Variation in the Caprine Keratin-Associated Protein 27-1 Gene is Associated with Cashmere Fiber Diameter

Variation in some caprine keratin-associated protein (KAP) genes has been associated with cashmere fiber traits, but many KAP genes remain unidentified in goats. In this study, we confirm the identification of a KAP27-1 gene (KRTAP27-1) and describe its effect on cashmere traits in 248 Longdong cashmere goats. A polymerase chain reaction–single strand conformation polymorphism (PCR-SSCP) analysis was used to screen for sequence variation in this gene, and three sequence variants (named A to C) were found. These sequences have the highest similarity (77% identity) to a human KRTAP27-1 sequence, while sharing some homology with a predicted caprine KRTAP27-1 sequence ENSCHIG00000023347 in the goat genome construct (ARS1:CM004562.1) at chromosome 1 position 3,966,193–3,973,677 in the forward strand. There were two single nucleotide polymorphisms (SNPs) detected in the coding sequence, including one nonsynonymous SNP (c.413C/T; p.Ala138Val) and one synonymous SNP (c.495C/T). The C variant differed from A and B at c.413C/T, having cytosine in its nucleotide sequence, while the B variant differed from A and C at c.495C/T, having thymine in its nucleotide sequence. Goats of the genotypes AB and BB produced cashmere fibers of higher mean fiber diameter (MFD) than goats of genotype AA, but no difference in MFD was detected between the AB and BB goats. These results suggest that B is associated with increased MFD. Expression of the caprine KRTAP27-1 sequence was predominantly detected in the skin tissue of goats but not or only weakly detected in other tissues, including longissimus dorsi muscle, heart, kidney, liver, lung and spleen.


Introduction
Cashmere fiber is produced by secondary hair follicles. The fiber has the characteristics of being light, soft, strong, elastic and a good thermal insulator, and it is accordingly referred to as 'soft gold', as it plays an important role in the textile industries [1]. The price paid for cashmere fibers is typically higher than that for mohair or wool. With cashmere fiber yield, curly fiber length and mean fiber diameter (MFD) are the most important traits in determining quality.
Keratin-associated proteins (KAPs) and keratins are the major components of cashmere fibers. The KAPs form a matrix and cross-link with cysteine residues on the keratins and other KAPs, through the formation of disulfide bonds. They typically possess a high content of either cysteine or glycine and tyrosine, and they have historically been classified into three broad groups based on their amino acid composition: the high sulfur (HS) group with less than 30 mol% cysteine, the ultrahigh sulfur (UHS) group with more than 30 mol% cysteine and the high glycine/tyrosine (HGT) group containing 35 to 60 mol% glycine and tyrosine [2].
In this study, we report the identification of intronless KRTAP27-1 sequences amplified directly from goat DNA and describe variation in these sequences and the locations of expression of this gene; we then describe the association of this gene with cashmere fiber traits in Longdong cashmere goats.

Cashmere Goats and Sample Collection
Two hundred forty-eight Longdong cashmere goats reared at the Yusheng Cashmere Goat Breeding Company in Huan County of the Gansu Province of China were investigated. The 248 Longdong cashmere goats were the progeny of eight unrelated sires. These goats included 196 females and 52 males. At twelve months of age (spring time), cashmere fibers were collected from individual goats (live weight 21.7 ± 0.92 kg) that were in the anagen phase of fiber growth by combing, and the raw weights of the collected secondary or cashmere fibers from individual goats were measured. A fiber sample from the mid-side region of each goat was taken for testing MFD and curly fiber length. The MFD and curly fiber length of each goat were measured using the Optical Fiber Length and Diameter Analyzer OFDA4000 (EPCO, Shanghai, China).
Blood samples from these goats were collected on Munktell TFN paper (Munktell Filter AB, Falun, Sweden). Genomic DNA for polymerase chain reaction (PCR) amplification was purified from 1.2-mm punches of the dry blood spots, following the procedure described by Zhou et al. [14].
Three 3-year-old female Longdong cashmere goats were slaughtered to collect tissue samples for reverse-transcription PCR (RT-PCR) analysis, including skin, longissimus dorsi muscle, heart, kidneys, liver, lung and spleen tissue samples. These samples were immediately frozen in liquid nitrogen upon collection.
The collection of caprine blood and skin samples were approved by Gansu Agricultural University, and the feeding of these goats was in accordance with the guidelines for care of experimental animals established by the Ministry of Science and Technology of the People's Republic of China (Approval number 2006-398).

