Introduction

Microsatellites, also known as simple sequence repeats (SSRs), are tandemly repeated DNA sequences that are generally 1–6 bp in length per unit (Tautz and Renz 1984). As one of the most popular source of genetic markers, SSRs are widely employed in population genetics, biogeography and microevolutionary studies (Guichoux et al. 2011). The yak (Bos grunniens) is endemic to central Asia, being adapted to the cold and high altitude environment. As one of the local important domestic animals, yaks play an indispensable role in the region. More than 14 million domestic yaks provide meat, milk, transportation, dung for fuel and hides for Tibetans and other nomadic pastoralists living at high altitudes (Wiener et al. 2003). In the past few years, using cattle-specific SSR markers, some researchers have studied the population and evolution genetics of yak (Ritz et al. 2000; Wang et al. 2003; Nguyen et al. 2005; Zhang et al. 2008; Qi et al. 2005, 2010; Ramesha et al. 2012). Recently, Cai et al. (2014) reported 19 novel yak-specific polymorphic microsatellites including nine perfect microsatellites and ten imperfect or compound repeats. These studies contributed a great deal of valuable information for the assessment, protection and management of yak as a genetic resource.

The availability of complete genome sequence for the yak has made it possible to carry out genome-wide analysis (Qiu et al. 2012). However, to date there are no reports on the abundance and density of microsatellites (1–6 bp) repeats in the yak genome. We screened the entire yak genome sequence to study the distribution and density of perfect SSRs, in order to facilitate the understanding of structure of the yak genome, and to build up a foundation for the isolation and identification of more yak-specific SSRs.

Materials and methods

The complete yak genome sequence with a total length of 2.66 Gb was downloaded (Hu et al. 2012) in FASTA file format to generate SSRs data. MSDB 2.4.2 (Microsatellite Search and Database) (http://msdb.biosv.com/) (Du et al. 2013) was used to scan the entire yak genome for abundance and density of perfect SSRs, using the “perfect” search mode. We identified six classes of microsatellites: mono-, di-, tri-, tetra-, penta- and hexa-nucleotide SSR motifs at a minimum repeat number of 12, 7, 5, 4, 4 and 4, respectively. The length of flanking sequence was constrained to 200 bp. Microsatellite statistics were selected using the “whole” mode, which means the program will generate one statistical Excel file for all sequence files as a whole. Repeats with unit patterns being circular permutations and/or complements were considered as one type for statistical analysis. For example, AGC denotes AGC, GCA, CAG, GCT, TGC and CTG in different reading frames or on the complementary strand. The software SPSS 19.0 was used to perform the data analysis and mapping. To facilitate the comparison among different repeat types or categories, the relative frequency, (SSR number per Mb of the sequence analyzed), and the relative density, [SSR length (in bp) per Mb of the sequence analyzed] were evaluated.

Results

Frequency and density of six classes of microsatellites

After scanning the genome sequence for six classes of SSRs, a total of 723,172 SSRs were identified in the yak genome assembly (Table 1). The total and mean lengths were 12,539,047 bp and 17.34 bp, respectively. The relative frequency and density were 272.18 loci/Mb and 4719.25 bp/Mb, respectively. About 0.47 % of the yak whole genome (2.66 Gb) was occupied by the perfect SSRs.

Table 1 Count, length, frequency, density and percentage of six types of perfect microsatellites in yak genome sequence
Full size table

The counts, length, frequency, density and percentages of the six classes of perfect SSRs are summarized in Table 1. Mono-nucleotides were the most abundant type, with the highest relative frequency (119.85 loci/Mb) and density (1762.75 bp/Mb), accounting for 44.04 % of all SSRs, followed by dinucleotides (24.11 %), tri-nucleotides (15.80 %), penta-nucleotides (9.50 %) and tetra-nucleotides (6.40 %). Hexa-nucleotides were much less abundant, accounting for only 0.15 % of all SSRs (Table 1).

Abundance and repeat numbers for different microsatellite categories

Mononucleotide repeats

Poly (A) [or poly (T)] was the predominant mononucleotide repeat category, with 312,471 loci accounting for 98.13 % of the mononucleotide SSRs. The total length, frequency and density of poly (A) was 4.59 Mb, 117.60 loci/Mb and 1727.86 bp/Mb, respectively, and the average length was 14.69 bp (Table 2). However, poly (C) [or poly (G)] was far less abundant than poly (A) [or poly (T)], accounting for only 1.87 % of the total. The abundance of poly (C) was also lower (namely 2.25 loci/Mb and 34.90 bp/Mb, respectively). The repeat numbers of mononucleotide repeats ranged from 12 to 967 times. But the repeat times ranged from 12 to 29 were predominant, which numbered 317,538 accounting for 99.72 % of the total count of mononucleotide SSRs (Fig. 1A).

