A Substantial Fraction of Barley (Hordeum vulgare L.) Low Phytic Acid Mutations Have Little or No Effect on Yield across Diverse Production Environments

The potential benefits of the low phytic acid (lpa) seed trait for human and animal nutrition, and for phosphorus management in non-ruminant animal production, are well documented. However, in many cases the lpa trait is associated with impaired seed or plant performance, resulting in reduced yield. This has given rise to the perception that the lpa trait is tightly correlated with reduced yield in diverse crop species. Here we report a powerful test of this correlation. We measured grain yield in lines homozygous for each of six barley (Hordeum vulgare L.) lpa mutations that greatly differ in their seed phytic acid levels. Performance comparisons were between sibling wild-type and mutant lines obtained following backcrossing, and across two years in five Idaho (USA) locations that greatly differ in crop yield potential. We found that one lpa mutation (Hvlpa1-1) had no detectable effect on yield and a second (Hvlpa4-1) resulted in yield losses of only 3.5%, across all locations. When comparing yields in three relatively non-stressful production environments, at least three lpa mutations (Hvlpa1-1, Hvlpa3-1, and Hvlpa4-1) typically had yields similar to or within 5% of the wild-type sibling isoline. Therefore in the case of barley, lpa mutations can be readily identified that when simply incorporated into a cultivar result in adequately performing lines, even with no additional breeding for performance within the lpa line. In conclusion, while some barley lpa mutations do impact field performance, a substantial fraction appears to have little or no effect on yield.

. Seed dry weight and seed phosphorus (P) fractions in 21 barley lines consisting of the wild-type control (cv. Harrington) and twenty mutants that display the "high inorganic P" seed phenotype. Seed was obtained from the 2004 Aberdeen, Idaho nursery. Please see Supplementary Methods for details of whole-grain and "half-seed" analysis and other methods used.  The barley cultivars "Morex" [6] and "Steptoe" [7] are widely used in barely mapping and genomics research and represent distinct germplasm pools as compared with cv. Harrington [8]. Therefore for chromosomal mapping, F2 populations were derived from crosses between each Harrington-derived seed P mutant and both Morex and Steptoe. F2's were grown at the Aberdeen Research and Extension Center in 2002, and in subsequent greenhouse nurseries, to provide populations ranging in number from approximately100 to 200 recombinants. F3 progeny seed for each F2 plant were harvested and stored until analysis.
Segregation analysis for each seed P mutation was conducted by testing individual F3 seed for the "high inorganic P" typical of each mutation. As homozygotes these mutations condition increases in seed inorganic P as compared with a heterozygote or homozygous wild-type seed that are typically large enough to permit clear-cut scoring on an individual kernel basis [9,10]. Typically a minimum of 30 individual kernels from each F3 progeny were tested. If all or nearly all (>27) F3 seed displayed the seed inorganic P phenotype typical of the parental mutant, the F2 parent plant was scored as homozygous mutant. If all F3 seed were scored as wild-type, then the F2 parent plant was scored as homozygous wild-type. If segregation for the HIP phenotype was observed in the F3 (three or more kernels per 30 F3 kernels scored as mutant/high inorganic P, with the remainder intermediate or wild-type/low inorganic P), then the parent F2s were scored as heterozygotes. If the result of the initial round of seed testing was inconclusive, additional F3 seed testing was conducted. Data were collected using either images of colorimetric reactions or digital readings from Synergy™ 2 Multi-Detection Microplate Reader (BioTEK, Winooski, VT, USA). For mapping purposes, F2-3 families homozygous for either mutant or wild-type alleles from each segregating population were used to simplify the mapping score. Leaf samples were collected, immediately frozen in liquid N2 and stored at −80 °C.