Polymerase Chain Reaction-Single Strand Conformation Polymorphism (SSCP) Analysis of Caprine KRTAP27-1
The coding sequence of human KRTAP27-1 (GenBank sequence AB096937) was used to BLAST search the Caprine Genome Assembly ASM170441v1. The sequence that shared the greatest similarity with the human KRTAP27-1 sequence was assumed to be the caprine KRTAP27-1 sequence. Based on this caprine sequence, two PCR primers (5 -ACTCAATCAAGGATTTTCAC-3 and 5 -GCAGATTCTTCACTCAGTAGG-3 ) were designed to amplify the entire coding region of the putative caprine KRTAP27-1. These primers were synthesized by the Sangon Biotechnology Company Limited (Shanghai, China).
Amplifications were performed in a 20-µL reaction containing the genomic DNA purified from a 1.2-mm punch of dried blood, 0.25 µM of each primer, 1 × PCR buffer (supplied with DNA polymerase enzyme), 150 µM of each deoxyribonucleoside triphosphate (dNTP) (Takara, Dalian, China), 2.5 mM Mg 2+ and 0.5 U of Taq DNA polymerase (Takara, Dalian, China). The PCR amplification profile was comprised of an initial denaturation at 94 • C for 5 min, followed by 35 cycles of 94 • C for 30 s, 60 • C for 30 s and 72 • C for 30 s, and then a final extension for 5 min at 72 • C. The thermal cycling was performed in Bio-Rad S1000 thermal cyclers (Bio-Rad, Hercules, CA, USA), and the presence of PCR products was confirmed and qualitatively assessed using 1.0% agarose gel electrophoresis in 1 × TBE buffer.

Sequencing of KRTAP27-1 Variants and Sequence Analyses
After PCR-SSCP analysis, amplicons from goats that appeared to be homozygous based on the observed PCR-SSCP banding patterns were sequenced directly, and in both directions, at the Beijing Genomics Institute (Beijing, China). Variants that were only found in a heterozygote form were sequenced using the method described by Gong et al. [16].

Reverse-Transcription Polymerase Chain Reaction (RT-PCR) Analysis
Total RNA from the tissue samples was extracted using Trizol reagent (Invitrogen, Carlsbad, CA, USA). The quality and concentration of RNA were ascertained using 2% agarose gel electrophoresis and UV spectrophotometry, respectively. Complementary DNA (cDNA) was synthesized using the PrimeScript RT Reagent Kit with gDNA Eraser (Perfect Real Time, Takara, Dalian, China) according to the manufacturer's instructions and then amplified using a set of inner primers (5 -TCGGCAGTGTTGGAGTGTGT-3 and 5 -GGACAAGACTTCGCCATGTTGC-3 ) that were designed to amplify a smaller portion (110 bp) of the putative caprine KRTAP27-1. The amplification conditions and thermal profile were the same as for the genomic amplifications described above, but the genomic DNA template was replaced with a 0.8-µL aliquot of the cDNA. The RT-PCR products were visualized using 1.0% agarose gel electrophoresis. The caprine β-actin gene (ACTB) was used as a positive control reference gene and amplified using the primers 5 -AGCCTTCCTTCCTGGGCATGGA-3 and 5 -GGACAGCACCGTGTTGGCGTAGA-3 .

