Springer 2006
Molecular Breeding (2006) 17: 69–77 DOI 10.1007/s11032-005-2453-6
-1
Genetic diversity and core collection evaluations in common wheat germplasm from the Northwestern Spring Wheat Region in China C.Y. Hao1, X.Y. Zhang1,*, L.F. Wang1, Y.S. Dong1, X.W. Shang2 and J.Z. Jia1 1
Key Laboratory of Crop Germplasm & Biotechnology, MOA, Institute of Crop Germplasm Resources, Chinese Academy of Agricultural Sciences, Beijing 100081, China; 2Gansu Agricultural University, Lanzhou 730070, China; *Author for correspondence (e-mail:
[email protected]; phone: +86-10-62135294; fax: +86-10-62135294) Received 13 December 2004; accepted in revised form 21 August 2005
Key words: Core collection, Genetic diversity, Microsatellite, Wheat
Abstract Fluorescence microsatellite markers were employed to reveal genetic diversity of 340 wheat accessions consisting of 229 landraces and 111 modern varieties from the Northwest Spring Wheat Region in China. The 340 accessions were chosen as candidate core collections for wheat germplasm in this region. A core collection representing the genetic diversity of these accessions was identified based on a cluster dendrogram of 78 SSR loci. A total of 967 alleles were detected with a mean of 13.6 alleles (5–32) per locus. Mean PIC was 0.64, ranged from 0.05 to 0.91. All loci were distributed relatively evenly in the A, B and D wheat genomes. Mean genetic richness of A, B and D genomes for both landraces and modern varieties was B>A>D. However, mean genetic diversity indices of landraces changed to B>D>A. As a whole, genetic diversity of the landraces was considerably higher than that of the modern varieties. The big difference of genetic diversity indices in the three genomes suggested that breeding has exerted greater selection pressure in the D than the A or B genomes in this region. Changes of allelic proportions represented in the proposed core collection at different sampling scales suggested that the sampling percentage of the core collection in the Northwest Spring Wheat Region should be greater than 4% of the base collection to ensure that more than 70% of the variation is represented by the core collection.
Introduction Abundant wheat germplasm resources provide considerable opportunities for genetic research and breeding. However, huge numbers of accessions represent challenges for conservation, evaluation, identification and utilization (Tanksley and McCouch 1997). For the convenience of management, research and application, scientists proposed the concept of the core collection (Brown 1989; Brown 1995; Hintum 1995). Zhang
et al. (2002a) and Dong et al. (2003) constructed the candidate core collection for Chinese wheat germplasm based on cluster analysis of passport data. The 5029 candidate core collections were analyzed for their high molecular weight glutenin subunit components (Zhang et al. 2002a). In 2003, a core collection was established based on the SSR data of the candidate core collections (unpublished data). The Northwest Spring Wheat Region of China is located at the crossroads of three high plateaus
70 (Loess, Qinhai-Tibeat, and Inner-Mongolia) on the upper reaches of the Yellow River. This wheat production region includes the bulk of Gansu and Ningxia Provinces and the eastern portion of Qinghai Province, which represents about 4% of the wheat production in China (Jin 1996; Zhuang 2003; You et al. 2004). This region, however, has the greatest phenotypic diversity in the five spring wheat production regions of China (Dong et al. 2003). In this paper, we used SSR analysis to characterize genetic variation in the candidate core collections from the Northwest Spring Wheat Region. We were interested in using these SSR results to identify a core collection to represent the genetic variation for wheat in this region. Results of this study will provide valuable information to geneticists and breeders for future wheat improvement.
Materials and methods Plant material A total of 340 wheat accessions including 229 landraces and 111 modern varieties were evaluated in this study (see supplementary 1) (http:// icgr.caas.net.cn). These accessions were candidates for a core collection of 1189 base collections from the Northwest Spring Wheat Region of China and represented 96.4% of the phenotypic variation of the base collection (Dong et al. 2003).
