|
|
||||||||
a Pioneer Hi-Bred International, Inc., Research Technology Services, 7300 NW 62nd Avenue - PO Box, 1004, Johnston, IA 50131-1004 USA
b Plant Genetic Resources Conservation Unit, USDA/ARS, 1109 Experiment Street, Griffin, GA 30223-1797 USA
c Iowa State Univ., 1553 Agronomy Building, Ames, IA 50011 USA
d Pioneer Hi-Bred International, Inc., Sorghum Research, 501 E. Pioneer Road, Plainview, TX 79072 USA
smiths{at}phibred.com
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
Tao et al. (1993) found that overall levels of polymorphisms among 36 sorghum lines were equally low for profiles obtained by means of 30 RAPD primers or 29 RFLP probes. Vierling et al. (1994) found 73 RAPD primers discriminated among sorghum lines but those data did not allow lines to be associated into groupings that reflected pedigrees. Associations among 34 lines determined by 19 RFLP probes, 21 RAPD primers, and 41 ISSRs were markedly different and dependent on the source of molecular profile data (Yang et al., 1996). Ahnert et al. (1996) reported a study of 105 sorghum inbreds that used 104 RFLP probes which showed higher levels of polymorphism and associations of lines that were congruent with pedigree information and breeder classifications of germplasm. Brown et al. (1996) in an explorative study surveyed diversity among 13 temperate adapted sorghum lines and four diverse genotypes using approximately 30 SSRs. Seventeen primer pairs revealed polymorphisms. Taramino et al. (1997) used 13 SSRs to reveal moderate to high levels of diversity among a group of nine sorghum lines of different racial classification and from different geographic origins.
The limited sampling of germplasm in these previous studies hinders an assessment of the ability of SSRs to detect genetic variation in sorghum. Few (913) inbred lines were included and they may not adequately represent the gene pool of many breeding programs. Furthermore, comparisons of discrimination abilities and associations of germplasm that stem from the use of various molecular profiling methodologies are confounded by the use of different inbred lines. Therefore, the potential utility of SSRs as a molecular profiling technology to aid in research and product development of sorghum remains to be evaluated.
The objective of this study was to determine the potential utility of SSR technology for applications in research, product development, seed production, intellectual property protection (IPP), and genetic resources conservation management for sorghum. To accomplish this goal, we report molecular profile and pedigree data for a set of 50 elite sorghum lines. We assessed the discrimination ability of data obtained from RFLP probes and from SSR loci and we compare associations among inbred lines that are revealed by these data with associations that would be expected on the basis of known pedigrees. We also discuss the cost effectiveness of acquiring SSR data with respect to the potential use of this technology by researchers, breeders, and conservators.
| Materials and methods |
|---|
|
|
|---|
|
In all, 15 primer pairs (multiplexed as three sets of five SSRs) were used for genotyping (Table 2) . These oligonucleotide sequences were derived from SSR-containing clones isolated from a size-fractionated genomic DNA library of S. bicolor cultivar RTx430 (Brown et al., 1996). Primers that amplified target sequences with nonoverlapping fragment sizes, determined on the basis of preliminary estimates from acrylamide gels (Brown et al., 1996), were labeled with the same fluorescent dye, either 6-carboxyfluorescein (6-FAM), tetrachloro-6-carboxyfluorescein (TET), or hexachloro-6-carboxyfluorescein (HEX); primers amplifying overlapping fragment sizes were labeled with different fluorescent dyes. In all pairs, only the "forward" primer was labeled with fluorescent dye (the "reverse" primer was unlabeled).
|
Electrophoresis and Detection
Samples containing 1-µL PCR product, 0.5-µL GeneScan 500 internal lane standard labeled with N,N,N',N'-tetramethyl-6-carboxyrhodamine (TAMRA) (Perkin Elmer/Applied Biosystems), and 50% (v/v) formamide were heated at 92° C for 5 min, placed on ice, then loaded on 4% (w/v) denaturing (6 M urea) acrylamide:bisacrylamide (19:1) gels (36 cm well-to-read). DNA samples were electrophoresed in 1x TBE buffer (89 mM Tris, 89 mM borate, 2 mM EDTA pH 8.3) at constant voltage (3000 V, 51° C) for 3.5 h on an automatic DNA sequencer (Perkin Elmer/Applied Biosystems, model 377) equipped with GeneScan software version 2.1 (Perkin Elmer/Applied Biosystems). Fragment sizes were automatically calculated to two decimal places by the "local Southern" algorithm (Elder and Southern, 1987).