Statistical Analyses
All data were analyzed using IBM SPSS Statistics version 24.0 (IBM, NY, USA) software. Pearson correlation coefficients between MFD, cashmere yield and curly fiber length were determined.
The effect of variation in KRTAP27-1 on the three cashmere traits was assessed using general linear mixed-effect models (GLMMs), but only those genotypes that occurred at a frequency of over 5% were analyzed. Sire (n = 8) and gender were found to affect all the fiber traits (p < 0.05), and they were therefore included in the GLMMs as a random factor and a fixed factor, respectively. Birth rank was not found to affect these traits, and thus it was not included in the models. Only main effects were tested. A Bonferroni correction was applied to reduce the probability of false positive results during the multiple comparisons in these models. Unless indicated, all p values were considered statistically significant when p < 0.05.

Figure 1.
KRTAPs identified on goat chromosome 1. Eleven identified KRTAPs and the newly identified KRTAP27-1 (boxed) are located on goat chromosome 1. Vertical bars and arrows represent KRTAPs and the direction of transcription, respectively, with the names of the genes being indicated below (i.e., 11-1 represents KRTAP11-1). The distances between these genes are approximate and refer to the caprine chromosome 1 sequence (NC_030808.1).

Sequence Variation in Caprine KRTAP27-1
Three different PCR-SSCP banding patterns representing three variants (named A, B and C) were identified in the Longdong cashmere goats investigated (Figure 2).
Sequencing of PCR amplicons representing the three different banding patterns revealed three unique nucleotide sequences (A, B and C), with sequence A being identical to the Caprine Genome Assembly (NC_030808.1). Two single nucleotide polymorphisms (SNPs) were found, with one (c.495C/T) being synonymous and the other (c.413C/T) being nonsynonymous, that would lead to a putative amino acid change of p.Ala138Val (Figure 3). A reverse complementary Chi sequence and a Chi-like sequence were detected in these variants. These variant sequences were deposited into GenBank and allocated the Accession Numbers MN934937, MN934938 and MN934939. In the 248 Longdong cashmere goats investigated, variants A, B and C were found at frequencies of 71.37%, 23.79% and 4.84%, respectively. Five genotypes (AA, BB, AB, AC and BC) were found, with frequencies of 52.02%, 5.65%, 32.66%, 6.05% and 3.63%, respectively.

Phylogenetic Analysis of Caprine KAP27-1
A phylogenetic analysis of the caprine KRTAP27-1 sequences, along with all of the other HS-KRTAPs identified in identified in goats, sheep and humans, revealed that the three caprine KRTAP27-1 sequences identified were most closely related to human KRTAP27-1, and less closely related to all of the other HS-KRTAPs ( Figure 4). This suggests the three sequences identified represent sequence variants of caprine KRTAP27-1, and thus they were named CAPHI-KRTAP27-1*A, CAPHI-KRTAP27-1*B and CAPHI-KRTAP27-1*C, based on the nomenclature proposed by Gong et al. [17].

Expression of Caprine KRTAP27-1
The RT-PCR analysis of seven different tissues from the Longdong cashmere goats revealed that caprine KRTAP27-1 expression was predominantly detected in the skin tissue but was not detected or only weakly detected in the longissimus dorsi muscle, heart, kidney, liver, lung and spleen tissues ( Figure 5).

Correlation between Cashmere Fiber Traits
A moderate correlation was found between curly fiber length and cashmere fiber yield (r = 0.496; p < 0.001), while MFD had a weak correlation with cashmere fiber yield (r = 0.252; p < 0.001) and curly fiber length (r = 0.209; p < 0.001).