Data collection and analysis Genomic DNA was extracted from lyophilized mixed young leaves of 10 seedlings as described by Sharp et al. (1989). The SSR primers were synthesized at Applied Biosystems Company according to Ro¨der et al. (1998). Seventy-four pairs of SSR primers were used to detect variation in the 340 candidate collections (Zhang et al. 2002b; You et al. 2004). These primers were distributed evenly on the 21 wheat chromosomes (Table 1, see supplementary 2). The basic PCR reaction procedure was at 94 C for 5 min, 94 C for 1 min, 55 C for 1 min and 72 C for 1 min
and then return to step 2 for 35 cycles. Different primers may have different annealing temperatures (Ro¨der et al. 1998). The amplified products were incubated at 72 C for 5 min and stored at 4 C. PCR amplification products were purified by 100% and 70% ethanol, re-suspended in ddH2O and analyzed with a Model 3700 capillary DNA analyzer. The fragment size (base pair) was determined by comparison with an inner size standard (LIZ500). The fluorescence data were processed by Genescan and Genotyper software from ABI Company. Details of the experimental procedures and data processing can be found in the ABI 3700 instruction manual and Hao et al. (2005). Microsatellite loci were scored individually, and alleles were recorded as 1 for presence and 0 for absence (see supplementary 3). Genetic diversity of each locus (i) was expressed by allelic richness (R Aij), and polymorphism information l P content (PIC) PICi ¼ 1 p2ij , where j reprej¼1
sents each allele at the i locus and pij is the frequency of each allele at locus i. Genetic dispersion index (Ht) for each genome was comk P puted by Ht ¼ PICi/k, where k is the total i¼1
number of loci studied within each genome, except those amplified in two or three genomes at the same time (Nei 1973). Genetic similarities between accessions were calculated using the DICE coefficient (NTSYS-pc version 2.1) (Rohlf 2000). Cluster analysis was performed based on the similarity matrices with the un-weighted pair-group method using arithmetic average (UPGMA), and a dendrogram was constructed based on the similarity matrices. Principal coordinate analysis (PCO, NTSYS-pc version 2.1) was used to reveal the relationships among the 340 accessions. Core entries were chosen based on the cluster tree. For each major branch, accessions with specific SSR alleles or very well known varieties were chosen as priority core entries. To reveal the influence of sampling proportion on allelic representation, we conducted two simulations choosing the entries at various proportions (13, 10, 7, 5, 4, 3, 2 and 1%). The mean value in the two times was used to evaluate the relationship between sampling proportion and percentage of alleles kept as well as the representative variation
71 Table 1. Genetic diversity of modern varieties and landraces from the Northwest Spring Wheat Region. Parameter
Loci Alleles (R Aij) Mean genetic richness (R Aij/Loci) Mean genetic diversity (Ht)
Population
Modern varieties Landraces Modern varieties Landraces Modern varieties Landraces
percentage, which was calculated by 96.4% · percentage of alleles kept.
Results and discussion Polymorphism of 71 loci within the 340 candidate core collections In the 74 pairs of primers, Xgwm47, Xgwm296, and Xgwm635 were multi-loci primers. They were not used in the evaluation of diversity of the three genomes (A, B and D). The number of alleles and polymorphism information content (PIC) for each locus was presented in Table 1 (see supplementary 2). A total of 967 alleles were detected. The average allelic richness was 13.6 alleles per locus, which ranged from 5 to 32. The PIC ranged from 0.05 to 0.91 for the 71 loci with a mean PIC of 0.64. After comparing allelic richness and the respective PIC, we found that allelic richness was positively associated with PIC. However, the relationship between allelic richness and PIC was not linear. At several loci, such as Xgwm358 (5D), Xgwm642 (1D) and Xgwm193 (6B), there was high allelic richness; however, their PIC values were very low (Table 1, see supplementary 2). The coefficient between allelic richness and PIC was 0.505 (p < 0.01), which was lower than that reported by Huang et al. (2002) (r = 0.73, p < 0.01). According to the data for the 71 loci, there was a quadratic relationship between allelic richness and PIC (Y = 0.257 + 0.039X0.001X2, R2 = 0.284) (Figure 1), which is not consistent with data for 24 loci by Huang et al. (2002) (Y = 23.933 + 54.777X, R2 = 0.535). This lack of agreement might be caused by loci with
Genome
Total
A
B
D
21 207 249 9.86 ± 1.09 11.86 ± 1.10 0.61 ± 0.04 0.58 ± 0.04
26 284 359 10.92 ± 0.91 13.81 ± 1.11 0.61 ± 0.03 0.64 ± 0.04
24 229 274 9.54 ± 0.80 11.42 ± 1.20 0.55 ± 0.06 0.61 ± 0.05
71 720 882 10.14 ± 0.54 12.42 ± 0.66 0.59 ± 0.03 0.61 ± 0.02
very low PIC values in our study (Table 1, see supplementary 2). For the 24 loci used in Huang et al. (2002), the locus with lowest PIC was Xgwm192c (4AS). It had a PIC of 0.41, which was considerably higher than the two lowest loci in our study, Xgwm358 (5D, 0.05) and Xgwm642 (1D, 0.1). The allelic frequency distributions for Xgwm193 (very low PIC) and Xgwm46 (highest PIC) were compared for landraces and modern varieties (Figures 2 and 3). The results showed that the alleles detected at Xgwm193 were distributed unevenly with 168 bp present in 90% of the accessions (Figure 2). These results suggest that this allele may be linked to important agronomic traits or environmental adaptation. This may help explain the strong hitchhiking effects commonly observed in domestication and breeding of wheat (Lander and Schork 1994; Andolfatto 2001). However, at Xgwm46 the alleles were distributed relatively evenly in landraces, although hitchhiking effects were observed at allele 146 bp in the modern varieties (Figure 3). Thus, both allelic richness and their respective PIC values should be considered to objectively evaluate polymorphism of different loci and genetic diversity of the collections.
