Crop Science Illumina
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 30 July 2007
Published in Crop Sci 47:1375-1383 (2007)
© 2007 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Perumal, R.
Right arrow Articles by Rooney, W. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Perumal, R.
Right arrow Articles by Rooney, W. L.
Agricola
Right arrow Articles by Perumal, R.
Right arrow Articles by Rooney, W. L.
Related Collections
Right arrow Sorghum
Right arrow Plant Genetic Resources
Right arrow Crop Genetics

CROP BREEDING & GENETICS

Genetic Diversity among Sorghum Races and Working Groups Based on AFLPs and SSRs

Ramasamy Perumala,*, Renganayaki Krishnaramanujamb, Monica A. Menzb, Seriba Katiléa, Jeff Dahlbergc, Clint W. Magilla and William L. Rooneyb

a Dep. of Plant Pathology and Microbiology, Texas A&M Univ., College Station, TX 77843-2132
b Dep. of Soil and Crop Science, Texas A&M Univ., College Station, TX 77843-2474
c National Sorghum Producers, 4201 N. Interstate 27, Lubbock, TX 79403

* Corresponding author (rperumal{at}ag.tamu.edu).


   ABSTRACT
TOP
 NOTES
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
Forty-six converted exotic sorghum lines representing all five races and nine intermediate races of sorghum were fingerprinted using amplified fragment length polymorphism (AFLP) and simple sequence repeat (SSR) markers. A total of 453 scored AFLP and SSR loci were used to calculate genetic similarities between the lines. The dendrogram constructed using UPGMA grouped 31 lines into three major clusters with Jaccard coefficients greater than 0.75. The remaining 15 lines were grouped into four small sub-clusters each with two lines and seven single accession nodes. Phenetic analysis using AFLP and SSR markers resulted in clusters corresponding to the kafir, guinea, caudatum and durra morphological groupings. Cluster and principal coordinate analyses indicate that the guinea, kafir and intermediate lines in this study are more closely related than phenotype would suggest. Likewise, caudatum and intermediates involving caudatum showed close genetic relationship with durra and durra intermediates. For the most part, morphological classification of race based on panicle traits was also reflected by similarity in DNA based polymorphisms. The molecular diversity of bicolor and associated intermediate races is not reflective of their common morphological classification, since this race and its intermediates are quite heterogenous.

Abbreviations: AFLP, amplified fragment length polymorphism • PCA, principal coordinate analysis • SSR, simple sequence repeats • UPGMA, unweighted pair group method with arithmetic average


   INTRODUCTION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
SORGHUM (Sorghum bicolor L. Moench) is a cereal grain that originated in Africa and is now grown throughout the semiarid tropical and semiarid temperate regions of the world. While it is a staple food for millions of people in India and Africa, livestock feeding accounts for most of the sorghum use in the developed world. The genus Sorghum is very diverse; all cultivated sorghums belong to Sorghum bicolor ssp. bicolor, which is divided, based on morphology into five races (bicolor, caudatum, guinea, durra, kafir), along with the ten intermediate races resulting from all possible inter-race crosses (Harlan and de Wet, 1972).

Information on the genetic diversity and characteristics of cultivars included in world collections is very important for characterization, management, utilization, and eventually for further collection of exotic germplasm. Traditional characterization of diversity in a germplasm collection is based on specific morphological traits. The genetic variation in sorghum has been evaluated in several studies using this methodology (Appa-Rao et al., 1996; Djè et al., 1998; Dahlberg, 2000). While this approach has been effective, it requires significant time and expertise to efficiently classify material and it is based on only a few traits. It is likely that important variation for other traits is not efficiently characterized.

