Genome analysis of American minks reveals link of mutations in Ras-related protein-38 gene to Moyle brown coat phenotype

Over 35 fur colours have been described in American mink (Neovison vison), only six of which have been previously linked to specific genes. Moyle fur colour belongs to a wide group of brownish colours that are highly similar to each other, which complicates selection and breeding procedures. We performed whole genome sequencing for two American minks with Moyle (m/m) and Violet (a/a m/m /p/p) phenotypes. We identified two frame-shift mutations in the gene encoding Ras-related protein-38 (RAB38), which regulates the trafficking of tyrosinase-containing vesicles to maturing melanosomes. The results highlight the role of RAB38 in the biogenesis of melanosomes and melanin and the genetic mechanism contributing to hair colour variety and intensity. These data are also useful for tracking economically valuable fur traits in mink breeding programmes.


Results
To the best of our knowledge, the present study is the first study to perform whole genome sequencing of American minks with Moyle (m/m) and Violet (a/a m/m p/p) fur colours. The resulting genome coverages were × 9 and × 40, respectively. We also used whole genome sequencing data of 3 standard dark brown (the mean genome coverage is × 8) and 3 Silverblue minks (the mean genome coverage is × 5) from our previous study 8 (Supplementary Table 1).
Mutations in the RAB38 gene. We used GATK software 9 to predict SNPs and InDels in sequenced mink genomes. We identified 13,827,261 variants across all genomes. To detect the genetic factor underlying the Moyle phenotype, we selected homozygous or compound heterozygous variants in Moyle (m/m) and Violet (a/a m/m p/p) minks that were not homozygous or compound heterozygous in standard dark brown (+/+) and Silverblue (p/p) animals. Among the selected variations, we found two homozygous mutations (one in the Moyle sample and one in the Violet sample) in Ras-related protein 38 gene (RAB38). Both mutations had a putative "HIGH" impact based on the VEP 10 prediction. We identified a homozygous 16-bp deletion [FNWR01000007.1:16075438-16075453del (RAB38:c.574-589del)], hereinafter referred to as RAB38 3del , in the Moyle sample, at the third exon of RAB38 gene. We found a homozygous 2-bp duplication [FNWR01000007.1:16132224_16132225dupCT (RAB38:c.20-21dup)], hereinafter referred to as RAB38 1dup , in the Violet sample, at the first exon of the RAB38 gene. Both mutations potentially resulted in the loss of function of RAB38 protein (Fig. 2).
We used Sanger sequencing and found that all 3 Moyle (m/m) minks were homozygous for RAB38 1dup mutations or were heterozygous for both the RAB38 3del and RAB38 1dup mutations. The single Lavender mink (a/a m/m) was also heterozygous for both mutations. Among four Violet minks (a/a m/m p/p) two were homozygous for RAB38 3del mutation, one was homozygous for RAB38 1dup and one was heterozygous for both (Table 1).
Twenty-six of the 27 standard brown or wild-type minks, and minks with other colour coats not postulated to have a Moyle allele, were homozygous wild type at the both tested mutations (Table 1), and a single mink was heterozygous for the RAB38 1dup mutation.
Allele-specific reverse transcription polymerase chain reaction (RT-PCR) was performed to confirm the chromosome location of RAB38 3del and RAB38 1dup mutations in double heterozygotes animals. Sequencing of the allele-specific cDNA amplicons encompassing exons 1-3 revealed that the RAB38 3del and RAB38 1dup mutations in double heterozygote animals were located on different chromosomes ( Supplementary Fig. 2).
Taken together, our data suggest that mutations RAB38 3del and RAB38 1dup are associated with the Moyle fur colour phenotype.

Discussion
The RAB38 gene encodes the member of the Rab small G protein family, which is involved in intracellular vesicle trafficking and melanosome biogenesis 11 . The RAB38 gene is highly expressed in melanocytes, and RAB38 protein co-localizes with end-stage melanosomes 12,13 . RAB38 participates in the transport of newly synthesized tyrosinase and Tyrp1, which are key enzymes in melanin production from the trans-Golgi network endosomes to maturing melanosomes 14,15 . Mutations in RAB38 gene were previously described in the dilution of coat colour in chocolate (cht) mice 12 and Ruby rats 13 .
The RAB38 3del mutation found in mink may result in a frame shift at the 192 protein position and lead to the loss of a stop-codon at the 212 position. A novel potential stop codon occurs only at the 277 protein position, which results in a 30% enlargement of protein size and a C-terminal end that is completely different from wild-type (Fig. 2). The mutant protein loses the C-terminal-interacting motif (amino acids 193-195), which is conserved in the RAB protein family, and seems to be involved in the interaction with Rab escort protein 16 . The mutant protein also lacks the C-terminal cysteine (position 208) within the geranylgeranylation motifs. This cysteine is the substrate for the covalent attachment of geranylgeranyl moieties, and it is highly important for RAB protein function, and present in all members of the RAB protein family 16 . Geranylgeranylated RAB proteins are inserted in a regulated manner into specific membranes where it interacts with GDI displacement factor, guanine nucleotide exchange factor and Rab effectors to control vesicular trafficking events 16,17 . We hypothesized that the mutant RAB38 protein had lower efficiency and/or specificity to interact with vesicles containing tyrosinase and Tyrp1, which would resulted in a decrease of melanin production and lead to the fur colour dilution from dark to light brown (Fig. 3). www.nature.com/scientificreports/ The RAB38 1dup mutation may result in a frame shift at the 8 protein position and lead to a premature stopcodon at the 15 position (Fig. 2).
We suggest that both of the identified mutations in RAB38 are causative for the Moyle (m/m) phenotype in the Novosibirsk mink population. Since at least two alleles Moyle (m) and Cameo (m c ) were described as autosomal recessive to standard dark brown (+) 3 it can be hypothesized that different allelic mutations in RAB38 gene may underlie the phenotype variabilities linked to the same locus. This suggestion has yet to be tested by direct RAB38 gene genotyping of Cameo animals. www.nature.com/scientificreports/  www.nature.com/scientificreports/ Our study adds the RAB38 gene to the list of genes with identified mutations which were associated with different mink fur colour phenotypes 4-8 and provides valuable data that contribute to improving global mink fur production via selective breeding programs.