PCR-based markers linked to each of the three previously-known Hvlpa loci and the barley genome's single gene encoding myo-Inositol-3-P1 synthase (MIPS) [11] were selected based on the previously published data [9][10][11]. For the Hvlpa1 locus on barley chromosome 2H, aMSU21 was identified as a closely linked STS-PCR marker [9,12]. New primer pairs for the aMSU21 locus were designed; aMSU21 Forward (5'-tggtctttcatgtacctacc-3') and aMSU21 Reverse (5'-tgtgtcatcaagcacaacca-3'). The newly designed primer pairs amplify a single strong 435 bp PCR product from Steptoe and Morex that is slightly shorter than the original 449 bp product, and about 200 bp bigger PCR products from Harrington and its derived mutations. The original primer pairs of aMSU21 detected two major fragments from each cultivar [12] while our modified primer pairs detected one major fragment from each genotype (data not shown). Thus the modified marker simplified the genotype scoring. The barley MIPS locus on chromosome 4H was first mapped close to the BCD453 marker [11]. The EBmac701 SSR marker was identified as flanking BCD453B (Barley Consensus Map 2005, GrainGenes, http://wheat.pw.usda.gov/cgi-bin/graingenes/report.cgi?class=mapdata&name=Barley,+Consensus+ 2005,+SNP) [11]. The EBmac701 primer pairs listed in GrainGenes (Forward, 5'-atgatgagaactcttcaccc-3' and Reverse, 5'-tggcactaaagcaaaagac-3') amplified a single strong PCR product. The polymorphic patterns were clearly scored as a ~150 bp product in Harrington and slightly smaller, but clearly scorable fragments in Steptoe and Morex (data not shown).
For the Hvlpa2 locus on barley chromosome 7H, the previously described flanking RFLP markers were MWG2301 and ABC310b [9]. Bmag0120 was chosen based on its close location (Barley shedar2 map, Grain Genes) to ABC310b that was loosely linked to Hvlpa2-1 [9]. The Bmag120 primers listed in GrainGenes (Forward, 5'-atttcatcccaaaggagac-3' and Reverse, 5'-gtcacatagacagttgtcttcc-3') amplified a ~250 bp single product in Harrington and a slightly smaller product in Steptoe. For the Hvlpa3 locus on barley chromosome 1H, the RFLP probes cWMG706 and ABG702B were shown to flank the locus and the closely-linked LP75 sequence-specific ISSR primer was developed [10]. Since LP75 requires ClaI digestion of PCR products to detect polymorphism, the Bmag382 SSR marker, closely linked to cWMG706 (Igri × Franka map, GraineGenes) was chosen for this experiment. The Bmag382 primer pairs listed in GrainGenes (Forward, 5'-tgaaacccatagagagtgaga-3' and Reverse, 5'-tcaaaagtttcgttccaaata-3') amplified a strong single PCR product in all barley cultivars in this experiment. The PCR product from Harrington is slightly above the 100 bp size marker distinguishing the slightly smaller fragments from Steptoe and Morex (data not shown).
Leaf samples collected from F2-3 plants were lyophilized prior to DNA extraction with cetyl trimethyl ammonium bromide (CTAB). Pulverized leaf tissue in a 96-well format was extracted with the protocol utilized by [13]. The only modification to the procedure was the use of chloroform without isoamyl alcohol.
PCR reactions were set up in a 96-well format. Each 25 μL reaction contained 50 ng of template DNA, 2 µL of each primer (10 µM), 2.5 µL of 10× buffer containing 1.1 mM Mg, 1 µL of dNTPs with 2.5 mM concentration for each nucleotide, and 1 unit of Taq polymerase (RedTaq, Sigma). The PCR program was: 94 °C for 3 min followed by 39 cycles of 94 °C for 30 s, 50 °C for 30 s, and 72 °C for 1 min, followed by a 4 °C hold. PCR reaction products were analyzed on 3% Agarose SFR high resolution gels (Amresco, Solon, OH, USA) stained with ethidium bromide. A 100 bp DNA ladder (Bio-Rad, Richmond, CA, USA) was used as a size marker in the same gel.
The PCR fragments were scored for each individual F2 plant from segregating populations and compared to the genotype patterns of both parents. Mutant or wild-type scores were recorded for all the plants used. The genotype scores were then compared to the seed inorganic P phenotype as determined in F3 seeds derived from F2 plants described above.