Experimental
Chinese Spring (CS) wheat seeds were acquired from various laboratories around the world and assigned an accession number (Table 1). The accession held in our laboratory is CS22. Seedlings from each source were grown in the greenhouse and harvested for DNA extraction using DNeasy extraction kits (Qiagen, Inc., Valencia, CA, USA) as described by the manufacturer. Quantity and quality of DNA were assessed by spectrophotometry and agarose gel electrophoresis. Polymerase chain reaction (PCR) primers (Supplementary Table S1, available online only at http://journals.cambridge.org) were designed to amplify selected targets as identified in salt tolerance subtraction [4909D4 (AY924304) and 4910D4 (AY924305); Mott, unpublished data] and microarray studies (Ta13485, Ta25111, Ta3109, Ta25815 and Ta10194; Mott and Wang, Reference Mott and Wang2007).
Table 1 Results of polymerase chain reaction (PCR) amplification in Chinese Spring wheat (Triticum aestivum L.) accessions with primers designed from gene sequences identified in salt-responsive expression studies
a R. Wang, USDA-ARS, Logan, UT, USA.
b A. Lukaszewski, UC-Riverside, Riverside, CA, USA.
c H. Ohm, Purdue University, West Lafayette, IN, USA.
d P. Gustafson, University of Missouri, Columbia, MO, USA.
e W. Cao, Agriculture and Agri-Food Canada, Ottawa, ON, Canada.
f J. Koenig, INRA, France.
g M. Sorrells, Cornell University, Ithaca, NY, USA.
h A. Martin, CSIG, Cordoba, Spain.
i T. Koniarov, Dobroudja Agricultural Institute, Bulgaria.
j A. Graner, IPK Genebank, Germany.
k T. Payne, CIMMYT, Mexico.
l M. Mackay, AWCC, Tamworth, Australia.
m J. Raupp, Kansas State University, Manhattan, KS, USA.
PCR [25 μl, containing 40 ng genomic DNA, 1 × reaction buffer containing 15 mM MgCl2, 0.2 μM gene-specific primers, 0.25 mM dNTP, 1.25 U Taq DNA polymerase (Promega, Madison, WI, USA)] were incubated as follows: 95°C for 2 min, then 36 cycles of 95°C for 15 s, 55°C for 15 s, 72°C for 30 s in a GeneAmp PCR system 9700 (Applied Biosystems, Foster City, CA, USA). PCR products were visualized by electrophoresis in 2% agarose gels followed by staining with ethidium bromide and illumination on a UV lightbox. The PCR results, shown in Table 1, varied from bands that amplified in all CS accessions (actin and Ta25111), bands that amplified in all accessions except one (Ta13485 and Ta10194), bands that did not amplify in several accessions (4910D4, Ta3109, Ta5503 and 4909D4), to bands that failed to amplify in most CS accessions (Ta25815).
For amplified fragment length polymorphism (AFLP) analysis, two plants from each source were analysed. The AFLP procedure followed the protocol of Vos et al. (Reference Vos, Hogers, Bleeker, Reijans, van de Lee, Hornes, Frijters, Pot, Peleman and Kuiper1995), using the selective primers E.ACA/M.CAT, E.ACG/M.CAA, E.AGG/M.CTG, E.ACG/M.CTC, E.AGG/M.CTA and E.ACT/M.CTC. Amplicons were separated on a capillary ABI 3730 instrument with the GS-500 LIZ size standard and GeneScan software (Applied Biosystems). Individual profiles were manually scored for the presence (1) or absence (0) of fragments with Genographer software (Benham, Reference Benham2001). Raw binary data were converted to Euclidian distance, with the resulting distance matrix used to calculate the population genetic distance using GeneAlEx (Peakall and Smouse, Reference Peakall and Smouse2006). The mean genetic distance matrix was the input for the assembly of a neighbour-joining (NJ) cladogram using PAUP* version 4.0b (Swofford, Reference Swofford2002). The six selective AFLP primer pairs generated 764 bands, of which 66% (501) had no polymorphisms while 34% were polymorphic. As would be expected for a highly self-pollinating species, 99% of the genetic variation was among the CS collections, while only 1% was within each CS source. Four accessions, all from the CIMMYT, Mexico collection, were grouped with Yecora Rojo (YR) and were distinct from all the other CS accessions (Fig. 1).