Methods
All methods were carried out in accordance with relevant guidelines and regulations for laboratory work. The local Ethics Committee of the Institute of Cytology and Genetics, Siberian Branch of the Russian Academy of Sciences, approved the study protocols.
Moyle (m/m 3 individuals), lavender (a/a m/m 1 individuals), violet (a/a m/m p/p 4 individuals), Silverblue (p/p, 2 individuals), royal pastel (b/b 2 individuals), shadow silverblue (S H /+ p/p 1 individual) and standard dark brown (+/+ 17 individuals) farm-bred American minks were maintained in the Experimental Fur Farm of the Institute of Cytology and Genetics (Novosibirsk mink population). The collected tissues (tail snips) were rapidly dissected, frozen in liquid nitrogen, and stored at − 70 °C until DNA and RNA extraction.
Sample collections from farm-bred American minks of royal pastel (b/b 4 individuals) and Cross sapphire (S/+ a/a p/p 1 individual) coat colour from the «Mermeriny» fur farm, Tver region, Russia (Tver mink population) were used in this study. The Tver mink population is unrelated to the Novosibirsk population.
Genomic DNA from mink tissues was extracted using QIAGEN Mini Spin Columns, following the manufacturer's protocol (QIAGEN, Germany). Library preparations (1 Moyle (m/m) and 1 violet (a/a m/m p/p) were performed using the TruSeq PCR Free Kit (Illumina, USA), following the manufacturer's protocol. Library validation was performed using an Agilent 2,100 Bioanalyzer with DNA High Sensitivity chip (Agilent, USA) and quantified via qPCR using a KAPA Library Quantification Illumina Kit protocol (KAPA Biosystems, USA). Paired-end libraries were sequenced in 2 × 101 cycles using the Illumina TruSeq SBS v3 kit (Illumina) on a HiSeq 2,000/2,500 sequencer (Illumina) at the Vavilov Institute of General Genetics RAS (Moscow, Russia) and 2 × 151 cycles using the Illumina NovaSeq S4 kit (Illumina) on a NovaSeq 6,000 sequencer at the Genetico Company (Moscow, Russia).
We also used whole genome sequencing data of 3 standard dark brown Silverblue minks from our previous study 8 .
Genetic variants in the sequenced mink genomes were predicted using HaplotypeCaller (with default arguments) from the Genome Analysis Toolkit (GATK) package version 4.0 9 .
To detect the genetic factor underlying the Moyle phenotype, we selected homozygous or compound heterozygous variants with a depth of coverage greater than 2 in m/m and a/a m/m p/p minks that were not homozygous or compound heterozygous in the standard dark brown wild-type and Silverblue animals.
Annotation and effect prediction of selected variants were performed in VEP 10 using the American mink genome annotation (Ensembl v97).
We performed Sanger sequencing to validate mutations. Primers for PCR amplification were designed in Primer3 software (Supplementary Table 2), and PCR was performed using the GenPack PCR Core (Isogen, Russia). Resultant amplicons were cleaned using a Cleanup Standard Kit (Evrogen, Russia) and processed using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, USA), following the manufacturers' protocols. Probes were purified using a DyeEx 2.0 Spin Kit (QIAGEN) and sequenced in a 3730xl DNA Analyzer (Applied Biosystems).
To identify the chromosome location of RAB38 3del and RAB38 1dup mutations in double heterozygotes animals, we used allele-specific RT-PCR ( Supplementary Fig. 1). Total RNA was extracted from tissues using RNeasy Mini Spin Columns, following the manufacturer's protocol (QIAGEN). Extracted RNA was treated with RNase-Free DNase I (Thermo Scientific, USA) and assayed for quantity and quality in a NanoDrop One-C (Thermo Scientific). All RNA samples were kept at − 80 °C. First-strand cDNA synthesis was performed using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems). The first amplification was performed using universal cDNA RAB38 ex 1-3 F and cDNA RAB38 ex 1-3 R primers. The resulting PCR products were used for the second amplification with (1) universal cDNA RAB38 ex 1-3 F and cDNA RAB38 ex 1-3 R primers, (2) wt-specific primers cDNA RAB38 ex 1-3 F and cDNA RAB38 ex 1-3 wt R, and (3) del-specific primers cDNA RAB38 ex 1-3 F and cDNA RAB38 ex 1-3 del R. Final PCR products were cleaned up using a Cleanup Standard Kit (Evrogen) and processed using the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems), following the manufacturers' protocols. Probes were purified using a DyeEx 2.0 Spin Kit (QIAGEN) and sequenced in a 3730xl DNA Analyzer (Applied Biosystems).

Data availability
The datasets generated during the current study were deposited into NCBI SRA database and can be accessed with the BioProject accession number PRJNA660737 (https ://www.ncbi.nlm.nih.gov/sra/PRJNA 66073 7).