Distribution of BoLA-DRB3 alleles between selected Myanmar cattle breeds
PCR-SBT genotyping allowed us to identify 72 BoLA-DRB3 alleles (69 previously reported variants and three new alleles; Table 1) for the breeds selected in this study. The number of alleles (na) was 57 in Pyer Sein cattle (54 previously reported and three new), 43 in Shwe Ni (41 previously reported and two new), and 33 in the Holstein-Friesian crossbreed (32 previously reported and one new) (Tables 1 and 2). Nucleotide and predicted amino acid sequences of the three new allele variants are shown in Fig. 1.
Table 1 BoLA-DRB3 allele frequencies in Myanmar cattle breeds
BoLA-DRB3 alleles
|
Pyer Sein
|
|
Shwe Ni
|
|
Holstein-Friesian Crossbreed
|
(N = 163)
|
|
(N = 69)
|
|
(N = 62)
|
BoLA-DRB3*001:01
|
1.53
|
|
1.45
|
|
6.45
|
BoLA-DRB3*002:01
|
3.68
|
|
6.52
|
|
6.45
|
BoLA-DRB3*002:02
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*003:01
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*003:02:01
|
1.23
|
|
0.00
|
|
0.00
|
BoLA-DRB3*005:01
|
0.31
|
|
0.00
|
|
4.84
|
BoLA-DRB3*005:02
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*005:03
|
0.00
|
|
0.72
|
|
0.00
|
BoLA-DRB3*005:04
|
0.00
|
|
0.00
|
|
0.81
|
BoLA-DRB3*006:01
|
0.00
|
|
0.00
|
|
0.81
|
BoLA-DRB3*007:01
|
2.15
|
|
2.17
|
|
0.81
|
BoLA-DRB3*008:01
|
0.00
|
|
0.00
|
|
1.61
|
BoLA-DRB3*009:01
|
4.29
|
|
0.72
|
|
0.81
|
BoLA-DRB3*009:02
|
1.84
|
|
2.90
|
|
4.84
|
BoLA-DRB3*010:01
|
2.76
|
|
3.62
|
|
4.84
|
BoLA-DRB3*011:01
|
0.00
|
|
0.00
|
|
11.29
|
BoLA-DRB3*011:03
|
0.92
|
|
2.90
|
|
0.00
|
BoLA-DRB3*011:04
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*012:01
|
1.53
|
|
1.45
|
|
11.29
|
BoLA-DRB3*013:01
|
1.53
|
|
2.17
|
|
1.61
|
BoLA-DRB3*014:01:01
|
3.68
|
|
1.45
|
|
4.03
|
BoLA-DRB3*013:02
|
0.61
|
|
0.00
|
|
0.00
|
BoLA-DRB3*013:03
|
0.00
|
|
0.72
|
|
0.00
|
BoLA-DRB3*015:01
|
2.76
|
|
1.45
|
|
3.23
|
BoLA-DRB3*016:01
|
2.15
|
|
2.17
|
|
2.42
|
BoLA-DRB3*016:02
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*017:01
|
0.00
|
|
0.00
|
|
3.23
|
BoLA-DRB3*017:03
|
6.13
|
|
5.07
|
|
0.00
|
BoLA-DRB3*018:01
|
1.53
|
|
7.97
|
|
0.00
|
BoLA-DRB3*01901
|
2.15
|
|
1.45
|
|
1.61
|
BoLA-DRB3*020:01:02
|
0.00
|
|
0.72
|
|
5.65
|
BoLA-DRB3*020:02
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*20:03
