The search for new food sources has promoted, along the years, a global spread of well-marketed breeds [1]. As a result, the current domesticated animal dispersion is essentially restricted to a few breeds, and almost exclusively involves transfers from developed to developing countries, thereby imposing a major threat to the conservation and utilization of indigenous animal genetic resource [2].

The need to conserve livestock genetic diversity within breeds has been recognized already in the early 1990s [3], and is widely reflected by FAO initiatives [2]. However, in order to encourage breeders and policy makers to keep and preserve indigenous breeds, sound arguments related to their superior performance over the exotic breeds, in terms of economic and biological efficiencies, should be provided. While basic conservation programs focus on managing production performance through the control of population size and genetic variability, to ensure maintenance of adaptability to a given environment in favor of persistent production under non-favorable conditions, less attention has been directed to product quality [4].

Baladi (B taurus; DAGRIS), a native cattle found throughout the entire Southern Mediterranean basin [5], is facing the danger of extinction, due to the introduction of larger and more productive European breeds in these regions. Baladi (BAL) cows are commonly reputed to be tolerant of poor care, meager diet, adverse climate conditions [6], and high disease resistance [7]. Recently, its unique foraging behavior was explored by Dolev et al. [8]. Comparing to a larger frame breeds, the small-framed BAL cows were found to travel longer distances between foraging habitats and extended foraging periods, reflecting their physiological plasticity to cope with conditions of changing herbage quality [9]. Moreover, BAL cows were more efficient in conditions of low herbage quality [10]. Although the above highlights the improved adaptability of BAL cattle to the harsh Mediterranean ecosystems, near future scenario predicts potential dilution of its genetic diversity and hence, the need for its conservation becomes critical.

Due to the small population size of the BAL, an ill-conceived conservation strategy could be detrimental to its survival [6]. In order to promote practical conservation decisions for the BAL in addition to its integration into the local herd, it is crucial to study yet unexplored production quality traits of the breed, and estimate their interaction with adaptability traits.

While European beef cattle breeds introduced to the Mediterranean ecosystems are superior to BAL in their productive performance by means of growth rate and carcass weight, a comparative analysis of their meat quality characteristics haven't been carried out yet. Since population size of BAL cattle in the Southern Mediterranean basin has been largely reduced during the last decades [6], it limits the use of phenotypic characterization of meat quality, and turns the genomic approach, at this stage, prerequisite.

From a genetic perspective, determination of allelic and genotypic distribution of markers linked with economically important traits may be a powerful tool to gain an immediate knowledge as for the productive potential of livestock breeds and populations [4]. Indeed, several genes have been successfully identified as markers for meat quality characteristics. Polymorphisms in the calpastatin (CAST) and µ-calpain (CAPN1) genes were significantly associated with meat tenderness in several beef cattle breeds [11-15]. Tenderness is classified as one of the most important organoleptic quality traits affecting customer's choice [16]. Both calpastatin and µ-calpain are proteolytic enzymes involved in the process of meat tenderization following rigor mortis [17]. A lysine to alanine polymorphism encoding diacylglycerol O-acyltransferase 1 (DGAT1), a microsomal enzyme that catalyzes the final step of triglyceride synthesis, has been shown to be associated both with milk fat (percentage of lipid content of milk) [18, 19] and intramuscular fat content (IMF; percentage of lipid content within muscle) [20]. The last is an important characteristic of cooked meat, affecting both juiciness and flavor [21]. Fatty acid synthase (FASN) is a multifunctional enzyme complex that regulates de-novo biosynthesis of long chain saturated fatty acids, and the thioesterase (TE) domain within this complex is responsible for termination of fatty acid synthesis [22]. Polymorphisms in the TE domain have been shown to affect fatty acid composition [23].

In the current study, we estimated the allelic and genotypic frequencies of single nucleotide polymorphisms (SNPs) from CAST, CAPN1, DGAT1 and FASN genes in five cattle breeds: BAL, Limousine (LIM; beef breed; Bos taurus), Holstein (HOL; dairy breed; Bos taurus), Simmental (SIM; dual-purpose breed; Bos taurus), and Brahman (BRH; beef breed; Bos indicus). During the course of evolution and domestication, the classification of BAL cattle as either dairy or beef breed has still been obscured. Hence, the inter-breed comparison carried out in the current study was also meant to locate the genetic potential of BAL for meat quality within dairy, beef, dual-purpose and indicine bovine members.

DNA samples of 5 cattle breeds were analyzed in the current study: BAL, (n=92; females and males); LIM, (n=33; males); HOL, (n=216; males); SIM, (n=135; females and males) and BRH, (n=24; females).

