Genetic variability of 10 microsatellite markers in the characterization of Brazilian Nellore cattle ( Bos indicus )

We assessed the polymorphism of 10 microsatellites in Brazilian Nellore cattle (Bos indicus) using a commercial multiplex system. Allele frequencies, polymorphism information content, heterozygosity and exclusion probability were calculated. Allele frequencies revealed that in the sample analyzed the markers were not equally polymorphic. The exclusion probabilities and the polymorphism information content of some loci in Nellore cattle were lower than in Bos taurus breeds. When all the microsatellites were considered the combined exclusion probability was 0.9989. This multiplex analysis can contribute toward pedigree information, adequate genetic improvements and breeding programs.


Introduction
The Bos indicus Nellore herd is one of the largest commercial beef herds in the world and is well-adapted to tropical regions.According to the Brazilian Ministry of Agriculture, Livestock and Supply (http://www.agricultura.gov.br), the Nellore herd is the most important beef herd in Brazil, where the total number of both purebred and crossbreed Nellore cattle totals over 140 million head.Due to genetic improvement programs and adequate international sanitary standards, Brazil has ranked as the top beef exporter since 2003, with an export volume exceeding 1.18 million tons (Brazilian Association of Meat Export Industries http://www.abiec.com.br).Accurate pedigree information is essential to maintaining the quality of breed improvement programs and molecular markers have become an important genetic tool in animal genetics studies, allowing the analysis of genetic variability within and between herds.Many of the current molecular marker techniques are based on variations of the polymerase chain reaction (PCR) such as random amplification of polymorphic DNA (RAPD).
Microsatellites markers have been widely used as a genetic markers in bovine population studies and pedigree verification (Visscher et al. 2002, Hansen et al, 2002, Ibeagha-Awemu and Erhardt, 2005), mainly because of their large polymorphism information content, widespread distribution in the eukaryotic genome (Tautz and Renz, 1984) and robust methodology.Microsatellites have been effective in evaluating differences within cattle breeds and in determining population substructures (MacHugh et al, 1998;Ciampolini et al., 1995).More than 1400 microsatellites have been mapped in the cattle genome (Luikart et al., 1999) and some of them have been employed in population genetics studies and parentage verification.Many microsatellite loci have been used in Nellore improvement programs but, to date, there have been no reports of pedigree verification studies using microsatellite markers, pedigree verification in Brazilian livestock currently being based on blood groups and biochemical polymorphism analyses.
The aim of the study described in this paper was to characterize Brazilian Nellore cattle through the analysis of the genetic variability of ten microsatellite markers and to evaluate if these markers are informative in parentage tests.

Sample collection and DNA extraction
We sampled 200 unrelated adult Nellore cattle (150 dams and 50 bulls) registered in their breeding associations and randomly selected from private and research herds belonging to 43 farms located in various regions of Brazil.Blood samples were collected in heparinized glass tubes and total genomic DNA isolated as described by Debomoy et al. (1991) and stored at -20 °C.

Microsatellite amplification
As recommended by the International Society of Animal Genetics (ISAG), ten microsatellites (Table 1) were selected for the analysis, using the Stockmarks for Cattle Bovine Genotyping Kit (Applied Biosystems Division, Perkin-Elmer, Foster City, CA).Multiplex amplification was carried out in a final volume of 15 mL containing 50 ng of template DNA, 0.5 units of AmpliTaq Gold TM polymerase (PE Applied Biosystems, Foster City, CA), 3.0 mL Stockmarks Buffer, 400 mM of each dNTP and 5.5 mL of primer mix (Table 1).The reactions were carried out using a Programmable Thermal Controller PTC-100 TM (MJ Research, INC) in an initial denaturation phase of 15 min at 95 °C, followed by 31 cycles of 45 s at 94 °C, 45 s at 61 °C and 1 min at 72 °C.A final extension was carried out at 72 °C for 1 h and then at 25 °C for 2 h.After amplification, 90 mL of water was added to the tubes and 0.4 mL of this solution was mixed with 2 mL loading mix (DI formamide: dye: GS350Rox -6:1:1) and analyzed in a 6% (w/v) denaturing gel using an ABI PRISM TM 377 DNA Sequencer.The fluorescence data was collected by GeneScan TM Analysis 2.0 and analyzed using Genotyper TM 2.0 software.

Data analysis
The GENEPOP package Version 3.4 (Raymond & Rousset, 1995) was used to calculate an exact test for deviation from Hardy-Weinberg equilibrium (HWE), allele frequencies and heterozygotic deficiency.Since the microsatellite loci have more than four alleles, an unbiased estimate of the exact HWE probability was calculated using the Markov chain method of Guo & Thompson (1992).The gene diversity (D) was calculated with FSTAT 2.9.3.2 (Goudet, 2001).Exclusion probability (EP), combined exclusion probability (CEP), expected heterozygosity (He) and observed heterozygosity (Ho), and polymorphism information content (PIC) were calculated using Cervus 2.0 software (Marshall et al., 1998).

Results
Ninety-four alleles were detected from the 10 loci surveyed, yielding a mean value of 9.4 alleles per locus.The allele frequencies of 10 microsatellites are listed in Table 2. Allele frequencies revealed that not all markers were equally informative.The TGLA227, BM1824 and TGLA53 loci each had one allele with a much higher frequency than the other alleles (75 bp, 180 bp and 160 bp respectively).The loci ETH10 and ETH3 each had two alleles with high frequencies (209 bp -207 bp and 115 bp -117 bp, respec-  The ETH3 locus did not show heterozygote deficiency, although its P value was close to p < 0.01.The mean PIC value was 0.640 and the mean expected heterozygosity value was 0.679.Expected and observed heterozygosity, probability of exclusion and PIC values are shown in Table 3.The combined probability of parentage exclusion for the 10 microsatellites was 0.9989.