|
0.61
|
|
0.00
|
|
0.00
|
BoLA-DRB3*20:05
|
0.00
|
|
0.72
|
|
0.00
|
BoLA-DRB3*022:01
|
6.13
|
|
5.07
|
|
2.42
|
BoLA-DRB3*025:01:01
|
2.45
|
|
0.72
|
|
1.61
|
BoLA-DRB3*025:01:02
|
0.31
|
|
0.72
|
|
0.00
|
BoLA-DRB3*025:02
|
0.00
|
|
0.00
|
|
0.81
|
BoLA-DRB3*026:01
|
4.60
|
|
5.80
|
|
3.23
|
BoLA-DRB3*027:03
|
2.45
|
|
2.90
|
|
0.00
|
BoLA-DRB3*027:05
|
0.00
|
|
0.00
|
|
0.00
|
BoLA-DRB3*027:07
|
0.31
|
|
2.17
|
|
0.00
|
BoLA-DRB3*027:10
|
0.61
|
|
0.00
|
|
0.00
|
BoLA-DRB3*028:01
|
3.68
|
|
2.90
|
|
0.00
|
BoLA-DRB3*028:02
|
0.61
|
|
0.72
|
|
1.61
|
BoLA-DRB3*030:01
|
1.84
|
|
2.17
|
|
0.81
|
BoLA-DRB3*31:01
|
6.75
|
|
1.45
|
|
3.23
|
BoLA-DRB3*31:03
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*033:01
|
0.61
|
|
0.72
|
|
0.81
|
BoLA-DRB3*34:01
|
2.45
|
|
0.00
|
|
0.00
|
BoLA-DRB3*034:03
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*035:01
|
1.53
|
|
2.17
|
|
0.00
|
BoLA-DRB3*036:01
|
1.23
|
|
0.72
|
|
0.00
|
BoLA-DRB3*037:01
|
0.61
|
|
0.00
|
|
0.00
|
BoLA-DRB3*038:01
|
0.92
|
|
0.00
|
|
0.81
|
BoLA-DRB3*039:01
|
1.84
|
|
0.00
|
|
0.00
|
BoLA-DRB3*041:01
|
1.53
|
|
2.90
|
|
0.00
|
BoLA-DRB3*042:01
|
0.61
|
|
0.00
|
|
1.61
|
BoLA-DRB3*043:01
|
3.37
|
|
1.45
|
|
4.03
|
BoLA-DRB3*043:02
|
0.00
|
|
0.00
|
|
0.81
|
BoLA-DRB3*048:02
|
1.53
|
|
5.07
|
|
0.00
|
BoLA-DRB3*049:01
|
0.00
|
|
0.72
|
|
0.00
|
BoLA-DRB3*50:01
|
0.31
|
|
0.00
|
|
0.00
|
BoLA-DRB3*50:11
|
0.00
|
|
0.72
|
|
0.00
|
BoLA-DRB3*57:02
|
0.61
|
|
2.17
|
|
0.00
|
BoLA-DRB3*58:01
|
0.92
|
|
0.00
|
|
0.00
|
BoLA-DRB3*63:01
|
0.00
|
|
0.72
|
|
0.00
|
BoLA-DRB3*64:02
|
0.61
|
|
0.00
|
|
0.00
|
BoLA-DRB3*72:01
|
1.53
|
|
3.62
|
|
0.00
|
BoLA-DRB3*73:01
|
1.53
|
|
2.17
|
|
0.81
|
BoLA-DRB3*079:01
|
3.68
|
|
4.35
|
|
0.81
|
BoLA-DRB3*080:01
|
0.61
|
|
1.45
|
|
0.00
|
aN, number of animals analyzed; bFrequent alleles in each breed are indicated in bold (>5%); cNovel alleles identified in this study are indicated in bold and underlined.
Table 2 Sample size (N), number of alleles (na), observed (ho) and expected (he) heterozygosity, Hardy Weinberg equilibrium (HWE) measured through FIS and Slatkin's exact test in the cattle breeds studied. FIS p-values are indicated between parentheses. Significant p values are indicated in bold.