The development of robust molecular marker technology now provides another means to examine genetic diversity. Molecular marker technology has several advantages; large numbers of markers are available, markers cover the genome and the environment does not affect their expression (Gepts, 1993). Genetic diversity in sorghum has been estimated using several types of molecular markers viz., allozymes (Aldrich et al., 1992; Djè et al., 1998), restriction fragment length polymorphism (Aldrich and Doebley., 1992; Tao et al., 1993; Cui et al., 1994; Vierling et al., 1994; Deu et al., 2006), randomly amplified polymorphic DNA (Vierling et al., 1994; Ayana et al., 2000; Uptmoor et al., 2003; Nkongolo and Nsapato, 2003; Agrama and Tuinstra, 2003), amplified fragment length polymorphism (Uptmoor et al., 2003; Menz et al., 2004), simple sequence repeats (Brown et al., 1996; Taramino et al., 1997; Djè et al., 1998; Smith et al., 2000; Uptmoor et al., 2003; Agrama and Tuinstra, 2003; Menz et al., 2004; Casa et al., 2005) and based on combined nuclear ITS and chloroplast ndhF DNA sequences (Dillon et al., 2004; Price et al., 2005). In each of these studies, authors were interested in a specific subset of sorghum germplasm. Evaluation of similarity and diversity among entries in the sorghum germplasm collection that have been previously classified into races and working groups based on phenotypic traits used by sorghum breeders is limited. Therefore, the objectives of this study were (i) to estimate the genetic diversity present among sorghum lines from diverse origins, (ii) to assess the relationship between races as defined by morphological classification with results based on molecular phenetic analysis, and (iii) to verify that a backcross conversion program to convert height and maturity characters for adaptation to U.S. agriculture did not alter race classification.


   MATERIALS AND METHODS
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 REFERENCES
 
Plant Material
Forty-six sorghum lines from the sorghum conversion program (Stephens et al., 1967) were selected to sample the variation present. Among these lines were representatives from all five races and nine intermediate races. Also included in this set were four inbred lines developed by the Texas Agricultural Experiment Station sorghum improvement program (Table 1). Fresh leaf samples were collected from 10-d-old seedlings and genomic DNA was extracted from one or more individual seedlings from all 46 pure breeding lines using the PEX method (Williams and Ronald, 1994).


View this table:
[in this window]
[in a new window]

  		Table 1. Phenotypic classifications and origin of sorghum lines used in this analysis.

 
AFLP and SSR Analysis
For AFLP analysis, DNA samples were digested separately in two sets with EcoRI plus MseI or PstI plus MseI restriction endonucleases. AFLP template preparation and PCR reaction conditions were as described by Klein et al. (2000) and Menz et al. (2002). Eight (+3/+3) AFLP primer combinations were used to amplify the tailed products of all EcoRI/MseI and PstI/MseI digests. For the seven SSR markers analyzed, forward primers were labeled with one of the IR fluorescent dyes (LI-COR Inc., Lincoln, NE). The products were examined using a dual-dye LI-COR 4200 IR2 gel detection system (Tables 2 and 3) (LI-COR Inc.). Primer information and PCR conditions for all markers in this study were reported by Menz et al. (2002) and they are also available from the Sorghum Genome website at http://sorgblast3.tamu.edu/ (verified 29 May 2007). The markers from eight AFLP- EcoRI+MseI primer combinations and seven SSRs were previously mapped in a population produced by crossing two inbred lines, BTx623 and IS3620C (Menz et al., 2002) (http://sorgblast3.tamu.edu/) and are distributed in all ten linkage groups of the sorghum genome. Virk et al. (2000) used unmapped AFLP markers and revealed a pattern of diversity that is very similar to that obtained using a range of other marker types and which reflects the known crossability groups within rice species. So in the present study, eight PstI/MseI primer combinations unmapped to the sorghum genome were also selected on a random basis. The selected SSRs were based on different simple and compound repeat motifs with varied numbers and the markers were mapped in linkage groups 3 and 8 of sorghum genome.


View this table:
[in this window]
[in a new window]

  		Table 2. The number of bands and degree of polymorphism revealed by AFLP primer combinations.

 

View this table:
[in this window]
[in a new window]

  		Table 3. Characteristics of SSR markers with the number of alleles observed over a set of 46 sorghum lines.

 
Data Analysis
Markers were scored manually for presence (1) and absence (0) of the band in both AFLP and SSR analysis. Bands of different electrophoretic mobility were assumed to be non-allelic, while bands of the same mobility were assumed to be alleles. A pair-wise similarity matrix was calculated using the Jaccard coefficient (Jaccard, 1908):

Formula
where MSij is the DNA marker similarity index between the ith and jth genotype, Nij is the number of bands present in both genotypes Nii is the number of bands present in the ith genotype, but lacking in the jth genotype, and Njj is the number of bands lacking in the ith genotype, but present in the jth genotype. By randomizing the data input order, a dendrogram was created from the similarity matrix using the unweighted pair group method with arithmetic average (UPGMA) described by Sneath and Sokal (1973). PAUP 4.0* was used to generate 2000 bootstrap replicates for testing the reliability of the dataset and to draw a consensus tree (Swofford, 2002). Furthermore, to test the goodness of fit for the clustering to represent the similarities between accessions, the cophenetic value matrix was generated for each of the dendrograms and compared with the corresponding similarity matrix by the Mantel matrix correspondence test (Mantel, 1967). Significance of Z was determined by comparing the observed Z values with a critical Z value obtained by calculating Z for one matrix with 1000 permuted variants of the second matrix. Ordination analysis was performed to study the relatedness within a matrix by converting the pairwise distance into Eigen vectors and values. Cluster analyses, ordination analyses and the Mantel test were performed using NTSYSpc (NTSYS– for Numerical Taxonomy SYStems) version 2.1 (Rohlf, 2000).