Table 2 Count, length, frequency and density percentage of different categories of SSRs (frequency above 1 loci/Mb) in yak genome sequence
Full size table
Fig. 1
figure 1

Repeat times of different types of SSRs in yak genome

Full size image

Dinucleotide repeats

Dinucleotide repeats include AC, AT, AG and CG categories of SSRs. Results showed that the frequencies of AC and AT were highest (40.22 loci/Mb and 19.00 loci/Mb, respectively). AG had the middle frequency of 6.29 loci/Mb (Table 2). These three categories of SSRs numbered 174,048 and accounted for 99.82 % of the total number of dinucleotide repeats. The CG repeat had the lowest frequency of 0.12 loci/Mb and numbered 315. The repeat times of dinucleotide repeats ranged from 7 to 1206 times. However, the predominate repeat times ranged from 7 to 25 which numbered 173,346 and accounted for 99.42 % of the total count of dinucleotide SSRs (Fig. 1B).

Trinucleotide repeats

Statistical analysis of all trimer repeats including AAC, AAG, AAT, ACC, ACG, ACT, AGC, AGG, ATC and CCG showed that AGC had the highest frequency of 32.61 loci/Mb. Three categories of AAC, AAT and ACC had the middle frequencies that were 3.48 loci/Mb, 2.57 loci/Mb and 1.68 loci/Mb (Table 2), respectively. The others had the lower frequencies which ranged from 0.01 to 0.82 loci/Mb. The repeat times of trinucleotide SSRs ranged between 5 and 1033 times. But 5–11 repeat times were predominant and numbered 113,808 and accounted for 99.59 % of the total count of trinucleotide SSRs (Fig. 1C).

Tetranucleotide repeats

A total of 33 categories of tetranucleotide repeats were obtained in this study. Analysis of frequencies and densities of each tetrameric repeat categories revealed that ATTT, GTTT, AATG, CTTT and ATGG were predominant across the genome, and had frequencies of 4.58 loci/Mb, 2.82 loci/Mb, 1.48 loci/Mb, 1.44 loci/Mb and 1.15 loci/Mb (Table 2), respectively. The overall frequencies of 24 tetrameric repeats namely ATGC, ACAT, CCTT, AGAT, AGTG, CCCT, ACTG, AATT, CTGT, AACC, AGGC, AAGT, GGGT, AATC, ACGC, AGCC, CTTG, ACCT, AGTT, AGCT, AGCG, CCCG, ACGG and CCGG were at the middle level, which ranged between 0.01 loci/Mb and 0.71 loci/Mb. There were four categories of tetrameric repeats namely CGAA, GGTC, GTAC and TCGA which had low densities (namely 0.01 bp/Mb). The repeat times of tetranucleotide SSRs ranged between 4 and 248 times but 4–8 repeat times were predominant, and numbered 46,201 and accounted for 99.78 % of the total count of trinucleotide SSRs (Fig. 1D).

Pentanucleotide repeats

In pentanucleotide repeats categories, AACTG and ATCTG had a higher frequency of 18.79 loci/Mb and 4.04 loci/Mb, and density of 401.32 bp/Mb and 86.46 bp/Mb (Table 2), respectively. The frequencies of remainder categories were <0.76 loci/Mb. The repeat times of pentanucleotide SSRs ranged between 4 and 339 times. However, repeats ranging between 4 and 6 times were predominant, and numbered 68,439 and accounted for 99.62 % of the total count of pentanucleotide SSRs (Fig. 1E).

Hexanucleotide repeats

The frequencies of all hexanucleotide repeat categories were lower than that of above five types of repeats, and ranged between 0.03 loci/Mb and 0.00 loci/Mb. The repeat times of hexanucleotide SSRs ranged between 4 and 89 times. However, the predominate repeat times ranged between 4 and 7, and numbered 1072 and accounted for 97.90 % of the total count of hexanucleotide SSRs (Fig. 1F).