Genetic diversity in the three genomes Mean genetic richness and diversity indices were calculated for the A, B and D genomes in the 229 landraces and 111 modern varieties (Table 1). The 71 loci were distributed relatively evenly in the A, B and D genomes with values of 21, 26 and 24, respectively. The total number of alleles amplified for landraces was 882 and for modern varieties was 720. Mean genetic richness and genetic
72
Figure 1. Relationship between allelic richness and PIC based on 71 loci.
Figure 2. Allelic frequency detected at Xgwm193 (6B).
diversity indices for landraces and modern varieties were shown in Figures 4 and 5. Mean genetic richness for the landraces (12.42 ± 0.66) was higher than that for modern varieties (10.14 ± 0.54). The genetic diversity index for landraces (0.61 ± 0.02) was higher than that for modern varieties (0.59 ± 0.03). For modern varieties, the B genome exhibited the greatest diversity and the D genome had the lowest diversity (Figures 4 and 5), which is consistent with the results of Jia et al. (2001). For the landraces, allelic rich-
ness of the three genomes was B > A > D, but the Ht index was B > D > A (Figures 4 and 5). This suggests that the breeding process may have exerted the strongest selection pressure in the D genome, resulting in significant decreases in Ht in this region (Lander and Schork 1994; Andolfatto 2001). Based on RFLP analysis of 15 wheat varieties worldwide, Jia et al. (2001) reported that genome distributions for genetic diversity in modern varieties were B > A > D. Our results for the modern
73
Figure 3. Allelic frequency detected at Xgwm46 (7B).
Figure 4. Mean genetic richness in landraces and modern varieties.
varieties agreed with this ranking; however, our results for the landraces did not. Furthermore, we found that major genetic differences existed between landraces and modern varieties for the D genome (Figure 5), which agreed with the results of Zhang et al. (2002b) and You et al. (2004). This indicates that modern hybridization and breeding has had a very strong selection in the D genome, which implies that the D genome contains traits that are very attractive to breeders (Lander and Schork 1994; Andolfatto 2001). Analysis of all Chinese wheat varieties showed that different wheat growing regions had different hitchhiking
effects for the same genome (unpublished data). Nevertheless rich diversity in the D genome for landraces has potential for widening genetic diversity of the D genome in modern varieties.
Clustering analysis of the 340 accessions The 340 accessions were clustered based on DICE derived from SSR data using UPGMA (Rohlf 2000). Results from this analysis showed two major cluster groups (Group I and II) (Figure 6, see supplementary 4). Group I mainly consisted of
74
Figure 5. Mean genetic diversity of landraces and modern varieties.
landraces, whereas Group II mainly consisted of modern varieties, suggesting that landraces and modern varieties are relatively independent genetic populations. This agrees with previous results (Zhang et al. 2002b; You et al. 2004). However, there were some exceptions including nine modern varieties (Xiangnong No.3, Beicheng 10, Huining
No.5, Huining 10, Jinmai 303, Ganchun 11, Ganchun 12, Ganchun 16 and Jiunong 10), which clustered into Group I. Similarly, 32 landraces including Gansu Honglanmai, Gansu Baimangzi, Qinghai Heshangmai and Ningxia Tongxinshanmai were clustered into Group II. Principal coordinate analysis of the 340 accessions made the
Figure 6. PCO analysis of the 340 accessions based on SSR data of the 78 loci.
75 boundary between the landraces and modern varieties more distinct (Figure 6). This agrees with the systematic pedigree analysis of modern varieties, which indicated that modern wheat breeding has mainly relied on the introduction of germplasm from other countries (Zhuang 2003).