Discussion
This study confirms the identity of a caprine KAP gene on goat chromosome 1, which putatively encodes a HS-KAP protein that is most similar to the human KAP27-1 and the predicted caprine KRTAP27-1 sequences in Ensembl and GenBank. However, the Ensembl-predicted caprine KRTAP27-1 sequence ENSCHIG00000023347 located in the goat genome construct (ARS1:CM004562.1) at chromosome 1 position 3,966,193-3,973,677 contains two introns, with this not having been observed in the human KRTAP27-1 sequence or any of the other caprine or ovine KRTAP sequences reported to date, including caprine KRTAP8-2 [8], KRTAP15-1 [5], KRTAP20-1 [9], KRTAP20-2 [10], KRTAP24-1 [6] and KRTAP28-1 [7] or ovine KRTAP26-1 [11] and KRTAP22-1 [12]. This would suggest the construct may contain assembly errors. Regions that are highly polymorphic and repetitive, such as the KRTAP genes, are particularly difficult to assemble [18], and the analysis of ARS1 by Li et al. [19] revealed that some sequences containing essential genes were missing from ARS1, despite their presence in the sequenced ARS1 individuals. This suggests that the long-read sequencing technologies or the assembling methods used to assemble ARS1 still have limitations that warrant further improvements.
In contrast to other HS-KAPs, this protein contains a high level of serine and moderate level of cysteine. Having a higher level of serine than cysteine has been described previously for a number of HS-KAPs, including caprine KAP15-1 [5], caprine KAP24-1 [6], ovine KAP13-3 [20], ovine KAP24-1 [21] and ovine KAP26-1 [11]. The precise role of KAP serine residues in animal fiber is unknown, but these serine residues may possibly be phosphorylated [22], especially as the phosphorylation of keratins can affect keratin assembly and organization [23]. It is therefore possible that serine may play a role in regulating interactions between KAPs and keratins and consequently affect the fiber structure.
The specific expression of caprine KRTAP27-1 in skin is consistent with the expression pattern of other KRTAPs described in goats, such as KRTAP15-1 [5], KRTAP20-1 [9] and KRTAP20-2 [10], but it contrasts with the expression pattern for human KRTAP27-1 [24], with expression of the gene apparently not found in all of the tissues taken from the Caucasians investigated, including skin and hair follicles. Accordingly, Rogers et al. [24] suggest the possibility of individual-or population-specific expression of this gene.
The detection of two SNPs in the 603-bp KRTAP27-1 coding sequence amplified gives a density of approximately 3.3 SNPs per Kb. This density is lower than a previously reported average density of 4.9 SNPs per Kb across the entire ovine genome [25], with this ovine density estimate based not only just on coding sequence but also on the more plentiful and arguably less functionally constrained noncoding parts of the genome. This stated, ovine KRTAPs do appear to have higher SNP densities that other parts of the sheep genome, with a density of over 20 SNPs/Kb having been reported for ovine KRTAP1-3 [26] and KRTAP1-4 [27]. This may be because there is evidence to suggest that gene conversion or nonreciprocal genetic exchange contributes to the accumulation of sequence variation in the KRTAPs [5,6,28,29], and these recombination events may be facilitated by the Chi (5 -GCTGGTGG-3 ) and Chi-like sequences within the genes [5,6,28,30]. In this respect, the presence of Chi and Chi-like sequences in KRTAP27-1 suggests recombination mechanisms may be creating sequence diversity in this gene as well.
Variants A and B differed by one synonymous SNP, c.495C/T; hence, the finding that these variants differentially affected MFD is notable. While synonymous SNPs do not lead to amino acid changes, they have been reported to affect the stability and secondary structure of mRNA, and hence they can have functional significance [31,32]. The possibility also exists that the effect detected may be due to other unidentified but functionally important SNPs located in proximity to the amplified region. For example, given that KRTAPs are clustered, it is possible that the association with MFD detected for KRTAP27-1 may be due to linkage with other nearby KRTAPs. This is supported by the findings of similar phenotypic effects for KRTAP15-1 [5], KRTAP24-1 [6] and KRTAP28-1 [7], which are physically closed to KRTAP27-1 on the same chromosome.
The finding that KRTAP27-1 variation is associated with increased cashmere fiber MFD is noteworthy. This stated, selection for lower MFD may lead to reduced cashmere yields, as these two traits have a weak positive correlation, although the result of the association analysis whereby KRTAP27-1 variation does not appear to affect yield would suggest it might be possible. Given the high heritabilities reported for cashmere MFD [33,34], it could be concluded that greater progress in breeding for reduced MFD may be achieved through traditional selection approaches.