Discussion

Currently, a SSR scan for the entire yak genome sequence using bioinformatics methodology has not been reported. Our research firstly examined the abundance of perfect SSRs composed of 1–6 bp motifs in yak genomic sequence. In our study, approximately 0.47 % of the yak genome comprised perfect SSRs from mono- to hexa-nucleotide repeats. This percentage is similar to the results that reported before on the cattle (0.48 %) (Qi et al. 2013), sheep (0.48 %) (Qi et al. 2013) and chicken (0.49 %) (Huang et al. 2012), but smaller than that of other species genomes such as human (3 %) (Subramanian et al. 2003), mosquitoes (2.14 %) (Yu et al. 2005) and mouse (2.85 %) (Tong et al. 2006). These differences also could be due to the variation in search criteria, size of the database and bioinformatics software tools used in different studies for identification of SSRs.

Unsurprisingly, the proportion of the six classes of perfect SSRs was not evenly distributed in the yak genome. Mononucleotide repeats, accounting for the largest proportion (44.04 %) in six types of SSRs, had the highest frequency (119.85 loci/Mb) and maximum density (1762.75 bp/Mb), followed by dinucleotide, trinucleotide, pentanucleotide and tetranucleotide repeats. Hexanucleotide repeats had the lowest frequency (0.41 loci/Mb) and minimum density (12.04 bp/Mb) (Table 1). This trend is similar to what has been found in human, cattle, sheep and chicken genomes (Subramanian et al. 2003; Huang et al. 2012; Qi et al. 2013), but is different from that of mouse, silkworm, drosophila, mosquito and zebra fish (Katti et al. 2001; Li et al. 2004; Yu et al. 2005; Tong et al. 2006). This difference in abundance might be due selection for or against mono-, di- and trimers to tetra-, penta- and hexamers repeats.

In the present investigation, the number and density of certain repeat categories are greater than others within each type of repeats. In the case of mononucleotide repeats, Poly (A) [or Poly (T)] exhibited a strong over-representation, accounting for 98.13 % of total number of mononucleotide SSR categories. Similarly, in the other five classes of SSRs, fourteen categories including AC, AT, AG, AGC, AAC, AAT, ACC, ATTT, GTTT, AATG, CTTT, ATGG, AACTG and ATCTG in yak genome were the predominant repeats, which all had a normal frequency above 1.00 loci/Mb (Table 2). It is possible that during SSR evolution the poly (A) stretches present in the genome might have mutated to produce the A-rich repeats. It is also possible that the abundance of repeats is influenced by their secondary structures and the effect on DNA replication. In addition, the repeat times of different categories of SSRs was also different. For example, the repeat times for mononucleotide SSRs mainly ranged between 12 and 29, for dinucleotide SSRs ranged between 7 and 25 times, for trinucleotide SSRs ranged between 5 and 11 times, and for tetranucleotide, pentanucleotide and hexanucleotide SSRs the repeats ranged between 4–8, 4–6 and 4–7 (Fig. 1), respectively. (Schlotterer 1998) showed that nucleotide sequences with higher GC content possessed fewer SSRs than those of higher AT content. Our results are consistent with this research, indicating that SSRs in yak genome are also AT-rich.

It should be noted that although the complete genome sequence of yak was obtained in 2012, it has not yet been physically mapped. So, in the future studies, after assembling the yak genome sequence to each chromosome, the following areas need to be further explored. Firstly, comparative analysis of abundance of SSRs on different chromosomes, and the association between the length of chromosomes and the distribution of SSRs on each chromosome needs to be investigated. Moreover, the difference in abundance of different classes of SSRs in coding and non-coding regions of yak genome (i.e. exon, intron and intergenic regions) should be studied. Some studies showed that SSRs plays an important role in the structure and function of the genome and may be associated with some diseases (Hefferon et al. 2004; Campregher et al. 2010). Therefore, another research focus should be to reveal genetic mechanisms, the function of SSRs in the yak genome and correlative analysis between some diseases and SSRs. Lastly, at present, the genome sequence of yak Y chromosome has not yet been obtained as the present complete genome sequence came from a female yak (Qiu et al. 2012). Pian Niu or Cattle-yak (Bos taurus × Bos grunniens), the first filial generation of yak and ordinary cattle, showed obvious hybrid vigor. However, an issue with crossbreeding and improvement of yak is that the males are sterile, thus it is not possible to reliably utilize the heterosis. Until now, although many studies have been done on the sterility of the male Pian Niu both at home and abroad, there has been no solution to the problem of male sterility (Luo et al. 2014). Therefore, it is necessary to obtain the genome sequence of yak Y chromosome and mine more Y-chromosome-specific SSR markers. Then, combining to the Y chromosome information from cattle, Pianniu (Bos taurus × Bos grunniens), zebu and others, the problem of male sterility of Pianniu can be explored.