Allelic (variation) representation percentage in the core collection at different sampling proportion A core collection consists of accessions selected from an existing germplasm collection, chosen to represent the genetic spectrum in the entire collection. The core should represent as much genetic diversity as possible for multiple uses and many users (Frankel and Brown 1984; Brown 1989, 1995). To identify a core collection of Chinese wheat germplasm, we conducted a pilot experiment with wheat geremplasm collected in the Northwestern Spring Wheat Region. Selection of the core accessions was based on the cluster results of SSR data in the 340 candidate core collections that were chosen based on cluster analyses of accession passport data (Zhang et al. 2002a; Dong et al. 2003). This allowed an optimal representation of diversity within the groups (Frankel 1984; Frankel and Brown 1984; Brown 1989). The 74 pairs of SSR primers were used to amplify 1035 alleles in the 340 candidate core collections. Then, allelic representation percentage and representation of variation at different sampling proportion were evaluated simultaneously for the 1189 accessions from the base collection and 1035 alleles (Table 2). At a 13% sampling proportion (155 accessions), the percentage of alleles kept reached 98.5% of the core collection. At a 10% sampling proportion, 93.4% of alleles were kept in the core. When the sampling proportion was reduced from 13% to 3%, the percentage of alleles consistently decreased. However, when sampling proportion reached less than 3%, the allelic or variation representation dropped dramatically (Table 2, Figure 7). With a 3% sampling proportion (36 accessions), 71.1% of the SSR alleles were represented, but variation representation in the core was less than 70%. Therefore, to ensure more than 70% of the variation in the base collection is represented, the core collection should be greater than 4% of the base collection (Table 2).
Table 2. Allelic representation of genetic variation at different sampling proportion. Number Sampling Number Percent of Representation of accessions percent of alleles alleles kept 1189 155 119 83 59 48 36 12
100 13 10 7 5 4 3 1
1035 1019 967 909 844 805 735 502
100 98.5 93.4 87.8 81.6 77.8 71.1 48.5
96.4 94.9 90.1 84.7 78.6 75.0 68.5 46.8
Figure 7. Representation percentage of genetic diversity at different sampling proportion.
To establish a core collection, the sampling proportion and variation representation of the base collection are the most important characteristics. Brown et al. (1987) recommended that the number of collections in the core should account for 5–10% of the base collection, the total number of accessions should not exceed 3000 and the core should represent at least 70% of the genetic diversity in the base collection. Diwan et al. (1994) indicated that core collection sampling should always be greater than 10%. Hintum (1999) suggested that the sampling proportion should depend on the particular objective of the core collection and should vary between 5% and 20% of the base collection. Hintum (1994) used a sampling proportion of 6.5% to establish a core collection of barley in Holland. For Bari faba bean, a sampling proportion of 26.6% was used (Scippa 2002). In establishing a core collection of rice germplasm in Yunnan Province, Li et al. (2000) found that 5%
76 of the base collection represented 96% of the phenotypic variation. For the candidate core collection of Chinese wheat germplasm, 21.5% of the basic accessions represented 96.4% of the genetic variation for 21 botanic and agronomic characters (Zhang et al. 2002a; Dong et al. 2003). In the Northwestern Spring Wheat Region of China, only 48 accessions (4% sampling proportion) could be used to represent 75% of the genetic variation in the base collection. This was considerably greater than the accepted standard (10% of the base accessions representing more than 70% of the genetic diversity). Our results show the great potential for combining molecular data with passport data to construct a core collection. Supplementary material to this paper is available in electronic format at http://dx.doi.org/10.1007/ s11032-005-2453-6. Acknowledgements We gratefully appreciate the assistance of Prof. Douglas Johnson of USDA-ARS in Logan, Utah for English editing. This research was supported by the Chinese Ministry of Science and Technology (G19980202, 2004CB117202) and CAAS funding to XYZ. References Andolfatto P. 2001. Adaptive hitchhiking effects on genome variability. Curr. Opin. Genet. Dev. 11: 635–641. Brown A.H.D., Grace J.P. and Speer S.S. 1987. Designation of a ‘‘core’’ collection of perennial Glycine. Soybean. Genet. Newsl. 14: 59–70. Brown A.H.D. 1989. Core collections: a practical approach to genetic resources management. Genome 31: 818–824. Brown A.H.D. 1995. The core collection at the crossroad. In: Hodgkin T., Brown A.H.D., Hintum Van T.H.L. and Morales E.A.V. (eds), Core Collection of Plant Genetic Resources. International Plant Genetic Resources Institute (IPGRI), Wiley-Science, pp. 3–19. Diwan N., Bauchan G.R. and Mclntosh M.S. 1994. A core collection for the US annual Medicago germplasm collection. Crop. Sci. 34: 279–285. Dong Y.S., Cao Y.S., Zhang X.Y., Liu S.C., Wang L.F., You G.X., Pang B.S., Li L.H. and Jia J.Z. 2003. Establishment of candidate core collections in Chinese common wheat germplasm. J. Plant Genet. Resour. 4(1): 1–8. Frankel O.H. 1984. Genetic perspectives of germplasm conservation. In: Arber W., Limensee K. and Peacock W.J. (eds), Genetic Manipulation: Impact on Man and Society. Cambridge University Press, Cambridge, UK, pp. 161–170.
