Amplifications were carried out using 50 ng of DNA and a "touchdown" PCR with annealing temperatures varying from 65 C to 55 C. The PCR products were visualized in 4% agarose/metaphor gels stained with ethidium bromide. In order to facilitate the genotyping of each microsatellite, an "allele pattern" was established among 28 inbreds. This pattern could represent the allele diversity and was organized as a molecular weight marker, which was used later for comparison with the other 67 inbreds.
To select the best primer pairs, three sets of tests were performed: the first among three lines, the second among eight lines and the third among 28 lines. The second was the most efficient of all, being able to reveal the polymorphism between the inbreds and not wasting much effort and time. A total of 193 microsatellite primer pairs were tested, and classifications according to amplification quality and genotyping difficulty were done. Most primer pairs had a good amplification quality; nevertheless, many of them were also classified as "difficult" for genotyping, with little difference in the number of base pairs among the amplification products obtained from the 95 inbreds. From the 193 microsatellites tested, 14 were chosen for the diversity analysis.
Results in Table 1 show the number of alleles for each locus studied and its PIC (or genic diversity) values. Among the 14 SSRs, a mean of 5.7 alleles/locus was observed and the mean for the genic diversity was 0.6. These results show the great discriminatory abilities of the loci analyzed and suggest large diversity among the germplasms being studied. The dendrogram presented in Figure 1 shows the genetic distances as provided by the modified Rogers coefficient (TFPGA and NTSYS/UPGMA 2.02 used). This index was chosen due to its Euclidian behavior and due to the fact that it may be compared to other measures of this kind, since it is calculated using means among the loci studied. Further, it considers allelic frequencies, which are extremely important when working with codominant markers. The cophenetic value was 0.57, which indicates that the schematic representation of the dendrogram was not able to show with confidence the distance matrix provided by the index. Moreover, it was not possible to visualize defined groups within the dendrogram, which may be caused by the insufficient number of loci analyzed until now. In order to provide more confident schematic representation, the genetic distances in the distance matrix were grouped and represented in Figure 2. Notably, the majority of the distances are between 0.6 and 0.9, which indicates great divergence among the inbreds.
The results presented in this work suggest that the tropical germplasm under analysis is an excellent material for hybrid breeding programs. A greater characterization of this germplasm will provide a larger understanding of the potential tropical maize has and will provide quality data for the maize improvement programs in Brazil.
Table 1. Number of alleles obtained for each SSR locus and the respective
PIC values.
| Locus | Number of Alleles | PIC |
| bnlg1621b | 10 | 0.8614 |
| bnlg1724 | 6 | 0.4676 |
| phi022 | 3 | 0.5199 |
| umc1069 | 8 | 0.7076 |
| umc1122 | 4 | 0.5058 |
| umc1221 | 9 | 0.7981 |
| umc1230 | 6 | 0.7898 |
| umc1252 | 2 | 0.3006 |
| umc1357 | 4 | 0.6409 |
| umc1395 | 3 | 0.5203 |
| umc1416 | 2 | 0.4943 |
| umc1639 | 5 | 0.7521 |
| umc1804 | 15 | 0.8754 |
| umc1943 | 4 | 0.3166 |
| Mean | 5.7 | 0.6107 |
Figure 1. Dendrogram showing the genetic distances according to the modified Rogers coefficient and to the UPGMA clustering method. Cophenetic value = 0,5767.
Return to the MNL 76 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage
