Polymorphism of locus DRB 3 . 2 in populations of Creole Cattle from Northern Mexico

The polymorphism of locus BoLA-DRB3.2 of the Major Histocompatibility Complex was evaluated in two northern Mexican Creole cattle populations, Chihuahua (n = 47) and Tamaulipas (n = 51). The BoLA-DRB3.2 locus was typed by amplification and digestion with restriction endonuclease enzymes (PCR-RFLP). Fifty-two alleles were detected (28 previously reported and 24 new ones). In the Chihuahua population, 18 alleles and 5.5 effective alleles were found, while in the Tamaulipas population there were 34 and 10.8, respectively. The allele frequencies ranged from 0.011 to 0.383 in Chihuahua and from 0.010 to 0.206 in Tamaulipas. The frequencies of the new alleles in both cattle populations were low (0.010 to 0.053). The expected heterozygosity was 0.827 and 0.916, respectively, for the Chihuahua and Tamaulipas populations. Both populations presented a heterozygote deficit: [Chihuahua FIS = 0.1 (p = 0.019) and Tamaulipas FIS = 0.317 (p < 0.001)]. In conclusion, this study showed that the Mexican Creole cattle have many low-frequency alleles, several of which are exclusive to these populations. Genetic distances obtained show that the Mexican Creole cattle population is composed of independent populations, far apart from other South American Creole populations.


Introduction
The Mexican Creole cattle originated from Bos taurus populations brought to Mexico by the Spaniards in the XVI century, mainly from the Iberian peninsula.Currently, the Mexican Creole cattle population consists of unselected populations with low productivity features, such as low body weight and low milk production.On the other hand, they seem to have the ability of surviving in extreme climates and of resisting to diseases.These cattle are seriously endangered by the rapid displacement due to improved modern breeds, but some populations still remain in remote areas in South, Central and North Mexico.In northern Mexico, Creole cattle are raised by small communities under precarious extensive conditions (low-input and lowmanagement) in a semi-arid climate.In the state of Chihua-hua, they are reared by peasants and by indigenous breeders of the Sierra Tarahumara, ever since their original introduction by Jesuit priests in 1667 (Ríos, 2004).In the state of Tamaulipas, similarly, the Creole cattle are raised by small farmers, although the breed was introduced from different regions of the country, mainly by a genetic rescue program that the Mexican government started around 1980.
Certain populations of Mexican Creole cattle have become a biotype of commercial interest for rodeo sporting events in the USA and Canada, because they possess features required for this activity.These features are hardiness, docility, speed, as well as tolerance to physical punishment and orientation, shape, length, thickness, and strength of the horns (Félix Portillo et al., 2006).
One of the characteristics that have probably evolved due to natural selection is the resistance to infectious diseases, because the Creole cattle are usually not immunized.The Major Histocompatibility Complex (MHC) is one of the main components of the immune system.Also called bovine lymphocyte antigen (BoLA) complex in cattle, it is located on chromosome 23 and contains some genes involved in antigen presentation necessary to trigger the immune response (Hess et al., 1999).The class II BoLA-DR region is composed of one DRA locus and at least three DRB loci (DP, DQ, and DR) (Muggli-Cockett and Stone, 1989).The MHC allele diversity is associated with the ability to recognize a large number of antigens, resulting in a more efficient immune response (Behl et al., 2007).So it may be interesting to assess the level of allele diversity in the BoLA-DRB3.2gene in a population that is under great pressure for survival under tough conditions, such as the Mexican Creole cattle populations.In addition, this polymorphism can be used to study the genetic relationships between populations and to assess their levels of genetic differentiation.Presently, extensive information is available on the levels of genetic diversity of exon 2 of the DRB3 gene in different populations of cattle, obtained by amplification of this segment by PCR, and subsequent digestion with endonucleases (Gilliespie et al., 1999;Kelly et al., 2003;Martinez et al., 2005;Behl et al., 2007).The objective of this study was to investigate the level of genetic diversity present in the MHC BoLA-DRB3.2locus in two Creole cattle populations from northern Mexico.

Materials and Methods
Samples were obtained from 98 Creole Bos taurus cattle from two northern states of Mexico: Chihuahua State samples were obtained from three localities (Morelos: n = 26, Chinipas: n = 7, and Guachochi: n = 14), all located in Sierra Tarahumara.All of the Tamaulipas samples came from a particular region (Soto la Marina; n = 51).

Amplification of BoLA-DRB3
DNA was purified from blood samples following standard procedures (Sambrook et al., 1989).A 284-bp segment of DNA containing the BoLA-DRB3.2gene was amplified by PCR.For this purpose, we used a pair of oligonucleotides described previously by van Eijk et al.

