Genetic Diversity of Kedah Kelantan Cattle Breed and Its Crossbred types in Malaysia Based on Microsatellite Markers

The Kedah Kelantan (KK) is the indigenous cattle breed of Malaysia and is mainly kept by small farmers for meat production because of its small and compact body, and low maintenance requirement. This breed faces risk of germplasm dilution due to extensive crossbreeding and breeds replacement practices in the country. The population size of purebred KK is fast decreasing and most of the commercial populations are actually crossbreds. There is a lack of information on the genetic characteristics of KK. The genetic relationships between the KK, the synthetic breeds developed using the KK as the maternal line, as well as the non-descriptive KK crossbred types are also unknown. It is with these in mind that the present study was conducted. The objective of the study was to evaluate the genetic variability within and among the indigenous KK cattle and its crossbred types in Malaysia using 30 microsatellites loci.


Background
The Kedah Kelantan (KK) is the indigenous cattle breed of Malaysia and is mainly kept by small farmers for meat production because of its small and compact body, and low maintenance requirement. This breed faces risk of germplasm dilution due to extensive crossbreeding and breeds replacement practices in the country. The population size of purebred KK is fast decreasing and most of the commercial populations are actually crossbreds. There is a lack of information on the genetic characteristics of KK. The genetic relationships between the KK, the synthetic breeds developed using the KK as the maternal line, as well as the non-descriptive KK crossbred types are also unknown. It is with these in mind that the present study was conducted. The objective of the study was to evaluate the genetic variability within and among the indigenous KK cattle and its crossbred types in Malaysia using 30 microsatellites loci.

Results
All the 30 microsatellites loci used were polymorphic in all populations. Heterozygosity values observed were moderate and lower than the expected values. The inbreeding was present in all populations and could lead to loss of genetic diversity if not addressed. In general, the genetic differentiation measures were moderate, with a mean FST of 0.054. The structure analysis grouped the populations into three clusters. Analysis of zebu and taurine diagnostic alleles showed that all population had high proportion of Indian zebu alleles and very low proportions of African taurine and European taurine diagnostic alleles.

Conclusions
It may be concluded that there is still some genetic variation present in the KK. However, this genetic diversity is at risk of being lost if no appropriate breeding practices are implemented.

Background
Indigenous breeds of livestock represent valuable resource to their owners and their countries [1].
However, these breeds are often not fully evaluated and are neglected in terms of genetic improvement. Many indigenous breeds, particularly those adapted to harsh environments of developing countries, have not yet been sufficiently characterized [2]. Moreover, many of these breeds, especially in the Asian countries, have been subjected to crossbreeding with improved breeds in order to improve their productivity [3]. The adoption of controlled crossbreeding strategy does not always produce the expected results [4], because of adaptation problems faced by the crossbred animals. Indiscriminate crossbreeding is the main threat for indigenous breeds especially under smallholder farming conditions [5], and this has led to severe reduction of indigenous breeds in many countries [6]. Industrial livestock farming is also a great threat to indigenous breeds; a few improved commercial breeds have already replaced or are fast replacing the latter. Indigenous breeds may perish before their potentials are fully recognized and exploited [2]. Therefore, there is an urgent need to conserve indigenous breeds [7]. Understanding the origin and subsequent history of indigenous breeds as well as their genetic diversity is essential to design strategies for their sustainable use and conservation [8].
In Malaysia the Kedah Kelantan (KK) is the only indigenous cattle breed and is mainly kept for meat production. It is popular among the small farm owners because of its small compact body size and low maintenance requirement [9]. However, the owners of bigger herds prefer the larger KK crossbreds or the imported exotic breeds. Genetic improvement of the KK has been by controlled crossbreeding strategies adopted by the government of Malaysia [10] through the Department of Veterinary Services Malaysia (DVS) and the Malaysian Agriculture Research and Development Institute (MARDI) to improve its productivity and increase the local beef production. However, uncontrolled crossbreeding of the KK with improved beef breeds has also been popular, especially among farmers and smallholders [11], resulting in non-descript KK crossbred types. As a result of these activities, the population size of the purebred KK population is fast decreasing, and most of the current commercial beef cattle populations are actually crossbreds of KK. The KK breed faces risk of germplasm dilution due extensive crossbreeding and breed replacement practices. There is a lack of information on the genetic characteristics of the KK. The proportions of KK genes in the KK crossbred populations are also unknown. These make it necessary to evaluate the genetic makeup of the KK and KK crossbred breed types and evaluate the genetic variation among these populations. The objective of this study was to evaluate the genetic variability within and among the indigenous KK cattle and its crossbred breed types in Malaysia.

