Duroc genetic ground, further confirming that the red coat color in Duroc is a recessive trait comp to the black coat color. In this regard, transgenic expression of foreign MC1R with th genotype can also achieve such coat color switching purposes. However, the transge strategy could bring an inconsistent gene expression level because of the diff transgene copies and integrated locations in the host genome, which may influenc coat color switching effect in transgenic pigs [17,18]. This point has been demonstra the light black pigs generated in our work. Although an intact E D1 is incorporated in genome in the light black pigs, a low expression level of E D1 still cannot support a dominant black coat color. In addition, the transgenesis strategy is disadvantageou subsequent pig breeding programs to culture the blackish lean pig breeds, a transgene integration cannot be controlled to keep an identical genotype in differen and pigs used for transgenesis [17,18]. Our HR-mediated point mutation strategy tates breeding of a new pig variety with black coat color and high meat yield.
Considering that multiple genetic loci can control coat color trait, the present M manipulation strategy may not work well in some pig breeds with complex coat col patterns. For example, many European white pigs such as Large White and Landrac bor the dominant white (I) locus, the molecular basis of which is duplication of the KIT [19]. Dominant white can mask other coat colors by allele dominance. The dominance tionship among these coat color gene loci can be reflected in pig crossbreeding pra Black pig and Duroc crossbred usually has the uniform black coat color, implying a " dominance" hereditary pattern in the Duroc genetic background. In contrast, Blac and European white pig crossbred displays large variations in coat colors [20]. Ther MC1R manipulation in the presence of the dominant white locus cannot cause a full s of coat color to black. For coat color manipulation in white pigs, inactivation of the nant white locus is required prior to conferring other desired coat colors by specific mutations.
In conclusion, we here present a gene editing approach to realize the coat color s without affecting gross phenotypes in Duroc or other red pig breeds. This approac be used to decipher animal coat color genetics precisely and cultivate new animal b with desired coat color meeting specific demands in a cost-effective and timesaving

MC1R Gene Editing Strategy
To introduce an active MC1R into the Duroc genome, we devised an CRISPR/Cas9mediated HR strategy to precisely replace the Duroc inactive MC1R allele (e) with the active MC1R allele (E D1 ) from Chinese indigenous black pigs, with all four missense SNP substitutions ( Figure 1B). A gRNA (GGTGTCCAGCCTCTGCTTCC with a TGG PAM site) mediating CRISPR cleavage in the SNP region was designed and ligated into U6-sgRNA cloning vector, as reported previously [21]. The homologous donor was a plasmid containing 2757 bp MC1R genome sequence of the E D1 allele covering all four SNPs with approximately 1000 bp flanking homology arms. The gRNA recognized sequence in the donor was mutated to GGTGTCCtctCTgTGtTTtCTcG (mutations are shown with lowercase letters) to avoid CRISPR cleavage at the recombinant MC1R locus. (The complete homologous donor sequence was shown in Supplemental File S1.) For HR manipulation, MC1R donor plasmid, MC1R gRNA plasmid, and hCas9 plasmid (Addgene # 41815) were co-transfected into Duroc fetal fibroblasts to screen gene-recombinant cells.

MC1R-Modified Fibroblast Screening
The fetal fibroblasts isolated from 30-day Duroc fetuses were primarily cultured and transfected with the CRISPR HR system (three plasmids as previously noted) by nucleofection (Lonza, Basel, Switzerland). The nucleofected cells were seeded in a low density in 10 cm culture dishes (approximately 500 cells/dish). When cells were attached, the medium was replaced with screening medium (high glucose DMEM containing 15% FBS and 500 µg/mL G418 (Thermo Fisher Scientific, Waltham, MA, USA)) for 10 days, with the screening medium refreshed 2-3 times. Following 10 days of screening, the single-cell colonies with compact morphology and suitable size were picked for separate culture in 48-well plates. When cells reached confluence, 1/10 cells in each well were collected for genotyping. The other cells were passaged to 24-well plates. The positive colonies with desired modified MC1R alleles were further cultured in 12-well plates and cryopreserved for future use.

Genotyping of MC1R-Modified Fibroblasts
The cell colonies were lysed with proteinase K at 56 • C for 30 min and then the proteinase K was inactivated at 95 • C for 5 min. The cell lysate was directly used as a PCR template to amplify the MC1R region using the following primer pair, MC1Rgenotyping-forward: CCTGCACTCGCCCATGTACTACT, and MC1R-genotyping-reverse: GCCAGAAAGAGGTTGACGTT. The primer location in the MC1R genomic region is shown in Figure 1B. The PCR product was sequenced using the MC1R-genotyping-forward primer. The positive colonies should have homozygous or heterozygous mutation in the four indicated SNPs and no Indel in the gRNA-recognized region.

Pig Cloning
The MC1R-modified pigs were generated by SCNT using the positive colonies as donor cells as previously described [21,22]. The reconstructed embryos were transferred into the uterus of estrus-synchronized sow recipients. Each recipient received approximately 200 embryos. The pregnancy of surrogates was monitored monthly by ultrasound, and the cloned piglets were delivered by natural birth.

Genotyping Assay of MC1R-Modified Duroc Pigs
The genomic DNA of cloned pigs were extracted from ear biopsies, and MC1R genotypes were PCR-identified using the same method described in Section 3.3. The E alleles were determined by the combinations of four SNP genotypes of the MC1R gene. To further confirm the origin breed of the cloned pigs, we used genome-wide SNP data for breed tracing [23,24]. Two dark black cloned piglets for test and 66 individuals from four commercial breeds (Pietrain, Duroc, Landrace, and Large White) as reference were sampled and genotyped with PorcineWENS55K SNP Array containing 56,380 SNPs across the whole genome (WENS foodstuff Group). The proportion of ancestries per pig was assessed using the ADMIXTURE 1.22 program and genome-wide SNP data to trace the breed of origin. The number of ancestries (K) was set at 4 [23].

MC1R mRNA Expression in Cloned Pigs
The ear skins of cloned pigs were lysed with TRIzol Reagent (Thermo Fisher Scientific) for total RNA extraction. The total RNA was subjected to reverse transcription with Hifair V one-step RT-gDNA digestion SuperMix (Yeasen Biotechnology, Shanghai, China) to generate cDNA. Real-time qPCR was performed with ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China) in a QuantStudio 5 Real-Time PCR System (Thermo Fisher Scientific) using the cDNA as templates and the primer pair, MC1R-qF, GCGTCTTCAA-GAACGTCAAC, and MC1R-qR, ATGGAGTTGCAGATGACGAG, to determine the MC1R mRNA expression level. Each sample was run in triplicate, and MC1R expression levels were normalized to GAPDH. We also amplified the complete mRNA coding region using cDNA as templates for Sanger sequencing to determine the allele-specific expression level of the MC1R gene of cloned pigs.