   RESULTS AND DISCUSSION
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
REFERENCES
 
AFLP and SSR Analysis
A total of 16 AFLP primer combinations [Eight (+3/+3) in each EcoRI/MseI and PstI/MseI] and seven SSRs produced clear banding patterns that could be scored for determining genetic diversity across the panel. A total of 1166 amplified AFLP-DNA fragments were generated across 46 lines, with 356 fragments being polymorphic (30.53% polymorphism) (Table 2). The number of bands from each primer combination ranged from 60 to 160 for EcoRI/MseI and 37 to 66 for PstI/MseI primer combinations. The EcoRI/MseI primer combinations amplified more bands (774) but at a lower rate of polymorphism (22.2%) compared to PstI/MseI primer combinations where 46.9% of 392 bands were polymorphic. These results seem consistent with other reports. In a study by Uptmoor et al. (2003) the average number of scored bands was 39.3 with 61.7% polymorphism for EcoRI/MseI primer combinations. For the seven SSRs used in this study, a total of 97 polymorphic alleles were detected. Among the seven SSRs, Xtxp336 revealed the lowest number of alleles (4) while Xtxp33 and Xtxp205 detected the largest number of alleles (21) (Table 3). Menz et al. (2004) detected between 2 and 19 alleles, with an average of 7.8 alleles in the whole set of sorghum inbred lines studied. Representative examples of the amplification products obtained with 46 sorghum lines using the AFLP primer combinations EcoRI-TGA/MseI-CAA, PstI-GAT/MseI-CTA and SSR (Xgap 236) are shown in Fig. 1a, 1b, and 1c .


Figure 1
Figure 1
Figure 1
View larger version (230K):
[in this window]
[in a new window]

  		Figure 1a, b, and c. AFLP amplification products from 46 Sorghum bicolor isolates using EcoRI-TGA/MseI-CAA (Fig. 1a) PstI-GAT/MseI-CTA (Fig. 1b) and SSR–Xgap 236 (Fig. 1c) primers as detected using a LI-COR system. (Lanes 1 to 46 represent Sorghum bicolor lines IS3620C, BTx623, SC195, SC614, SC1103, SC214, SC224, SC950, SC15, SC315, SC319, SC635, SC641, SC647, SC663, SC671, SC322, SC258, SC66, SC115, SC237, SC52, SC93, SC392, SC407, SC749, SC411, SC418, SC228, SC171, SC425, SC233, SC193, SC465, SC910, SC166, SC131, SC199, SC209, SC303, SC06, SC1156, SC941, RTx 436, RTx 430, and BTx 378). Ladder sizes are given in base pairs (bp).

 
Cluster Analysis
A total of 453 markers from both AFLP and SSRs were combined for cluster analysis. Although most amplified bands were common to all of the sorghum cultivars, as indicated by a Jaccard coefficient of 0.67 or greater, the dendrogram constructed using UPGMA grouped 31 lines into three major clusters A, B, and C with Jaccard coefficients greater than 0.75 and with confidence limits of 77, 75, and 82% respectively (Fig. 2 ). The remaining 15 lines were grouped into four small sub-clusters each with two lines with confidence limits ranging 86 to 100% and seven single accession nodes (Fig. 2). The high cophentic correlation (rcoph = 0.866) indicated a good agreement between the tree and the original similarity matrix, thus corroborating the consistency of the tree. The Mantel Z test statistic was also significant (Z = 2.86; p = 0.002) suggesting a good correspondence between the matrices. Cluster A is comprised of 11 lines, 10 of which are kafir and intermediates involving kafir (kafir/intermediates) (n = 6) and guinea/intermediates (n = 4). SC635 (kafir) and BTx378 in cluster A showed the highest coefficient of similarity (closest pair). Cluster B is the largest: based on phenotypic characters, 13 of the 14 lines are classified as caudatum or caudatum intermediates. However, not all lines classified as caudatum in this study were present in this group. Lines SC171 and SC425 form one of the nodes outside the major clusters, as does SC322. In cluster C, five durra/intermediates were grouped together. Bicolor/intermediate lines SC614, SC15, and SC214 had patterns that placed them in clusters A, B, and C respectively; the remaining three bicolor lines and six bicolor derived races were scattered and formed independent clusters. Finally, SC303 (guinea-margaritiferum) and its derived accession IS3620C in small sub-cluster also showed high coefficient of similarity.