Polymorphism information content (PIC) provides an estimate of the discriminatory power of a locus, or loci, by taking into account, not only the number of alleles that are expressed, but also the relative frequencies of those alleles. PIC values were calculated by the algorithm:
![]() |
PIC values range from 0 (monomorphic) to 1 (very highly discriminative, with many alleles each in equal and low frequency).
Genetic distances between pairs of inbred lines for SSR and RFLP data were calculated from comparisons of the band scores by a modified Nei's distance (Nei and Li, 1979). Pedigree distances between pairs of inbreds were calculated from 1- Malécot's Coefficient of Relatedness (Malécot, 1948). Associations among inbreds were revealed by UPGMA corresponding to Ahnert et al. (1996).
| Results |
|---|
|
|
|---|
Associations among inbred lines that would be expected on the basis of pedigree are presented in Fig. 1 . There is a clear demarcation between most of the lines according to their status as B or R lines. Associations among lines on the basis of their SSR profiles are shown in Fig. 2 . Viewing these associations from the top of the figure, the lines fall into five groups. First, there are six lines (no. 65 to no. 2) that are a mix of R and B genotypes. Second, there is an association of 13 lines, 11 of which are R lines. Third, there is a mixed group of six lines; three of which are closely associated R lines. Fourth, there is a cluster of 13 B lines. Fifth, there is a loose association of the remaining 12 inbreds that include both R and B lines.
|
|
| Discussion |
|---|
|
|
|---|
The 50 lines that were included in this study represent germplasm that is routinely used in sorghum breeding in the USA. The inbreds encompass a relatively broad array of germplasm diversity. For example, the set of germplasm includes lines developed in several different geographic areas (including India, Mexico, South America, the USA and Australia). Maturity ranges from the very early that is adapted to South Dakota to the very late (sub-tropical). Germplasm groups that are represented include Feterita, Zera Zera, Kaura, Kaoliang, Kafir, Redbine, and lines developed from inter-group crosses. Within the groups, there is a variety of kernel colors, plant heights, and maturities. Some of the lines have been derived from well adapted, high performing inbreds while others have been developed from crosses with partially or fully converted lines from the world sorghum collection.
Comparisons of discrimination ability for SSRs and their ability to provide data that are reflective of pedigree backgrounds to those obtained by RFLPs, RAPDs, or ISSRs in the studies reported by Tao et al. (1993), Vierling et al. (1994), and Yang et al. (1996) are difficult to make for they are confounded by the use of different germplasm in each study. RFLPs are reported to reveal more polymorphisms than RAPDs (Vierling et al., 1994) or ISSRs (Yang et al., 1996). In addition, RAPDs and ISSRs can contribute to problems with reproducibility among experiments (Yang et al., 1996) and the meaningful comparison of data among laboratories. Taramino et al. (1997) note that prior to their studies of SSRs on nine diverse sorghum genotypes, isozymes and RFLPs emerge as the most discriminating reproducible molecular profiling method for discriminating sorghum germplasm.
The results provided by Taramino et al. (1997) showed that SSRs could have great potential in discriminating among sorghum inbreds. Information content of the SSR loci was high; data from just one SSR locus could alone allow all inbred lines to be uniquely identified. However, that set of lines does not allow an evaluation of the discrimination ability of SSRs among an elite set of germplasm used routinely in product development nor does it provide the basis for a direct comparison of discrimination ability from RFLP data. The present study allows for both of these analyses.
Forty-eight (96%) of the elite inbred lines included in this study could be uniquely identified by just 15 SSR loci. The analyses of SSR and RFLP data were congruent in identifying similarities between lines that are closely related by pedigree. For example, associations made on the basis of SSR profiles identified the following groups of lines that would be expected on the basis of known pedigrees: (i) Lines 80 and 104; (ii) Lines 88 and 98; (iii) Lines 30 and 22; (iv) Lines 3, 18, and 38; and (v) Lines 2, 10, and 36. But unlike the association of lines from RFLP data, the analysis of SSR data did not result in such a discrete demarcation of B and R lines according to their respective fertility groups. Nevertheless, other associations of lines revealed by SSRs have support from pedigree or from performance information. For example, Lines 45, 65, and 94 were not associated according to RFLP data but their association on the basis of SSR profiles does reflect the fact that these lines are all females with Zera Zera germplasm in their pedigrees. Lines 1, 4, 23, and 69 represent a mix of male and female lines but they are all of early maturity. Lines 62, 73, and 76 are all females with pedigrees that trace to an Australian source that are associated by SSR data but which are not closely associated on the basis of RFLP data. SSRs show an association of Line 90 with Lines 88 and 98 that reflects known pedigrees, whereas RFLP data showed Line 90 to be more different to Lines 88 and 98. Both SSRs and RFLPs showed an association of Lines 39 and 52, which are yellow endosperm lines that were not associated according to pedigree data.