Breed
|
N
|
na
|
ho
|
he
|
HWE
|
Slatkin's p value
|
FIS (p value)
|
Pyer Seina
|
163
|
58
|
0.94
|
0.97
|
0.0276 (0.0534)
|
0.006
|
Shwe Nia
|
69
|
43
|
0.86
|
0.97
|
0.1197 (< 0.0001)
|
0.010
|
Holstein-Friesian crossbreeda
|
62
|
33
|
0.94
|
0.95
|
0.0183 (0.7794)
|
0.149
|
Yacumeño Creole b
|
112
|
35
|
0.92
|
0.95
|
0.0344 (0.7363)
|
0.001
|
Hartón del Valle Creole b
|
66
|
24
|
0.97
|
0.94
|
-0.0360 (< 0.0001)
|
0.138
|
Bolivian Nelore c
|
116
|
26
|
0.78
|
0.87
|
0.0990 (0.6921)
|
0.306
|
Bolivian Gir c
|
110
|
19
|
0.88
|
0.92
|
0.0406 (0.0926)
|
0.009
|
Nellore x Brahman c
|
195
|
33
|
0.76
|
0.86
|
0.1131 (0.1985)
|
0.473
|
Japanese Holstein b
|
102
|
18
|
0.92
|
0.90
|
-0.0215 (0.4481)
|
0.091
|
Japanese Shorthorn b
|
100
|
20
|
0.92
|
0.91
|
-0.0086 (0.0692)
|
0.069
|
Japanese Jersey b
|
69
|
14
|
0.91
|
0.89
|
-0.0297 (0.0023)
|
0.042
|
Japanese Black b
|
200
|
23
|
0.91
|
0.91
|
0.0095 (0.4043)
|
0.004
|
Chilean Hereford e
|
49
|
15
|
0.82
|
0.87
|
0.0574 (0.4980)
|
0.582
|
Chilean Black Angus e
|
100
|
26
|
0.61
|
0.90
|
0.3246 (< 0.0001)
|
0.447
|
Chilean Red Angus e
|
99
|
29
|
0.71
|
0.93
|
0.2415 (< 0.0001)
|
0.080
|
Philippine Native d
|
482
|
71
|
0.91
|
0.96
|
0.0480 (< 0.0001)
|
0.068
|
Philippine Brahman d
|
236
|
58
|
0.89
|
0.95
|
0.0687 (< 0.0001)
|
0.134
|
aPresent work; bGiovambattista et al., 2013; cTakeshima et al., 2018; dTakeshima et al., 2003; eTakeshima et al., 2015; and fPolat et al., 2014.
The three new variants were assigned allele names by IPD-MHC, namely, BoLA-DRB3*002:02, which differs from BoLA-DRB3*002:01 at 147 positions; BoLA-DRB3*079:01, which differs from DRB3*001:01 at seven positions (76, 108, 129, 206, 218, 220 and 254); and BoLA-DRB3*080:01, which differs from BoLA-DRB3*003:01 at five positions (108, 157, 173, 199 and 207). All three new BoLA-DRB3 allele variants shared about 90-94% and 83.7-90.7% nucleotide and amino acid similarity with the BoLA-DRB3 cDNA clone NR1 (correspond to GenBank accession number D45357 and the allele BoLA-DRB3*016:01), respectively [5].
A Venn diagram was constructed using data obtained in this study and from previous reports [18,23,29] which include 102 BoLA-DRB3 alleles. Data were grouped in terms of the breed’s geographical origin as follows: Myanmar native breeds (Pyer Sein and Shwe Ni) and Holstein-Friesian crossbreed; Asian native (Philippine native and Japanese Black); Zebu (Bolivian Nellore, Bolivian Gir, Peruvian Nellore-Brahman and Philippine Brahman); and European (Chilean Hereford, Chilean Black Angus, Chilean Red Angus, Japanese Jersey, Japanese Shorthorn and Japanese Holstein) breeds. Six alleles were not present in any of these breed groups. This analysis revealed that out of the 96 alleles identified in the five cattle groups, only nine were detected in the Myanmar native breeds (Fig. 2a), three of which exhibited gene frequencies that were higher than 0.5% (Fig. 2b). Two other variants were only present in Myanmar native breeds and the Holstein-Friesian crossbreed. Together, these 11 alleles represent about 15% of the 73 alleles detected in the Myanmar native cattle and Holstein-Friesian crossbreed. Twenty-six other alleles were only found in Myanmar cattle populations and Asian native or Zebu breeds, or a combination of these groups. In addition, the BoLA-DRB3 NJ tree including all the previously reported alleles and the three new variants ones showed that the variants detected in Myanmar cattle populations were interspersed among the various clusters (Fig. 3).
