Comparison between cluster analyses from RFLP and pedigree data of inbred lines related to Stiff Stalk Synthetic heterotic group

--P. Ajmone Marsan, C. Livini, M. M. Messmer1, E. Melchinger1, P. Franceschini, G. Monfredini and M. Motto

1Univ. Hohenheim, Stuttgart

Several criteria have been used extensively to characterize important inbred lines and hybrids of maize. During the past decade a new class of molecular markers, named restriction fragment length polymorphisms (RFLPs), have been introduced. They allow the detection of differences between genotypes at the DNA level (Helentjaris, T et al., Plant Mol. Biol. 5:109-118, 1985). Moreover, RFLPs have been used to produce well populated linkage maps (Helentjaris, T, Trends Genet. 3:217-221, 1987; Burr et al., Genetics 118:519-526, 1988) permitting a more detailed sampling of the genome. Objectives of this study were: 1) to determine the utility of RFLPs for estimation of genetic similarities among 16 maize inbred lines from Iowa Stiff Stalk Synthetic (BSSS) heterotic group and 2) to compare genetic similarities based on molecular markers with those based on pedigree information.

Table 1. Inbreds used in the analysis of RFLP and pedigree data.
Linesa Backgroundb
B14A (CuzcoxB148)rust. res. sel.c
B68 (41.2504BxB143) sel.
CM109 V3xB142
Lo950 Pioneer3183
Lo951 Pioneer3183
Lo999 (B37xteosinte)xB73
N28 Stiff Stalk Synthetic
A1 50% B14
A2 50% A1
A3 Commercial hybrid
A4 Commercial hybrid
A5 B3 recovered sel.
A8 Commercial hybrid
aLines designated by A1 . . . A10 are private property.

bAnonymous (1989), Hallauer et al. (1983), Henderson (1984), and Guarneri (1990, personal communication).

cPower refers to m-1, where m is the number of backcross generations to the respective parent.

Sixteen BSSS maize inbred lines (Table 1) were chosen as selected directly from the BSSS population (B37, B73, B84, N28), or derived from lines originating from BSSS (B14A, B68, CM109, Lo999, A1, A2). A5, a selection recovered from B3, is not a direct derivative of the previous inbreds but has Reid Yellow Dent in its pedigree. Lo950, Lo951, A3, A4, A8 were derived from commercial hybrids between BSSS and Lancaster Sure Crop lines, yet based on their breeding behaviour they were classified to the BSSS heterotic group. These 16 lines were analyzed with 149 clone-enzyme combinations for their RFLP profile according to published protocols (Livini, C et al., Theor. Appl. Genet., 1992, in press). For each inbred line, genomic DNA was separately digested with two restriction enzymes (EcoRI and HindIII). Genetic similarity between pairs of lines was calculated from molecular data according to Dice similarity coefficient (GS) (Dice, Ecology 26:297-302, 1945). The coancestry coefficient (f) (Malecot, Les mathematiques de l'hérédité, Masson et Cies, Paris, 1948) was used to calculate the genetic distance between pairs or lines from pedigree data. Pedigree information of these lines was obtained primarily from Henderson (Maize Research and Breeders' Manual, No. 10, Illinois Foundation Seeds Inc., Champaign, IL, 1984), Bertolini et al. (Maydica 36:87-106, 1991), or from maize breeders working with these materials. Collected data were subjected to cluster analysis by using the average linkage (UPGMA) method. Simple correlations (r) were calculated between GS and f values for related pairs of lines (f>0.10).

The dendrogram of BSSS lines obtained from cluster analysis of RFLP data is shown in Figure 1. The BSSS lines clustered in two main groups: B14A-related lines (B14A, B68, A1, CM109, A2); and a group of nine lines subdivided into three subclusters, (a) B37 and Lo999, (b) B73 and lines derived from commercial hybrids (A3, A4, Lo950, Lo951, A8), and (c) B84, A5 and N28, did not cluster to any of these groups.

Cluster analyses based on pedigree (Fig. 2) and on RFLP (Fig. 1) data revealed similar associations among lines with minor discrepancies. RFLPs associated inbred lines according to known pedigrees (Table 1). Furthermore molecular analysis was able to classify inbreds derived from commercial hybrids having unavailable pedigrees. The correlation between GS and f values for 18 pairs of related (f>0.10) BSSS lines was highly significant with r=0.70.

The two indices are both valid for revealing relationships among the BSSS lines assayed. However, f values based on pedigree information explained only 50% of the variation of RFLP based GS values, suggesting that the latter offers a better tool for accurate quantification of the degree of relatedness between lines of known, uncertain or even unknown pedigree.

Figure 1. Association among BSSS related inbreds revealed by UPGMA cluster analysis of RFLP data.

Figure 2. Association among BSSS related inbreds revealed by UPGMA cluster analysis of pedigree data.

Please Note: Notes submitted to the Maize Genetics Cooperation Newsletter may be cited only with consent of the authors

Return to the MNL 66 On-Line Index
Return to the Maize Newsletter Index
Return to the Maize Genome Database Page