Figure 2
View larger version (32K):
[in this window]
[in a new window]

  		Figure 2. Dendrogram of 46 Sorghum bicolor lines using AFLP and SSR markers as per unweighted pair group method with arithmetic average clustering. Numbers shown at different nodes represent percentage confidence limits obtained in the bootstrap analysis and are shown for clusters with > 60%. The scale shown below is the measure of genetic similarity coefficients calculated according to Jaccard (1908).

 
To visualize the genetic relationships among the sorghum lines in greater detail, principal coordinate analysis (PCA) was performed for display in two dimensions (Fig. 3 ). In this analysis, the first two principal components (having eigen value > 1) explained about 74% of the total variation. Like the UPGMA clustering dendrogram, the PCA analysis placed the 46 lines into distinct groups. Group I had two sub-groups of both guinea and kafir and their intermediates. Group II had two subgroups of caudatum and durra. As in the dendrogram, bicolor/intermediate races were scattered in the two-dimensional picture.


Figure 3
View larger version (14K):
[in this window]
[in a new window]

  		Figure 3. Two-dimensional picture of principal coordinate analysis estimated using AFLP and SSRs genetic similarity matrix of 46 Sorghum bicolor lines.

 
In cluster A of the dendrogram, SC647 (kafir-caudatum) and SC663 (guinea-kafir) were grouped with four kafir lines implying that these lines are genetically more similar to kafir rather than caudatum and guinea races. The six kafir/intermediates in cluster A were a mixed group of restorer and maintainer lines in the A1 cytoplasmic male sterility system. Menz et al. (2004) reported similar results, where the elite sorghum germplasm lines were grouped by genetic background and not by existing B- or R-line classification. Likewise, in the next subgroup in cluster A, SC465 (guinea-durra) and SC315 (guinea-bicolor) were grouped with two other guinea lines. In this analysis, the genetic relatedness of these intermediates is closer to guinea rather than durra and bicolor. All four guinea/intermediates in this group are restorer (R) lines, while another of the pair of guinea lines separated from cluster A are maintainer (B) lines. All of these results imply that genetic diversity and the basis of race are not influenced by the capability to restore fertility to A1 cytoplasm. Likewise, geographical origins had little influence on this cluster as entries in this group are from very different regions. Our phenetic results agreed with those observed by Deu et al. (2006). They reported that Asian collections of cultivated sorghum were clustered in different equatorial groups.

The caudatum race and its derivatives have been very important in the development of improved sorghums (Harlan and de Wet, 1972). In cluster B, five caudatum lines and eight intermediate races (SC418, SC641, SC115, SC66, SC407, SC411, RTx430, and SC319) were distinctly grouped together. Two caudatum lines (SC171 and SC425) formed a separate sub-cluster, but in the two-dimensional view of PCA analysis, these two lines were grouped along with other caudatum lines. The caudatum accession SC322, a partial maintainer, was also separated from cluster B.

Three durra lines (SC209, SC233, SC199) and two intermediates (SC193 and SC910) from India were grouped together in cluster C, indicating that for these two intermediates alleles typical of durra were more common than those of bicolor or guinea. However, three other durra-bicolor intermediates (SC166, SC06, and SC131) from Ethiopia and SC1156 from India were not included in this group. Finally, SC749 and SC950, with origins reported as Japan and the USA, respectively were unique. The last six lines mentioned are all intermediate races involving bicolor, the genetically most heterogeneous race. Mixtures of varied bicolor alleles with those of another race could also explain the scattering of these intermediates into separate clusters. In a study by Deu et al. (2006), bicolor accessions were scattered widely across different clusters due to its wide geographic distribution and diversity of uses (forage, broom corn, and sweet stems) as reported by Doggett (1988).