Three, six and five alleles appeared with frequencies of > 5% in Pyer Sein, Shwe Ni and Holstein-Friesians, respectively. Three of these high-frequency (> 5%) alleles (BoLA-DRB3*002:01, *017:03 and *022:01) were common in at least two out of three Myanmar populations (Table 1). These common alleles accounted for a low proportion of the cumulative gene frequencies (19.02, 34.78 and 43.13% in Pyer Sein, Shwe Ni and Holstein-Friesians, respectively), revealing an even gene frequency distribution (Fig. S1).
Nucleotide and amino acid diversity in the BoLA-DRB3 alleles found in Myanmar cattle populations
The results of genetic diversity at the DNA and amino acid levels obtained for Myanmar cattle breeds and for breeds previously reported are shown in Table 3. The π values within Pyer Sein, Shwe Ni and Holstein-Friesian crossbreed were 0.090, 0.080 and 0.080, respectively, while the mean number of pairwise differences was 20.96, 17.89 and 20.09, respectively. These nucleotide diversity values all fall within the upper end of the range reported (πrange = 0.068-0.083; NPDrange = 16.31-20.04) for other bovine breeds when using PCR-SBT genotyping methods [18,22,23,26,29]. The average dN and dS substitutions in Myanmar cattle breeds was calculated across BoLA-DRB3 exon 2 and the antigen-binding site (ABS). As expected, the dN/dS ratio was higher when only the ABS was analyzed. As shown in Table 3, the values obtained in Myanmar cattle were similar to those estimated for other cattle breeds (dN/dS total = 0.054–0.067; dN/dS ABS = 0.247–0.282).
Table 3 Nucleotide diversity (π), mean number of pairwise differences (NPD) and mean number of non-synonymous (dn) and synonymous (ds) nucleotide substitutions per site. ABS = antigen-binding site.
Breed
|
π
|
NPD
|
Total
|
ABS
|
|
|
|
ds
|
dn
|
dn / ds
|
ds
|
dn
|
dn / ds
|
Pyer Seina
|
0.090
|
20.96
|
0.038
|
0.096
|
2.53
|
0.127
|
0.387
|
3.05
|
Shwe Nia
|
0.080
|
17.89
|
0.039
|
0.098
|
2.51
|
0.122
|
0.397
|
3.25
|
Holstein-Friesian crossbreeda
|
0.080
|
20.09
|
0.043
|
0.099
|
2.30
|
0.141
|
0.4
|
2.84
|
Yacumeño b
|
0.078
|
19.59
|
0.036
|
0.099
|
2.75
|
0.128
|
0.391
|
3.05
|
Hartón del Valle b
|
0.076
|
19.00
|
0.029
|
0.096
|
3.31
|
0.109
|
0.386
|
3.54
|
Bolivian Nellore c
|
0.070
|
17.54
|
0.035
|
0.097
|
2.77
|
0.117
|
0.388
|
3.32
|
Bolivian Gir c
|
0.078
|
19.45
|
0.038
|
0.096
|
2.53
|
0.133
|
0.385
|
2.89
|
Nellore x Brahman c
|
0.068
|
16.95
|
0.039
|
0.097
|
2.49
|
0.128
|
0.376
|
2.94
|
Japanese Holstein d
|
0.079
|
19.86
|
0.038
|
0.096
|
2.53
|
0.132
|
0.393
|
2.98
|
Japanese Shorthorn d
|
0.079
|
19.80
|
0.041
|
0.097
|
2.37
|
0.128
|
0.41
|
3.20
|
Japanese Jersey d
|
0.073
|
16.31
|
0.041
|
0.099
|
2.41
|
0.122
|
0.402
|
3.30
|
Japanese Black d
|
0.071
|
18.56
|
0.043
|
0.096
|
2.23
|
0.139
|
0.365
|
2.63
|
Chilean Hereford e
|
0.070
|
17.41
|
0.033
|
0.098
|
2.97
|
0.112
|
0.46
|
4.11
|
Chilean Black Angus e
|
0.077
|
19.17
|
0.037
|
0.097
|
2.62
|
0.123
|
0.385
|
3.13
|
Chilean Red Angus e
|
0.080
|
20.03
|
0.041
|
0.097
|
2.37
|
0.124
|
0.389
|
3.14
|
Philippine Brahman f
|
0.080
|
20.04
|
0.04
|
0.096
|
2.40
|
0.133
|
0.379
|
2.85
|
Philippine Native f
|
0.083
|
19.60
|
0.036
|
0.096
|
2.67
|
0.12
|
0.381
|
3.18
|
a Present work; b Giovambattista et al., 2013; c Takeshima et al., 2018; d Takeshima et al., 2003; e Takeshima et al., 2015; and f Polat et al., 2014.
Gene diversity, HWE, and neutrality testing of BoLA-DRB3 variants found in Myanmar cattle populations
As mentioned above, na ranged from 33 in the Holstein-Friesian crossbreed to 58 in the Pyer Sein native breed, while he and ho were both higher than 0.86 in all three populations (Table 2). These indexes are evidence of high diversity values for Myanmar cattle populations, which is similar to the results reported for other bovine breeds which have been evaluated by PCR-SBT, and characteristic of MHC class II DR genes [18,22,23,26,29]. When we evaluated the populations using the HWE test, two of the three Myanmar populations were in equilibrium, while native breed Shwe Ni significantly deviated from its theoretical values (Table 2), probably because of the significant proportions of homozygous animals found in this study (FIS = 0.1197, p < 0.0001). As demonstrated in Table 2, other breeds have also been seen to be in disequilibrium, as a result of an excess or deficit in the proportion of homozygous animals within the population.
It is widely accepted that the genetic diversity of MHC class II genes can be maintained by balancing selection. Thus, we performed a Slatkin’s exact neutrality test (Table 2) to evaluate this phenomenon in our populations. The BoLA-DRB3 gene frequency profile in both the Pyer Sein and Shwe Ni cattle showed an even distribution (p = 0.006 and 0.010, respectively), consistent with the theoretical proportion expected under balancing selection pressures as opposed to positive or neutral selection (p > 0.025). A similarly even BoLA-DRB3 gene frequency was observed in other cattle breeds, including Japanese Black, Yacumeño Creole and Bolivian Gir. Conversely, BoLA-DRB3 gene frequency distributions in the Holstein-Friesian crossbreed were more compatible with neutral selection, which is similar to the results obtained for the majority of the cattle breeds analyzed to date (Table 2).
BoLA-DRB3 genetic structure and levels of population differentiation in Myanmar cattle
The average FST analysis showed a low level of genetic differentiation between Myanmar native breeds (FST = 0.003), similar to those estimated in Holstein populations (0-0.0067) [23]. FST values between the native breeds and the Holstein-Friesian crossbreed varied from 0.019 to 0.021. These values were within the range estimated for differences within Taurine or Zebu breeds and lower than those obtained when comparing breeds from different groups (Fig. 4 and Table S2).
The average FST values across all Myanmar Holstein-Friesian and native breed populations were 0.0136 and 0.0121, respectively (p < 0.001). Significant differences were observed in nine out of the fifteen native and one out of six Holstein-Friesian crossbreed populations (p < 0.05). In addition, FST values for comparisons between Myanmar native breeds ranged between 0.003 and 0.024 and between 0 and 0.031 for Myanmar Holstein-Friesian crossbreed populations (Table S3a and b). As mentioned above, similar genetic distance values were observed among Holstein populations from different countries [23].