Methods
A total of 312 animals were used in this study. These animals could be classified into three groups: indigenous KK, KK crossbreds and exotic breed. The indigenous KK used in this study were from the nucleus herd at the DVS Livestock Centre in Tanah Merah, Kelantan. The crossbreds were of two types: the crossbreds developed through planned crossbreeding programmes, which included the Brakmas and Charoke, and the non-descriptive KK crosses. The Brakmas animals used in the present study were from the nucleus herd maintained at the MARDI Station in Muadzam Shah, Pahang, while the Charoke animals were from the MARDI Station in Kluang, Johor. The two non-descriptive crosses used are referred to as KK cross 1 (KKX1) and KK cross 2 (KKX2). The KKX1 herd initially began as a nucleus herd of pure KK. However, it was not maintained as such and the animals were mixed with other breeds of both Zebu and Taurine types. KKX2 was a herd belonging to a commercial meat production farm in Kluang, Johor (Kulim Livestock Sdn. Bhd.). This herd was the result of crossing KK with the Brahman breed. Since no breeding design was followed and mating was random these animals were considered as a non-descriptive KK cross. These latter two KK crossbred types were included in the present study as they represent many of the cattle herds in the country. The exotic breed used in the study was the Brahman breed. The animals were from the nucleus herd maintained at the DVS Livestock Centre in Kuala Berang, Terengganu. The Brahman has contributed to the beef industry development in Malaysia, and many owners of big farms have Brahman herds and smaller farm owners cross their cattle with the Brahman; therefore, this breed was included as an out-group, a non KK breed type. Figure 1 shows the locations of the sampled populations for each breed.

Samples
Blood samples from 56 random animals from each herd were used, except for the Brahman where samples from only 32 animals were used.

Microsatellite markers
Thirty microsatellite markers were investigated in the present study. These markers were those recommended by FAO/ ISAG advisory group for genetic diversity studies in cattle [12] (Table1).  The PCR products were separated using an automated capillary sequencer (CEQ 8000; Beckman Coulter). The CEQ 6.0 software was used to automate allele sizing by comparisons with size standard 400 (Beckman Coulter, USA), which includes DNA fragments from 60 to 420 bp. Data analysis was performed using the AE2 subroutine (dye mobility correction). This analysis included analysis of raw data to estimate the fragment sizes, and filtration of low quality and unwanted samples. Binning was also performed to estimate the allele sizes.

Data analyses
The mean of observed and effective number of alleles (MNA and MNE) and Shannon information index per locus per population were estimated using the Popgene software [13]. The informativeness of the marker and polymorphism information content (PIC) were calculated according to Botstein [14] using the Power Marker software [15]. inferring population structure using genotype data. To choose the appropriate number of inferred clusters (K), two to seven inferred clusters were performed with three independent runs each. All analyses used a burn-in period of 10000 and 10000 iterations for data collection. The phenomenon that once the real K is reached the likelihood for larger K's (LnP(D)) plateaus and the variance among run increases were not observed. When K was increased, the LnP(D) increased continuously.
Therefore, the Evanno method was used. The values of LnP(D) for each K were plotted and the delta K (DK) statistics were estimated, which is based on the rate of change in LnP(D) between successive K values [17]. One to ten inferred clusters were performed with 20 independent runs each. Based on references [18][19][20] that suggest groups of alleles which may be used as diagnostic markers of Indian zebu, African taurine and European taurine breeds, the frequencies of these alleles at the loci were averaged to estimate of proportion of introgression from the individual cattle groups.
Pairwise genetic distances between the six breed-types were estimated using Nei's standard genetic distances (DS) [21]. Phylogenetic trees were constructed based on four genetic distance measures: Cavalli-Sforza [22], Nei's DA genetic distances [23], Goldstein [24] and Shriver [25] using neighbour joining (NJ) method [26]. The robustness of tree topologies was evaluated with a bootstrap test of 1000 resampling across loci; the PowerMarker program [15] was used for this purpose. The phylogenetic trees were edited using MEGA 5 program [27]. The principal components analysis (PCA) was performed using the allele frequencies according to Cavalli -Sforza [28] with the aid of the Unscrambler X 10.1 software.

Microsatellite Loci Analysis
All 30 loci were amplified successfully and were polymorphic. As shown in Table 2        heterozygotes in a group of populations, F ST = the degree of gene differentiation among populations in terms of allele frequencies.