PCR-RFLP analysis
Three aliquots of 60 μL were digested independently, each containing 10 μL of the PCR-amplified products, 1X of the recommended digestion buffer, and 5 U of each of the restriction enzymes RsaI, BstYI, and HaeIII (New England Biolabs Inc., Mass., USA), and incubated at the recommended temperature for 14 h.As a control of BstYI complete digestion, 10 μL of the digestion mix were added to 160 ng of pBR322 DNA and incubated in parallel.

Polyacrylamide gel electrophoresis
The digested products were resolved by vertical gel electrophoresis (8 x 10 cm) on an 8% polyacrylamide gel (Sambrook et al., 1989), and stained with silver nitrate using a commercial kit (BIO-RAD Silver Stain Plus 161-0448).As a molecular weight marker, pBR322 DNA digested with MspI was used.The size of the digestion products was estimated with the RFLP Molecular Weight Program, version 2.17 (UVP, Inc., Upland, CA, USA).Allele identification from the restriction patterns and its nomenclature have been previously reported by van Eijk et al. (1992), Gelhaus et al. (1995), Maillard et al. (1999), and Gilliespie et al. (1999).

Definition of alleles in heterozygote individuals
The PCR products were cloned into a T vector (pUC19 DNA was digested with EcoRV and T tails added at the 3' end, as described by Sambrook et al., 1989).The cloned products were subjected to PCR and the amplified products digested with restriction enzymes.

Statistical analysis
Statistical analysis was carried out using the GENEPOP v.3.4 program (Raymond and Rousset, 1995a) to calculate allele frequencies, the value F IS (Weir and Cockerham, 1984) that measures the heterozygote deficit (Rousset and Raymond, 1995), and the exact test for population differentiation (Raymond and Rousset, 1995b).Statistical significance was measured by the Markov Chains method (Guo and Thompson, 1992).The allele frequencies used for comparison and genetic analysis were those published for Argentine Creole cattle (Giovambattista et al., 1996), Uruguay Creole cattle (Kelly et al., 2003), Blancorejinegro Creole of Colombia (Martinez et al., 2005), Saavedreño cattle of Bolivia (Ripoli et al., 2004), Holstein and Jersey (Sharif et al., 1998), andZebu Gir (da Mota et al., 2002).The genetic distances were calculated according to the method described by Reynolds et al. (1983), using the Gendist program, implemented in the PHYLIP package (Felsenstein, 1993).Genetic distances were used to construct a neighbor-joining phylogenetic tree (Huson and Bryant, 2006).The unbiased expected heterozygosity or Fernández et al.
gene diversity was calculated from the previously reported frequencies (Nei and Chesser, 1983), and the effective allele number (n e ) was obtained as the inverse of the homozygosity (Nei, 1987).

Results
In the 98 animals tested, we identified 52 BoLA-DRB3.2alleles; 28 had already been reported (van Eijk et al.,1992;Gelhaus et al.,1995;Maillard et al.,1999;Gilliespie et al.,1999;Behl et al., 2007) and 24 were new patterns.The identified alleles and their frequencies are shown in Table 1.In the Chihuahua Creole cattle, we found 18 alleles of which seven were new, and in the Tamaulipas Creole cattle we found 34 alleles, of which 17 were new; these two populations shared seven alleles, of which two were new ones (nbd and sba).
Genetic diversity was 0.83 and 0.92 for Chihuahua and Tamaulipas Creole cattle, respectively.The effective allele number was 5.5 for Chihuahua and 10.8 for Tamaulipas.As shown in The F IS parameter assesses the deviation between expected and observed heterozygosity (Weir and Cockerham, 1984).In our study, the F IS values obtained for the Chihuahua Creole and Tamaulipas populations were 0.1 (p = 0.0196) and 0.317 (p < 0.001), respectively.Positive values mean that both groups present a heterozygote deficit.The F IS values for the three subpopulations of the Chihuahua population (Chinipas, Guachochi and Morelos) were also estimated (-0.063, 0.233 and 0.075, respectively), and only the Morelos population presented heterozygote deficiency at a significant level (p = 0.034).
The F ST value measures the degree of genetic differentiation between populations.This value was F ST = 0.043 between the Chihuahua and Tamaulipas groups, which means that 4% of the genetic variation corresponded to differences between populations, whereas the remaining 96% were due to differences among individuals within each population.These results indicate little genetic differentiation between these two populations.We also performed a genetic differentiation test among the subpopulations of Chihuahua (Morelos, Chinipas and Guachochi) and Tamaulipas (F ST = 0.033).The allele distribution was similar among the three Chihuahua subpopulations (p > 0.05) and between Chinipas and Tamaulipas (p > 0.05); however, it was different between Guachochi and Morelos vs. Tamaulipas (p < 0.001).In order to analyze the genetic relationships of the Mexican Creole populations with other American Creole cattle and with other commercial breed cattle, we evaluated their genetic distances by the Reynolds method, using the allele frequencies reported for Argentina (Giovambattista et al., 1996), Uruguay (Kelly et al., 2003), Colombia (Martinez et al., 2005), Bolivia (Ripoli et al., 2004), Holstein and Jersey breeds (Sharif et al., 1998), and for the non-taurine population Zebu Gir (da Mota et al., 2002).The genetic relationships among the different populations are shown in Figure 1.