Genetic differentiation of BoLA-DRB3 alleles in Myanmar breeds. Comparison with zebu and taurine breeds
First, BoLA-DRB3 allele frequencies from Myanmar and previously reported breeds included in our dataset were used to generate Nei’s DA and DS genetic distance matrices. Then, dendrograms were constructed from these distance matrices using UPGMA and NJ algorithms. All trees revealed congruent topologies, which were consistent with the historical and geographical origin of the breeds. As expected, this tree revealed two main clusters which included most of the Taurine and Zebuine breeds (Fig. 5a), with Japanese Jersey and Chilean Hereford located outside of these clusters. The Holstein-Friesian crossbreed fell into the Taurine cluster, while the Myanmar native breeds were located in a sub-cluster within the Zebuine cluster close to Philippine populations.
Second, we used BoLA-DRB3 allele frequencies to perform three PCA analyses among breeds. In these PCA, the first two components accounted for 41.11% of the data variability. The first PC accounted for 27.12% of the total variance and, as shown in a previous study [29], first component clearly exhibited a differentiation pattern between Zebu (negative values) and Taurine (positive values) breeds, while native breeds from Myanmar and the Philippines were located near the origin of the plot (Fig. 5b). The first PC was primarily determined by differences in the frequency of the same alleles reported by Takeshima et al. [29]. The second PC explained 13.99% of the total variation and showed a gradient among Taurine breeds, with Japanese Black and Japanese Jersey located at opposite ends. Furthermore, this component discriminated between Myanmar and Philippine native breeds. The second PC was identical to PC1 reported in the study mentioned above [29]. Finally, the third PC accounted for 13.64% of the variance and allowed the differentiation of Chilean Hereford cattle from other Taurine breeds. As shown in Fig. 5b and c, the Myanmar Holstein-Friesian crossbreed was located within the Taurine cloud but in an intermediate position between the Japanese Holstein and Myanmar native breeds, supporting the presence of the same level of gene introgression in Myanmar Holstein populations, which is also supported by the presence of unique BoLA-DRB3 alleles within these populations. These PCA results agree with the overall clustering observed after NJ or UPGMA tree construction.
In addition, we analyzed protein pockets (pocket 1, pocket 4, pocket 6, pocket 7, and pocket 9) involved in the antigen-binding function of the MHC complex using PCA. As shown in Fig. S3 a-e, only the distribution pattern of pocket 4 was similar to the PCAs created using the allelic frequency data, while PCA of the remaining pockets did not exhibit a spatial distribution related to the geographical or historical origin of the breeds. The position of the Myanmar native breeds in pocket 4 was the result of positive PC1 and PC2 values for the presence of amino acid motifs GFDQKEV, SYDRENY, SFDREYY, SFDDEAY, KFDRAAY, and GYDREYY (amino acid positions 13, 26, 28, 70, 71, 74 and 78). Also, pocket 1 showed quite similar distribution pattern, but for PC2.
Finally, PCA was performed at the Myanmar population level, to evaluate the impact of the ten sampling sites (six for native breeds and four for Holstein-Friesians) on our results. This analysis showed that Myanmar native populations grouped in a narrow cluster that diverged clearly from the Myanmar Holstein-Friesian crossbred populations, in agreement with the FST analysis described above. Furthermore, PCA showed that some Myanmar native populations (Bago, Mandalay and Yangon) seemed to be closer to Zebu breeds (Nellore, Gir and Brahman), while others (Kayin, Magway and Sagaing) were more closely related to the Philippine native breeds. However, PCA results at the Myanmar population level did not show a clear correlation between the genetic relationship of BoLA-DRB3 alleles and geographical distribution. By contrast, Myanmar Holstein-Friesian populations showed a more dispersed distribution when compared to the compact cloud reported for Holstein populations from other countries [23], which may be the result of the differences in the degree of admixture between these populations.