Genetic Variation and Relationship between KK and KK Crossbred Types
The genetic structure of the study populations is shown by the clustering assignment of the 312 animals representing the six breed types (Figure 3). The value of K = 3 was chosen as this showed the highest delta K (Figure 3) as suggested by Evanno [17]. Table 6 shows the average proportions of memberships (q) to the three clusters. KK, Charoke and Brahman were grouped in cluster 1, 2 and 3, respectively, with q ≥ 0.80; while KKX2 and Brakmas were grouped in cluster 1 and 3, respectively, with q ≤ 0.80. The KKX1 animals were split among the three clusters. The genetic compositions of the 56 KK individuals (q-values) are shown in Figure 4.  Table 7 shows the zebu and taurine diagnostic alleles and the frequencies of these alleles in the KK and its crossbred types. As shown in Figure 5 Table 8 shows the genetic distance among the breed types. The KK breed was found to be very    The results of the PCA on the allele frequences of the 30 microsatellite loci in the six breed types are shown in Figure 8. The first axis, which accounted for 36% of the variation, separated KK, KKX2, KKX1 and Charoke from Brahman and Brakmas. The second axis accounted for 32% of the variation, and separated the KKX1 and Charoke from KK and KKX2, each in a different cluster.

Genetic Diversity at the Microsatellite Loci
A TNA of 360 was observed for the six breed types in the present study. This was higher than that reported for five Cuban cattle breeds (n = 317; TNA = 299) screened for the same 30 microsatellite loci [29]. This may be due to the fact that the Cuban cattle breeds were purebreds, while the present study included crossbred animals with a number of contributing breeds. The TNA was, however, lower than that reported for 27 indigenous Chinese cattle breeds (n = 1638; TNA = 480) for the same loci [30]. This most likely is due to the big sample size and larger number of breeds from different parts of China used in the Chinese study. The range of the PIC values, however, was similar to that reported for seven of the native Chinese cattle breeds (PIC = 0.74 -0.75) [30] and for the five Cuban breeds [29]. In general, the high number of alleles and PIC values for the individual loci showed that the 30 microsatellite loci recommended by ISAG/FAO for genetic diversity evaluation of cattle are highly informative and suitable for the purpose.

Genetic Diversity within KK Cattle and KK Crossbred Types
The genetic diversity within the breed types was evaluated using the allelic variation. The MNA is consider as a good indicator of genetic variation. The Brahman breed exhibited the lowest MNA (7.7) among the six breed types. This may be due to the smaller sample size used (n = 32), but most likely due to founder effects and population bottlenecks as a result of breeding practices. The Brahman were introduced into Malaysia to be used for crossbreeding as well as to increase the local cattle population. The KK breed had lower MNA than the KK crossbred types. This may be attributed to crossbreeding incorporating the alleles of the parental breeds into the crossbred types [31]; the KK crossbred types would have the KK alleles as well as those of the other parental breeds, increasing the number of allele.
The MNA of KK and KK breed types were lower than that reported for four Chinese native cattle breeds (10.1-10.5) [30], but was higher than those reported for six Spanish native cattle breeds (4.9 -6.7) [32]. This may be attribute to the differences in population sizes and the sampling technique used. The sample size for the Spanish breeds were between 29 to 50 individuals. Moreover, the breeds occurred as isolated populations and four of them were considered as endanger. Isolated small populations lose genetic variability over time but they become distinctively different.
The average heterozygosity is the best general measure of genetic variation [33]. Selection and controlled mating were practised to a certain extent in the farms concerned, especially the nucleus farms. However, proper record keeping was generally lacking; records were often available for growth traits but pedigree records were often not available. The observed F IS values were higher than that reported for Chinese and Ugandan native cattle breeds [30,37] screened using the same 30 microsatellite loci.
A population is said to be in Hardy-Weinberg equilibrium (HWE) when the gene and genotype frequencies remain constant from generation to generation, and the latter is of a definite proportion [38]. In the present study the deviations from HWE was observed in all six breed types. Deviation from HWE could be attributed to many causes, among which are selection, assortative mating, migration and small population size, all of which could have influenced these populations. The overall numbers of loci that deviated from HWE were high compared to that reported for 10 Ethiopian and 10 Portuguese native cattle for the 30 microsatellite loci [35,39].
In general, the low genetic diversity observed in the present study in terms of low mean number of the alleles, heterozygote deficiency and deviation from HWE could be attributed to many reasons, but the most probable reasons are inbreeding, small population sizes and assortative mating. The KK, Brakmas, Cheroke and Brahman animals used in the present study were from single nucleus herds.
According to Phillips [40] there are many factors associated with establishing and managing nucleus cattle herd which lead to inbreeding. These include the nucleus herd size, whether the nucleus is open or close, the desired age structure of the nucleus, selection criteria and selection accuracy for the bulls and replacement cows, and completeness of the performance and pedigree records. When these factors were investigated, it was noticed that the records in these nucleus farms were limited and often incomplete. Vital pedigree information was often missing. Consequently, this would have affected the selection accuracy for the bulls and replacement cows. The record keeping at these farms has to be improved and the system reviewed regularly if genetic variability is to be maintained and herd performance is to be improved.