The loci that are represented by these SSRs map to at least nine different linkage groups (Dean et al., 1999). This number of SSR loci, which collectively had a moderately high PIC, provided an ability to identify uniquely most (96%) of the inbred lines that were profiled. Data from 104 RFLP loci provided a discrimination power that was only marginally greater among this same set of lines. Inbred Lines 88 and 98, that were indistinguishable according to data from 15 SSR loci, were also indistinguishable for the majority of RFLP loci. The ability to provide distance measures between inbred lines that reflect pedigrees provides a more stringent evaluation of the adequacy of marker profile data. The use of additional SSR loci will collectively allow the sorghum genome to be surveyed more comprehensively and will also give smaller standard errors around estimates of genetic distance between individuals, inbred lines, hybrids, or populations. Usage of such an expanded set of SSR loci would also provide associations among germplasm entries that are profiled that are more reflective of pedigree. The availability of more SSR loci would facilitate mapping and selection upon quantitative trait loci (QTL) in sorghum.
This relatively small set of SSRs is convenient to use because of multiplexed amplification and gel running. These SSRs can provide data that are useful to support genetic conservation and IPP to underpin sorghum research and product development because of good discrimination power and an ability to validate pedigrees. Pedigree validation is possible because of an ability to reveal adequate polymorphisms that obey the laws of simple Mendelian inheritance in contrast to the more complex and less completely understood inheritance of randomly primed amplification products. SSR data can be obtained with a high degree of reproducibility between laboratories (Jones et al., 1997) and this class of data are now being used to "help guarantee the intellectual property rights of soybean breeders", (Diwan and Cregan, 1997). Nonetheless, data from additional SSR loci that collectively sample the sorghum genome with greater thoroughness will be required before data from SSRs can be the sole source of molecular evidence to help ascertain whether a line is essentially derived according to revised laws for plant variety protection (PVP).
Yang et al. (1996) compared the costs of profiling by RFLPs, RAPDs, and ISSRs. They estimated that ISSR was the most cost effective method for obtaining varietal profile data at $1 per data point; costs of acquiring data from the use of RFLPs and RAPDs were substantially higher at $2.90 and $5.70 per data point, respectively. Sample throughput costs and timeliness have been the major factors limiting the application of molecular marker technologies in marker-assisted selection and in the monitoring of seed purity. Our estimates are that profile data from the use of SSR technology can be obtained and databased at a cost of $0.50 per data point when both DNA amplifications and gel separations are multiplexed. Advances in gel separation technologies (e.g., automated capillary gel electrophoresis) could further reduce this cost. The high initial costs of discovering SSRs and developing multiplexed sets for profiling can be shared through public investments or by private consortia.
We anticipate that SSR profiling will replace RFLPs and other PCR-based methods for the identification of sorghum inbred lines and hybrids for many applications in research, product development, the support of IPP and that SSRs can help in the more effective management and utilization of conserved genetic resources. However, still further improvements in throughput and cost effectiveness will be required before SSR technology can be widely used in marker aided selection, the monitoring of seed purity, or for studies of diversity in large numbers of landrace populations, farmer varieties, or germplasm collections. Indeed, the basic limitations in throughput that are inevitably associated with all gel-based technologies will limit the use of any gel technology for applications such as purity monitoring that demand ultra-high throughput at ultra-low cost.
Received for publication March 6, 1998.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
R. Perumal, R. Krishnaramanujam, M. A. Menz, S. Katile, J. Dahlberg, C. W. Magill, and W. L. Rooney Genetic Diversity among Sorghum Races and Working Groups Based on AFLPs and SSRs Crop Sci., July 30, 2007; 47(4): 1375 - 1383. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. T. Tesso, L. E. Claflin, and M. R. Tuinstra Analysis of Stalk Rot Resistance and Genetic Diversity among Drought Tolerant Sorghum Genotypes Crop Sci., February 23, 2005; 45(2): 645 - 652. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. A. Menz, R. R. Klein, N. C. Unruh, W. L. Rooney, P. E. Klein, and J. E. Mullet Genetic Diversity of Public Inbreds of Sorghum Determined by Mapped AFLP and SSR Markers Crop Sci., July 1, 2004; 44(4): 1236 - 1244. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||