Genomic DNA was extracted from blood samples of each animal using a commercial kit (Sigma, NA2020-1KT). Four single nucleotide polymorphisms (SNPs) were genotyped, one in each of the following genes: CAST, CAPN1, DGAT1 and FASN (Table 1). The CAST SNP is an A→G variation, present in exon 7 of the gene, causing an amino acid substitution of threonine to alanine at position 182 (T182A). This polymorphism was previously associated with meat tenderness (Table 1) [24]. The CAPN1 SNP is a C→G polymorphism, causing an amino substitution of glycine to alanine at position 316 (G316A) of the protein (Table 1). This polymorphism was previously reported to be associated with meat tenderness in cattle [12-14]. The DGAT1 polymorphism is a lysine to alanine amino acid substitution at position 232 (K232A), presents in exon 8 of the gene (Table 1). This polymorphism was first linked to a quantitative trait locus (QTL) in BTA14, associated with milk fat content and other milk characteristics [18, 19]. Later, the K232A polymorphism was associated with intramuscular fat content in cattle [20]. The fourth tested variant is an A→G polymorphism present in the FASN gene, causing an amino acid substitution of alanine to threonine at position 2266 of the protein (A2266T) (Table 1). This polymorphism was previously shown to be associated with fatty acid composition in cattle, as animals carrying the GG genotype had significantly higher oleic acid content and total mono unsaturated fatty acid concentration comparing to these carrying the AA genotype of the SNP [23].

SNP genotyping was carried out using custom TaqMan allelic discrimination assay, designed by Applied Biosystems according to manufacturer's protocol. Primers and TaqMan^® MGB probes were designed to amplify and target the two alleles of each SNP, using Primer Express^® software (Applied Biosystems, Foster City, CA). The list of oligos and probes used for genetic analysis is presented in Table 2.

All statistical analyzes were performed using JMP statistical software (JMP10 software, SAS institute). Gender-based allelic and genotypic frequencies for each polymorphism were studied in the different breeds. Pair-wise tests were performed to calculate for genotypic frequencies. For females the test included a comparison between BAL, SIM and BRH breeds. For males BAL, LIM, HOL and SIM were compared. Deviations from Hardy-Weinberg (HW) equilibrium were calculated using a contingency analysis - Chi Square likelihood test with a threshold of P≤0.05.

The populations examined in the current study were composed of females and males. While in BAL and SIM there was a mix of both genders, BRH was exclusively represented by females, and LIM and HOL by males. Inter-gender comparison was performed for the BAL and SIM breeds. Genotypic and allelic frequencies identified in females (n=74) versus males (n=18) of the BAL breed are presented in Table 3. As seen, significant differences were found between genotypic frequencies in the CAST SNP (P≤0.01), showing predominance of the heterozygous genotype AG in females (Table 3).

In the DGAT1 SNP, significant differences were identified between males and females for both, allelic and genotypic frequencies (P≤0.05; Table 3). The AK genotype was the highest for both genders. As for the homozygous genotypes, while the AA was higher in females, the KK was higher in males (P≤0.05). Within the allelic frequencies the A allele was higher in females and the K allele was higher in males (P≤0.05).

Unlike for CAST and DGAT1, no differences were detected between BAL females and males in the CAPN1 and FASN SNPs (Table 3).

In the SIM breed, no significant differences were detected between females (n=30) and males (n=105); gender-based analysis revealed the following P values for the following polymorphisms: CAST (P=0.17), CAPN1 (P=0.71), DGAT1 (P=0.39) and FASN (P=0.58). In the DGAT1 polymorphism, none of the SIM females and only 3 SIM males were homozygous for the KK genotype. Similarly, in the FASN SNP, none of the females and only 2 males were homozygous for the AA genotype (Table 4A, 4B).

Due to potential confounding effect of gender, highlighted by the differences found in BAL (Table 3), separate estimations of genotypic and allelic frequencies were carried out for females (BAL, SIM and BRH; Table 4A and males (BAL, LIM, HOL and SIM; Table 4B).

For females, no differences were revealed in genotypic and allelic frequencies of the CAST SNP, between BAL, SIM and BRH (Table 4A).

In the case of the CAPN1 marker, allelic (P≤0.05) and genotypic (P≤0.01) frequencies of BAL females were significantly lower than those of SIM and BRH (Table 4A). It is noteworthy, that all three breeds favored the CC genotype and C allele.

Interestingly, both, the allelic and genotypic frequencies of the DGAT1 marker differed among females of BAL, SIM and BRH (P≤0.0001; Table 4A). Estimated genotypic and allelic frequencies in the SIM showed the highest values of the AA genotype and A allele, followed by those of BAL. As expected, females of the BRH breed presented the highest frequency of the KK genotype and K allele (Table 4A).

For the FASN marker, females of BAL, SIM and BRH presented higher GG genotype and G allele frequencies compared to the AA genotype and A allele, but without significance among breeds (Table 4A).