Fig. 1 NJ phylogram. The tree was based on the pairwise matrix of the mean binary genetic distance generated from 42 plants from 21 CS and YR wheat accessions using 764 AFLP bands.
Discussion
CS wheat (Triticum aestivum L.) is commonly used for genetic and molecular analyses, including genome sequencing, cytogenetics and germplasm development. Two salinity-tolerant wheat recombinant lines, W4909 and W4910 (Wang et al., Reference Wang, Larson, Horton and Chatterton2003b), were generated by crossing the disomic addition line AJDAj5 with the Ph inhibitor line PhI. AJDAj5 was developed in France and harbours a pair of Thinopyrum junceum chromosomes (Forster et al., Reference Forster, Miller and Law1988; Charpentier, Reference Charpentier1992; Wang et al., Reference Wang, Larson and Jensen2010). PhI was developed in Kansas and harbours the wheat Ph inhibitor gene derived from Aegilops speltoides that promotes homoeologous recombination (Chen et al., Reference Chen, Tsujimoto and Gill1994). Both the AJDAj5 and PhI lines were developed in CS wheat. W4909 and W4910 have been shown to contain non-CS22 DNA markers and both lines have greater salt tolerance than either parent (AJDAj5 and PhI), which, in turn, have greater salt tolerance than the CS22 background (Wang et al., Reference Wang, Li, Hu, Zhang, Larson, Zhang, Grieve and Shannon2003a; Mott and Wang, Reference Mott and Wang2007). Using subtraction suppression hybridisation, we isolated two genes, 4910D4 and 4909D4, in AJDAj5 and PhI that were not expressed in CS22 (Mott, unpublished data). Furthermore, microarray analysis (Mott and Wang, Reference Mott and Wang2007) identified additional genes not expressed in CS22, but expressed in both the AJDAj5 and PhI parental lines, which suggests that the source of these genes was CS, not the alien chromatin in AJDAj5 and PhI. Given that AJDAj5, PhI, W4909, and W4910 all share CS background, it seemed likely that our CS22 accession might have appreciable genetic differences with the very CS accessions used to generate PhI and AJDAj5, and, possibly, with CS accessions held at other laboratories.
CS seeds from 13 laboratories were screened for the presence or absence of PCR amplified bands from eight primer pairs designed from differentially expressed genes identified in the microarray (Mott and Wang, Reference Mott and Wang2007) or subtraction experiments that did not amplify in CS22 (Table 1). The results of these PCR assays ranged from primers that amplified from all CS accessions (actin and Ta25111) to primers that only amplified from four of the CS accessions. In every case that CS22 lacked amplification, at least one other CS accession did amplify with that primer pair. This supports our hypothesis that some genes not expressed in CS22, but expressed in W4909 or W4910, have CS origins, and did not originate from the alien chromatin present in PhI or AJDAj5.
AFLP data were used to further assess the genetic diversity among the accessions of CS wheat. The most prominent structure in the cladogram was four CIMMYT accessions grouped with YR that were distinct from all the other CS accessions (Fig. 1). Because YR has been used extensively in the CIMMYT breeding programme, some of its CS accessions might have been contaminated by outcrossing with YR. Interestingly, the salt-responsive gene Ta25815 is only present in W4910, PhI YR, and the three CS lines most similar to YR (CS13, CS16 and CS17) (Table 1 and Fig. 1). Only one accession (CS19) of the CIMMYT collection is similar to the other CS accessions from 12 different laboratories. Variations among these CS accessions could have resulted from seed contamination, outcrossing and a variety of mutations involving transposable elements in the largely repetitive DNA. Because variation exists among the CS accessions maintained in different laboratories, appropriate actions should be exercised when using CS plants to conduct any research. The exact parental plant used in making hybrids needs to be properly identified and its seeds preserved for subsequent studies to act as historic controls. The genetic background of past and subsequent generations that resulted from different breeding programmes needs to be carefully documented to ensure valid collection and interpretation of experimental data.
Acknowledgements
The authors thank A. Lukaszewski, H. Ohm, P. Gustafson, W. Cao, J. Koenig, M. Sorrells, A. Martin, T. Koniarov, A. Graner, T. Payne, M. Mackay and J. Raupp for contributing the CS seeds for this study, and Kim Thorsted for AFLP genotyping.