Discussion
The results of this study show that Chihuahua and Tamaulipas Creole cattle have a high degree of BoLA-DRB3.2polymorphism, in concordance with findings in other populations, such as the American Creole cattle (Giovambattista et al., 1996;Kelly et al., 2003;Ripoli et al., 2004;Martinez et al., 2005), European taurine breeds (Sharif et al., 1998), Asian taurine breeds (Takeshima et al., 2003), and the Zebu Gir breed (da Mota et al., 2002;Behl et al., 2007).In this study, allele DRB3.2*24 showed the highest frequency (0.38) in Chihuahua Creole cattle, and it has also been reported to be present at frequencies higher than 0.1 in all taurine populations (Giovambattista et al., 1996;Sharif et al., 1998;Gilliespie et al., 1999;Kelly et al., 2003).This allele was further reported to be associated with a higher antibody-mediated immune response and with a lower cell-mediated immune response (Rupp et al., 2007).In the Tamaulipas population, the most frequently found allele was DRB3.2*23 (0.206).This allele has been detected in almost all populations of taurine cattle (Giovambattista et al., 1996;Sharif et al., 1998;Gilliespie et al., 1999), but not in Uruguayan Creole cattle (Kelly et al., 2003).The high frequency of alleles DRB3.2*23 and DRB3.2*24 found in Tamaulipas and Chihuahua Creole cattle may be the result of positive natural selection, but this remains to be demonstrated.
It is known that the MHC genes are highly polymorphic in the populations studied so far (Dietz et al., 1997).Although we found a high number of alleles in Mexican populations, the Chihuahua Creole cattle had the lowest effective allele number (5.5), even lower than that of Holstein (7.4) and Jersey (8.2) populations, which are known to be undergoing intense selection and inbreeding.Actually, in the Mexican Creole cattle many alleles at low frequencies were found, but presenting some exclusiveness.In fact, considering the ratio between the effective allele number and the total number of alleles, the Argentine and Uruguayan Creole showed a high relationship (0.8 and 0.72, respectively), whereas in the Chihuahua and Tamaulipas Creole cattle this relationship was very low (0.3 and 0.317, respectively).This may be the result of natural selection on a particular BoLA allele pool, leading to the development of resistance to local infectious diseases.Of the novel alleles reported in this study, sequence analysis was made in 10 of them by Félix Portillo et al. (2006), who found that two of these corresponded to the previously reported alleles DRB3*1602 and DRB3*1501.The remaining eight alleles rendered unique nucleic acid sequences after a search in the official BoLA database.Even though the nucleotide sequencing of BoLA genes is a more precise way of allele identification and comparison, for laboratories with limited resources the PCR-RFLP technique is a more useful and effective tool for BoLA-DRB3.2genotyping, because it is easier to perform, cheaper and faster.
In the Chihuahua and Tamaulipas Creole cattle, we observed an excess of homozygosity that may be due to the combined effect of inbreeding and the Wahlund effect.The Chihuahua Creole cattle animals belonged to three subpopulations from the Sierra Tarahumara, which are isolated and have only few sires available.In turn, the Tamaulipas Creole cattle, although the population descended from animals from different regions of Mexico, are currently confined and consist of small, closed herds.
Finally, Figure 1 presents a dendrogram that summarizes the genetic distances and shows that the Mexican Creole cattle is an independent branch, far apart from the South American Creole populations.This may indicate that this population has a different origin from those populations, or that it has experienced extreme genetic differentiation, probably due to a founder effect.It is also evident that there is extensive differentiation among the different populations of Creole cattle in America, although the Argentinean and Uruguayan populations are closer related, being placed in a common node.Furthermore, although the Chihuahua and Tamaulipas Creole populations are in the same branch, there is a great genetic distance between them.This may be the result of the ancient isolation and independent origin of the Chihuahua group.This information is important for the knowledge of this endangered genetic resource and for the establishment of strategic conservation programs.Furthermore, our findings suggest that these local cattle populations may be a valuable reservoir of genetic diversity for the development of defined local breeds or in the improvement of commercial cattle lines.

Figure 1 -
Figure 1 -Dendrogram depicting the genetic distances among Chihuahua and Tamaulipas Creole cattle, South American Creole cattle and taurine and Zebu breeds.

Table 1 -
Allele distribution and gene frequency of locus DRB3.2 in Creole cattle and in improved breeds.