Frankel O.H. and Brown A.H.D. 1984. Plant genetic resources today: A critical appraisal. In: Holden J.H.W. and Williams J.T. (eds), Crop Genetic Resources: Conservation and Evaluation. George Allen & Urwin Ltd, London, pp. 249– 257. Hao C.Y., Wang L.F., Jia J.Z., Dong Y.S. and Zhang X.Y. 2005. Comparison of fluorescence and silver-staining detection systems of microsatellite markers. Acta Agronomica Sinica 31: 144–149. Hintum Th. JL. Van. 1994. Comparison of marker systems and construction of a core collection in a pedigree of European spring barley. Theor. Appl. Genet. 89: 991–997. Hintum Th. JL. Van. 1995. Hierarchical approaches to the analysis of genetic diversity in crop plants. In: Hodgkin T., Brown A.H.D., Hintum Th.J.L. and Morales E.A.V. (eds), Core Collection of Plant Genetic Resources. John Wiley & Sons, pp.23–34. Hintum Th. JL. Van. 1999. The general methodology for creating a core collection. In: Johnson R.C. and Hodgkin T. (eds), Core Collections for Today and Tomorrow. International Plant Genetic Resources Institute, Italy, pp. 10–17. Huang X.Q., Bo¨rner A., Ro¨der M.S. and Ganal M.W. 2002. Assessing genetic diversity of wheat (Triticum aestivum L.) germplasm using microsatellite markers. Theor. Appl. Genet. 105: 699–707. Jia J.Z., Zhang Z.B., Devos K. and Gale M.D. 2001. Genetic diversity of RFLP mapping loci for 21 wheat chromosomes. Science in China Ser.C 31: 13–21. Jin S.B. 1996. Wheat Science in China. Agricultural Press, Beijing (In Chinese). Lander E.S. and Schork N.J. 1994. Genetic dissection of complex traits. Science 265: 2037–2048. Li Z.C., Zhang H.L., Zeng Y.W., Yang Z.Y., Shen S.Q., Sun C.Q. and Wang X.K. 2000. Study on sampling schemes of core collection of local varieties of rice in Yunnan, China. Scientia Agricultura Sinica 33(5): 1–7. Nei M. 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. USA 70: 3321–3323. Ro¨der M.S., Korzun V., Wendehake K., Plaschke J., Tixier M.H., Leroy P. and Ganal M.W. 1998. A microsatellite map of wheat. Genetics 149: 2007–2023. Rohlf F.J. 2000. NTSYS-pc: numerical taxonomy and multivariate analysis system, version 2.1. Exeter Software, Setauket, N.Y. Scippa G., Polignano G.B. and Uggenti P. 2002. Diversity analysis and core collection formation in Bari faba bean germplasm. Plant Genet. Res. Newsl. 125: 33–38. Sharp P.J., Chao S., Desai S. and Gale M.D. 1989. The isolation, characterization and application in Triticeae of a set of wheat RFLP probes identifying each homoecologous chromosome arm. Theor. Appl. Genet. 78: 342–348. Tanksley S.D. and McCouch S.R. 1997. Seed bank and molecular maps: unlocking genetic potential from the wild. Science 277: 1063–1066. You G.X., Zhang X.Y. and Wang L.F. 2004. An estimation of the minimum number of SSR loci needed to reveal genetic relationships in wheat varieties: information from 96 random samples with maximized genetic diversity. Mol. Breeding 14: 397–406. Zhang X.Y., Pang B.S., You G.X., Wang L.F., Jia J.Z. and Dong Y.S. 2002a. Allelic variation and genetic diversity at
77 Glu-1 loci in Chinese wheat (Triticum aestivum L.) germplasms. Scientia Agricultura Sinica 35(11): 1302–1310. Zhang X.Y., Li C.W., Wang L.F., Wang H.M., You G.X. and Dong Y.S. 2002b. An estimation of the minimum number of SSR alleles needed to reveal genetic relationships in wheat varieties. I. Information from large-scale planted
varieties and corner-stone breeding parents in Chinese wheat improvement and production. Theor. Appl. Genet. 106: 112–117. Zhuang Q.S. 2003. Chinese Wheat Improvement and Pedigree Analysis (in Chinese). China Agricultural Publishing House, Beijing.