Genetic Diversity
The exotic germplasm in this study was taken from the sorghum conversion (SC) program (Stephens et al., 1967) and all of the lines were fully converted to daylength insensitivity and short stature. Fully converted implies a minimum of four backcrosses to the exotic accession. The donor parent of the alleles for photoperiod insensitivity and shorter stature were provided by Tx406 which is classified as a kafir (D.T. Rosenow, personal communication, 2006). Therefore, the SC lines are genetically identical in the genomic regions controlling photoperiod sensitivity and dwarfing (Lin et al., 1995), but these BC4 individuals still retain about 3% of alleles from the elite parent in addition to the "converted regions". Any bands amplified from selected regions would not be expected to show polymorphisms, and thus would not contribute to the diversity seen. For all other characters the lines are expected to reflect the diversity present in their photoperiod sensitive isoline. The observations here support that expectation. Other than in two lines that are morphologically classed as kafir-intermediates clustering with kafir rather than the other parent or a separate node, there was no indication of excess kafir alleles.

Classification of the entries in this study and in the sorghum germplasm collection is challenging due to the relatively high level of introgression that has occurred during the evolution of this diversity (Doggett, 1988). Initially, Snowden (1936) used simple traits such as grain color, glume color, awns, and persistence of pedicellate spikelets. However, all of these characters vary widely within related forms to the point that they have little taxonomic value. Traits such as height, tillering, juiciness of stalk, and daylength response are useful for agronomic purposes, but also vary greatly among related forms and are not useful for classification purposes. Instead, Harlan and de Wet (1972) based their classification on spikelet morphology and grain characteristics and identified five main races of sorghum from the mature sessile spikelets alone. Their reasoning was that spikelet characters are considered to be the most stable, the least influenced by environment, and the most revealing with respect to relationships.

The five different races and their intermediates used in this study represent the natural distribution of morphological and geographical variability present in the collection. Marker analysis detected significant allelic variation in the 46 lines. Phenetic analysis using AFLP and SSR markers resulted in clusters loosely corresponding to the kafir, guinea, caudatum, and durra morphological groupings. Many of the intermediate races fell into the same cluster as one of the donor parents. For the most part then, morphological classification of race based on panicle traits was also reflected by similarity in DNA based polymorphisms. Cluster and PCA analyses together clearly revealed a close genetic relationship of guinea and kafir and their intermediates. Likewise, with the PCA comparisons, caudatum and many caudatum intermediates shared many common alleles with durra/intermediates. Given that the centers of diversity for these two races often overlap, a greater degree of relation is not unexpected (Kimber, 2000).

The molecular diversity of bicolor/intermediate races is not reflected in their common morphological classification. de Wet (1978) found the race bicolor to be the most heterogeneous and suggested it was most closely related to wild sorghum. The results of this study clearly support this hypothesis since SC lines derived from bicolor and its intermediate races were widely distributed and present nearly throughout the dendrogram.

Other than the race bicolor, the general correspondence between morphological and molecular diversity measures is interesting as each similarity measure is based on two different characteristics of the plant and because diversity at molecular markers, which are a priori neutral, may not reflect diversity of quantitative traits (Karhu et al., 1996). It may be the result of an adequate representation of genetic relationships by the observed morphological traits and large variation of these traits among the set of lines used in our study.

The most accurate characterization of genetic diversity is completed through the simultaneous use of several relatedness measures (Cox et al., 1985; Schut et al., 1998). The construction of a more comprehensive, composite index of relationships based on pedigree, morphological, biochemical, and molecular markers data is expected to improve accuracy. Both AFLP and SSRs markers have unprecedented utility for analysis of population genetics and phenetic diversity of sorghum. The use of SSRs potentially could remove most, if not all, of the limitations in revealing polymorphisms and in obtaining more complete genomic coverage for plants, as has been achieved already for the human genome (Smith and Helentjaris, 1996). The utility of PCR-based markers such as AFLP and SSRs for measuring diversity, for assigning lines to heterotic groups, and for genetic fingerprinting should prove valuable for sorghum breeding programs.


   NOTES
TOP
 NOTES
ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 
All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

Received for publication August 18, 2006.


   REFERENCES
TOP
 NOTES
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Perumal, R.
Right arrow Articles by Rooney, W. L.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Perumal, R.
Right arrow Articles by Rooney, W. L.
Agricola
Right arrow Articles by Perumal, R.
Right arrow Articles by Rooney, W. L.
Related Collections
Right arrow Sorghum
Right arrow Plant Genetic Resources
Right arrow Crop Genetics

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