For males, no differences were revealed in genotypic and allelic frequencies of the CAST SNP, between BAL and SIM (Table 4A). Of all tested males, LIM presented the highest AA genotype and A allele, followed by BAL, SIM and HOL (P≤0.0001; Table 4B).

In the case of the CAPN1 marker, genotypic frequencies of BAL males were significantly higher than those of LIM and HOL, but similar to those of SIM (P≤0.0001; Table 4B). With the exception of HOL the other three breeds favored the CC genotype.

The allelic and genotypic frequencies of the DGAT1 marker was similar in BAL and LIM, but differed from HOL and SIM males (P≤0.0001; Table 4B). Interestingly, unlike in females, BAL males exhibited higher frequencies of the KK genotype and K allele, not as in HOL and SIM, which favored higher values of the AA genotype and A allele Table 4B).

For the FASN marker, BAL and SIM males presented higher GG genotype and G allele frequencies compared to the LIM and HOL males (P≤0.0001; Table 4A).

In order to evaluate the genetic dynamics of meat quality in the above populations, with respect to the four polymorphisms, departures from HW equilibrium were estimated. Deviations from HW were identified for the CAST polymorphism in the SIM breed (P≤0.0001; Table 5). For the CAPN1 polymorphism, departures were observed in the BAL (P≤0.05), and HOL (P≤0.001) populations, as a lower than expected number of heterozygous was identified. Such departure was identified also in the BRH population (P≤0.0001), but in the opposite direction (Table 5). Both DGAT1 and FASN polymorphisms were in HW equilibrium in all breeds. In LIM, none of the four tested markers were in HW disequilibrium (Table 5).

The indigenous breed of Israel and other Southern Mediterranean countries like Egypt is the BAL. The BAL populations in these regions are facing the danger of extinction. In the current study we aimed at checking whether in addition to its superb adaptability to the harsh Mediterranean conditions, BAL conservation may be justified also on basis of the yet unexplored meat quality traits. Lacking a sufficient sample size that would enable phenotypic characterization of meat quality parameters, we concentrated herein on a genetic approach. It is well established that meat quality is determined by a large number of genetic and environmental factors [25, 26]. Therefore, in the current study, we focused on four SNPs in genes involved in meat tenderness (CAST and CAPN1 genes), intra-muscular fat content (DGAT1) and fatty acid composition (FASN).

As previously reported, the CAST genotypes (AA, AG) and A allele have been associated with improved meat tenderness [24]. In the current study, frequencies of the AA genotype and A allele were highest in LIM males, a breed selected for meat tenderness [27, 28]. In HOL males, a breed selected for milk production, frequencies of the AA genotype were the lowest (Table 4B). Interestingly, in females and males of the BAL breed, frequencies of the AA genotype and A allele were higher, comparing to the GG genotype and G allele, respectively. In addition, the AG genotype was found at high frequency in BAL females. A gender-based comparison between BAL and SIM (the predominant beef breed in Israeli herds), revealed similar frequencies of the AA genotype and A allele in females and males. Altogether, these findings imply a possible genetic potential for meat tenderness in the BAL breed. According to Rodero et al. [4], similarities only in CAST allele frequencies, between continental and indigenous cattle, might not fulfill the requirements to improve meat tenderness of local breeds.

The CAPN1 polymorphism was significantly associated with increased meat tenderness in several cattle breeds and populations [12, 29, 30]. In the study of Van Eenennaam et al. [30] this marker (G316A) was evaluated for meat tenderness within a two marker haplotype together with the marker CAPN1 4751-T3. The CAPN1 316/4751 C-C haplotype was significantly associated with increased meat tenderness in a mixed cattle population [31]. In the current study, a high frequency of the CAPN1 C allele was detected in both genders of the BAL and SIM breeds. For SIM, this is not surprising, as in addition to dairy this breed has also been selected for meat production. In BRH females, a fixation towards the CC genotype and C allele was detected. Assuming C is the favorable allele [29, 31-35], this finding is intriguing since BRH breed is apparently known for its low meat tenderness, emphasizing the need to test additional markers associated with the trait.

In males, CAPN1 CC genotype and C allele frequencies were significantly lower in LIM when compared to BAL and SIM, but higher as related to HOL, which demonstrated the lowest frequency. The HOL breed was characterized by a close to normal genotypic distribution in the CAPN1 marker, supporting the common notion that HOL was not selected for meat quality. An interaction between the CAST and CAPN genes was brought by De Tullio et al. [36]. Calvo et al. [24] suggested that a specific substitution in the CAST gene (T182A, located in the well conserved protein in its inactive calcium free form) might be responsible for the variation in meat tenderness, found seven days post-mortem in animals from the Parda de Montaña Spanish cattle breed. Although both, CAPN1 and CAST are proteolytic enzymes involved in the process of meat tenderization, calpastatin serves as µ-calpain inhibitor, moderating its activity post-mortem [17, 37]. The substitution of threonine by alanine at position 182 is generating a more stable union between µ-calpain and calpastatin, affecting meat tenderization [38]. Calvo et al. [24] has shown strong linkage disequilibrium (LD) between the decreasing G allele in CAPN1 and the increasing A allele in CAST in the Parda de Montaña and Pirenaica breeds. It is possible, that a similar situation exists in the LIM population tested herein, and this might explain the relatively low frequency of the CAPN1 G allele in this breed, known for its meat quality. Future studies should be conducted in the Israeli herd and BAL population to verify whether this LD exists in one or both populations, and if it does, which is the proper strategy to break it.