Genetic Variation and Relationship between the KK and KK Crossbred Breed Types
The genetic variation between KK and KK crossbred types were evaluated by estimation of Fstatistics, gene flow, genetic admixture and genetic distance, as well as by phylogenetic analysis and principal component analysis (PCA In the present study Chinese indigenous cattle breeds (F ST = 8%) [30]. This was probably due to the fact that the breeds used in these earlier studies originated from different parts of the respective countries. For example, in the Indian study the Sahiwal breed was native to Pakistan and found along the India-Pakistan border in the North, while the Hariana and Deoni were found in northern and western India respectively. In the study by Zhang [30] the 27 breeds were representatives from all parts of the vast land area of China, from north, south, east and west. The F ST value in the present study was, however, higher than that reported by [35] for 10 Ethiopian cattle breeds (F ST = 1.3 %). The most probable cause of this low level of genetic differentiation in these Ethiopian breeds is the fact that Ethiopian cattle breeds have common historical origins, and shared common grazing lands and watering points. Moreover, an uncontrolled mating practice, which is predominant in Ethiopia, increases the gene flow among the breeds.
It is very clear that the F ST between KK and KKX1 (F ST = 1.5%) and between KK and KKX2 (F ST = 1.8%), were lower than the F ST between KK and Brakmas (F ST = 3%) and between KK and Charoke (F ST = 4%). KKX1 and KKX2 represent unplanned breed types and the KK was probably the most common breed used in the mating, thus being a major gene contributor. The Brakmas and Charoke are synthetic breeds developed using planned breeding design, and, therefore they are more different from the KK than the other KK crosses. The genetic makeup of a synthetic breed is not easy to manage and monitor; it is influenced by inbreeding and selection for fitness and desired traits (fertility, fleshing ability, mature weight and coat colour, etc.) which may be bias to one of the parental breeds. In this study, it is clear that all breed types were bias towards the KK breed genes.
The degree between breed differentiations indicated a relatively moderate to high gene flow between the six cattle breeds (Nm = 4.38). The highest gene flow (16.84 %) was between the KK and KKX1 population. The gene flow between the KK and KKX1 populations (16.84%), between KKX1 and KKX2 populations (13.65 %), and between KK and KKX2 populations (13.39%) reflect the genetic similarity between these three breeds, supporting the findings based on the F statistics. High gene flow was also observed between the Brakmas and KKX2 (9.71%) and the Charoke and KKX1 (9.04%) also indicating their genetic closeness between these two pairs. The earlier association may be due to the fact that both Brakmas and KKX2 are crossbreds of KK and Brahman. As for the second pair, both Charoke and KKX1 shared the same ancestral KK population which was then kept as a nucleus herd at MARDI Station in Kluang, Johor. However, due to funding shortage this KK herd was not maintained and the animals were crossed with the different available breeds. Moreover, since both these breed types were from the same farm, there is a possibility that at times there was interbreeding between the two herds. Although the Brakmas and Charoke are crossbreds of KK they showed low inter breed gene flow (5.04%) compared to the others pairs. This could be attributed to the physical separation of the two breeds, and the breeding and selection programs practised in the respective farms; the latter may have been bias towards the genes of the exotic breeds.
The genetic admixture in the KK cattle and the KK breed types was estimated using structure analysis and frequency analysis of the zebu and taurine diagnostic alleles. The results of the structure analysis showed that the studied populations were split into three clusters: KK and KKX2 in the first cluster, (DVS) as part of their routine screening of animal herds in the country. Random samples were obtained by the researchers from DVS for the present study. We obtained written informed consent to use these animals in this study from the owner(s) of the animals.

Consent for publication
Not applicable

Availability of data and material
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Figure 2
Clustering assignments of 312 animals representing the six cattle breed types. K = number of clusters. When K (the inferred number of cluster) was two, KK and KKX2 were grouped into one cluster, and Charoke and Brahman were separated in the other cluster, while Brakmas and KKX1 showed admixture from both clusters. For K=3, KK and KKX2 were grouped again into one cluster, Brakmas and Brahman were grouped into another cluster, while Charoke was clearly separated into a different cluster; KKXI showed admixture from the three clusters. With K= 6, the six populations were inferred to six cluster.