Discussion
Accurate cattle pedigree information is essential for the optimal development of breed and selection programs, improving productivity in the animal industry.Misidentification of parentage can lead to breeding inaccuracy, causing great financial losses in herd management and in the beef industry.Geldermann et al. (1986) estimated misidentification rates of 13% using blood group factors and biochemical polymorphisms in cattle.Ron et al. (1996) found a 5% misidentification rate using microsatellite analysis in Israeli dairy cattle.Rosa (1997)   with high polymorphism information content would help to correctly identify individual cattle, allowing for the better operation of cattle breeding programs.
Little information is available regarding the allele frequencies of the ten microsatellites studied in this work and the other variability estimates for Nellore cattle and, to date, there are no estimates of the multiplex variability in Brazilian cattle breeds.Since the evaluation of polymorphism is strictly dependent on the allele number and the frequency distribution of the alleles, estimates of allele frequencies are essential.
A comparison of the results obtained for B. indicus Nellore cattle with those of B. taurus breeds indicated a difference in variability for some loci, which are highly informative in B. taurus but less informative in B. indicus (Nellore).The exclusion probability values for the TGLA227, ETH10 and TGLA53 loci of taurine cattle described by Peelman et al. (1998) and Heyen et al. (1997) are much higher than that of Nellore cattle.
According to Peelman et al. (1998), who analyzed Belgium cattle, the number of TGLA53 locus alleles in Holstein Friesian (13 alleles), Belgian Red Pied (12 alleles), East Flemish (12 alleles) and Belgian Blue (10 alleles) cattle were very similar to that found in Nellore cattle (13 alleles).However, we found that the exclusion probability for the TGLA53 locus in Brazilian Nellore (EP = 0.256) cattle is much lower than in the four Belgian breeds (Holstein Friesian = 0.742, Belgian Red Pied = 0.711, East Flemish = 0.698 and Belgian Blue = 0.682).We obtained similar results for the TGLA227 locus (EP 0.230), much lower than the values described by Heyen (1997) for Holstein (0.69), Red Angus (0.63) and Gelbvieh (0.68) cattle.Thus, the effectiveness of these markers in European B. taurus cattle is not always the same for Indian B. indicus zebu (Brahman, Nellore) cattle.The substitution of the markers with low variability values for others with improved EP values could render this multiplex more efficient for pedigree verification and individual identification in Nellore.Characterization of Brazilian cattle breeds with microsatellite loci is useful to identify informative markers for each breed, optimizing parentage tests along with the variability values of each marker, thus using the least number of markers with higher levels of information while simultaneously facilitating genotypic identification.
The combined exclusion probability value for the 10 loci was 0.9989, an acceptable value more than ideal for parentage tests (Baron et al, 2002).Jia et al (2004) showed that the CEP value was 0.9957 for Holstein Friesian cattle using six microsatellite markers, while Radko et al (2002) obtained a CEP value of 0.9999 using 11 microsatellites and it is known that the CEP values found in Nellore cattle is lower than that of other taurine breeds (Heyen et al., 1997).
In our study we found significant (p < 0.01) deviations from HWE for six loci (TGLA122, INRA023, TGLA53, ETH10, ETH225 and ETH3).Machado et al (2003) also found significant deviations from HWE for Nellore, Gyr and Guzerat cattle breeds using microsatellites markers.Almeida et al (2000) found that the TGLA122 locus was in HWE in Brazilian hybrid bovine breed (5/8 Aberdeen Angus x 3/8 Nellore).We found deviations from HWE caused by heterozygote deficiency at the TGLA122, INRA023, TGLA53, ETH10 and ETH225 loci. Beja et al. (2003) and Loftus et al. (1999) found deviations from HWE in other European bovine populations, also caused by a heterozygosity deficit, and similar results have been reported by Loftus (1999) in six populations, including Indian Nellore cattle.
Several factors can lead to heterozygote deficiency, including null alleles, assortive mating, the Wahlund effect, selection against heterozygotes, inbreeding, or a combination of these.Null alleles are alleles that are not amplified (usually due to a mutation in one of the primer binding sites) and are commonly reported in microsatellite studies as being the source of heterozygosity deficit (Pemberton et al., 1995).The frequency of microsatellite loci containing null alleles has proved to be as high as 30% in humans (Callen et al., 1993).In paternity tests, an undetected null allele may have profound consequences, since it may cause rejection of an otherwise correctly assigned parent (Holm et al., 2001).
To date, there are no reports of studies on Nellore cattle indicating the presence of null alleles for the markers analyzed, although the presence of null alleles has previously been observed in segregation analyses using other microsatellite loci in Nellore cattle (Tambasco et al., 2000).This hypothesis cannot be excluded because segregation analysis using the loci evaluated in this study has not yet been undertaken for Nellore cattle.
Despite the paucity of information provided by some of the loci analyzed in this study, the use of this multiplex analysis proved efficient in Nellore characterization and can be used in pedigree verification.

Table 1 -
Details of the ten microsatellite loci analyzed.

Table 3 -
Variability measures of ten microsatellite markers in Brazilian Nellore cattle.