In females, the estimated frequency of DGAT1 AA genotype was significantly lower in the BAL than in SIM. BRH females showed the lowest frequencies of the alanine variant (A allele) and highest frequencies of the lysine variant (K allele). There is a lack of consensus regarding the association of the DGAT1 polymorphism with IMF [39]. In the study of Thaller et al. [20], the K allele was found to be significantly associated with higher IMF content in German Holstein animals. In a different study, Pannier et al. [39] found no association of the K allele with IMF content. Allelic frequencies estimated in the study of Pannier et al. [39] for several purebred cattle breeds were similar to these identified in our study for the SIM (both genders) and HOL (males) breeds, showing higher A allele frequency. The findings of Pannier et al. [39] were in agreement with these reported by Casas et al. [15], who found no association of the K allele with IMF content. Winter et al. [19]. suggested that the K allele is likely the ancestral state of DGAT1, as the lysine variant was identified in yak (Bos grunniens) and water buffalo (Bubalus bubalus), in addition to several Bos taurus and Bos indicus breeds. The presence of the alanine variant in the Anatolian Black local breed might point out that the K232A substitution probably raised early in the history of cattle domestication or even previous to that, especially since this breed is indigenous in a region known as a site of domestication of the European Bos taurus [40]. As for the alanine variant, it was not detected in subspecies of Bos indicus which domestication occurred independently [41]. Based on the above and as BAL does not present a fixation for either genotypes, it can be postulated that selection towards each allele might be equally achievable.

The GG genotype of the FASN marker is associated with health index and higher proportions of monounsaturated fatty acids [23]. While GG genotype and G allele frequencies were similar in BAL and SIM breeds, their distribution in LIM and HOL males, was significantly lower. Theoretically, the high frequency of GG genotype observed in BAL and Israeli SIM may partially be related to their adaptation to poor nutritional conditions, likely to occur in the Southern Mediterranean basin. Indeed, when lactating SIM cows were exposed to low energy and protein diet, they switched their milk fatty acids in favor of mono- and polyunsaturated ones [42].

A similar finding was detected in the CAST polymorphism for which the SIM breed showed significantly lower observed frequencies, presumably as a result of its selection as a dual breed.

Deviation from HW equilibrium may serve to estimate the genetic dynamics of population at given loci. In the current study, it was implemented, in order to evaluate the rate of divergence in allelic frequencies of meat quality related polymorphisms, with respect to CAST, CAPN1, DGAT1 and FASN markers. Departures from HW equilibrium were revealed only in the case of CAST and CAPN1. The lower observed frequencies in the CAST polymorphism in the SIM breed may presumably stem from its selection for dual purpose. In the case of CAPN1, deviations from HW equilibrium were detected in three breeds. While for BAL and HOL populations lower than expected number of heterozygous was identified, in BRH this departure was the other way around. In the BAL population, the favorable allele C of the CAPN1 marker was significantly higher, emphasizing the plausible potential of the breed with respect to meat tenderness. However, as already indicated, such findings are only speculative until verified by proper meat phenotyping. Similar findings were shown in the study of Rodero et al. [4] where two endangered Spanish cattle breeds presented lower observed than expected heterozygosity. As for the HOL breed, the deviation from HW could be explained by the fact that this breed is selected for dairy, with less emphasize on meat related traits.

In summary, for the first time, the current study examines the potential of BAL cattle with respect to meat quality traits, by means of genotypic / allelic frequencies of four genetic polymorphisms. Relative to cattle lineages that represent various classifications and agricultural specializations [Bos taurus (HOL, LIM, SIM) and Bos indicus (BRH), [dual purpose (SIM), dairy (HOL) and meat (LIM) breeds], the BAL population exhibited high frequencies of increasing alleles and genotypes in all four SNPs. Although from a conservation point of view this information sounds very encouraging, the upcoming research should focus on additional polymorphisms associated with these traits. In the future, adding these meat quality related markers to the already established repertoire of adaptive traits may add a great value to the “conservation initiative index” and promote conservation policy of BAL by stakeholders. However, this could be attained only when population size exceeds the critical number that defines a breed as endangered.