Overall 90.4% bread wheat production (Triticum aestivum L.) in Mexico is processed by the domestic milling industry, thus physical and chemical variables of grain quality must be met, as well as dough strength and extensibility (^{CANIMOLT, 2010}). These last two variables are partially dependent on quantity and quality of the glutenin and gliadin proteins, present in the flour (^{Souza et al., 2013}), which are encoded by the Glu-A1, Glu-A3, Gli-A1 y Gli-A2 loci (^{Payne and Lawrence, 1983}). Since glutenins and gliadins are under genetic control, they can be manipulated and recombined through plant breeding.

Several national (^{Martínez-Cruz et al., 2010b}) and international (^{Liang et al., 2010}) studies indicate the effect of HMW-G, LMW-G and gliadins on dough strength and extensibility. A prerequisite in breeding programs is knowledge about the germplasm used in crosses to generate variability for the desired trait. In Mexico, developing an improved bread wheat variety require the combination of high yield potential, drought tolerance, disease resistance (yellow and leaf rust) and specific industrial quality (bread and biscuits preparation). Therefore this research aimed to identify the HMW-G and LMW-G alleles in the parent group genotypes from the rainfed bread wheat program of the National Institute for Forestry, Agriculture and Livestock (INIFAP).

Overall 99 genotypes from the parent group in the INIFAP rainfed wheat program were used. The materials were established during the spring-summer season 2011 in Chapingo, Mexico. The experimental plot consisted of four rows, 3 m long with a 30 cm spacing among them. Electrophoretic analysis was performed on 40 mg of flour, obtained from an UDY cyclone mill with a 0.5 mm mesh, grinding grains from one spike per plot. The electrophoretic method used was described by ^{Peña et al. (2004)}, using polyacrylamide gels in basic pH conditions and sodium dodecyl sulfate as denaturing agent. The HMW-Gs were identified by the nomenclature proposed by ^{Payne and Lawrence (1983)} and the LMW-Gs at Glu-A3 and Glu-B3 were designated according to ^{Singh et al.(1991)} and ^{Jackson et al. (1996)}, while Glu-D3 alleles were named according to ^{Branlard et al. (2003)}. Allele frequencies were calculated for the Glu-A1, Glu-B1, Glu-D1, Glu-A3, Glu-B3 and Glu-D3 loci as well as for the HMW-G and LMW-G combinations.

The highest HMW-G polymorphism was observed in the Glu-B1 locus, while Glu-A1 and Glu-D1 loci had lower and equal number of allelic variants. The most frequent allele were: 2* in Glu-A1, 17+18 in Glu-B1 and 5+10 in Glu-D1, with 71.1, 52.5 and 82.8% respectively (Table 1). These frequencies are comparable to those reported by ^{Liang et al. (2010)} who evaluated 273 genotypes from the irrigation program of the International Maize and Wheat Improvement Center (CIMMYT). Less frequent variants were 0 in Glu-A1, 13+16, 7*+15 and 7+8 in Glu-B1 and 2+T2/1.5+T2 in Glu-D1. Importantly, the 7*+15 and 2+T2/1.5+T2 variants, had not been previously identified in the parent group, according to reports from ^{Espitia et al. (2008)} and ^{Martínez-Cruz et al. (2010a)} who analyzed them in the spring-summer 2001 and spring-summer 2008 cycles respectively.

Table 1 HMW-G allele frequency in the parent group from the INIFAP national rainfed bread wheat program. Chapingo spring-summer 2011. 

For 2+T2/1.5+T2 a genotypic mixture of two alleles was found (separated by the slash), which comes from Triticum tauschii L. one of the wheat ancestors (^{Pilch, 2007}). The Figure 1 shows the variants encoding LMW-G. Greater -polymorphism was observed at locus Glu-B3 where seven allelic variants and four allele mixtures were identified. Five loci were identified at each of the Glu-A3 and Glu-B3 loci. Thus, the c alleles in Glu-A3, h in Glu-B3 y b in Glu-D3, appeared in 57.6, 38.4 and 59.6% of the genotypes analyzed, respectively. Individual alleles of lowest frequency were f in Glu-A3, j in Glu-B3 as well as d in Glu-D3. The h allele in Glu-B3 was found at similar frequencies to genotypes evaluated by ^{Liang et al. (2010)}, while the j variant associated with adverse effects on dough strength and extensibility (^{Martínez-Cruz et al., 2012}) appeared at low frequency.

Figure 1 LMW-G allele frequency in the parent group from the INIFAP national rainfed bread wheat program. Chapingo spring-summer 2011. 

The HMW-G 2*, 17+18, 5+10 combination was the most frequent (30.3%), followed by 2*, 7+9, 5+10 and 1, 17+18, 5+10 with 15.2% and 14.1% , respectively, while there were nine combinations that occurred in 1% (Table 2). For the LMW-G, the c, h, b combination showed the highest frequency, appearing in 15 genotypes, followed by c, h, b and c, b, b associated with 6 materials. Furthermore 23 combinations appeared with percentages equal to 1, Figure 2. In addition, the 2*, 17+18, 5+10 and c, h, b combinations, in higher proportion, showed alelles suitable for line development for bakery products consistent with reports from ^{Martinez-Cruz et al. (2010b)}. But also the 2*, 17+18, 2+12 and c, g, b combinations could recombine to generate genotypes suitable for biscuits production.

Table 2 HMW-G combinations frequency in the parent group from the INIFAP national rainfed bread wheat program. Chapingo spring-summer 2011. 

Figure 2 LMW-G combinations frequency in the parent group from the INIFAP national rainfed bread wheat program. Chapingo spring- summer 2011. 

The parent group from the INIFAP rainfed bread wheat, contains allelic variants and several HMW-G and LMW-G combinations, from which progenies can be derived, with specific quality for bread, biscuit and pastry products preparation. Also two new alleles were identified, broadening the HMW-G genetic diversity at the Glu-B1 and Glu-D1 loci.