This study used BC₁F₅ seeds harvested from 392 BC₁F₄ MSD panel (Elbashir et al., 2017b), which was developed through crossing and backcrossing of the Japanese common wheat cultivar “Norin 61” (hereafter referred to as N61) with 43 lines of SHW (Tsujimoto et al., 2015; Gorafi et al., 2018). The 43 lines of SHW were derived from crosses between 43 diverse genotypes of Ae. tauschii and T. turgidum var. durum cv. “Langdon” (LDN; Matsuoka and Nasuda, 2004; Kajimura et al., 2011).

We detected HMW-GS in the 43 lines of SHW and subsequently used the data to confirm HMW-GS in the MSD panel. To identify the HMW-GS' alleles in the 392 MSD panel and the 43 lines of SHW, the recurrent parent N61 and LDN were used in each electrophoresis assay.

The 392 BC₁F₄ MSD panel was grown in season 2015/2016 in the field of the Arid Land Research Center, Tottori, Japan (35°32′N, 134°13′E, 11 m above sea level), where the soil contains 95% sand, 1.3% silt, and 3.7% clay (Fujiyama and Nagai, 1989). The field experiment was arranged in an augmented randomized complete block design with eight blocks and four replicated checks, one of which was the recurrent parent N61. The plot size was one row with five plants spaced 0.2 m apart. Before sowing, three types of fertilizers were used: Kumiai Fukugo PKN 366 at a rate of 60 kg ha⁻¹ (MC FERTICOM Co., Ltd., Japan), Hitachi Fukugo 1 at a rate of 40 kg/ha (HITACHI CHEMICAL INDUSTRIES Co., Ltd., Japan), and granular carbonated magnesium lime at a rate of 100 kg/ha (SHIMIZU INDUSTRIAL Co., Ltd., Japan). At the tillering stage, the fertilizer Koudokasei 444 (Mitsubishi Shoji Agri-Service Co., Japan) was used at a rate of 500 kg/ha.

The composition of HMW-GS in the MSD panel was determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) according to the method used by Tanaka et al. (2003). The process of staining and decolorizing the gel was done according to that used by Dyballa and Metzger (2009). The gel was scanned using an image scanner (ES-2200, Seiko Epson Co., Japan).

Since the BC₁F₅ seeds in the MSD panel might still be genetically segregated, we investigated the composition of HMW-GS in three grains per line. We considered the MSD lines to have HMW-GS introduced from the SHW if at least one out of the three tested seeds had the HMW-GS of SHW.

The HMW-GS alleles at Glu-A1, Glu-B1, and Glu-D1 loci were identified based on the numbering system of Payne and Lawrence (1983). The subunits derived from Ae. tauschii were followed by the superscript designation “t” to refer to their origin. The nomenclature of subunits derived from Ae. tauschii is tentative in this paper because it is only based on electrophoresis and has not been confirmed by DNA analysis.

Grain yield (g) per plant was calculated from an average of five plants. Whole wheat flour was obtained by grinding 4 g from each line of the MSD panel using a UDY cyclone sample mill (UDY Co., USA) equipped with a 1-mm screen. The protein content of the samples was measured as a percentage of the total weight by near-infrared spectroscopy (NIR composition analyzer KJT-270, Kett Electric Laboratory Co. Ltd., Japan). To assess the gluten quantity and quality, we measured the SDS sedimentation volume in 1 g of flour, using the method according to Takata et al. (1999). The sedimentation volume is highly correlated with bread loaf volume (Axtord et al., 1979), where dough strength is the main factor. For lines that derived their HMW-GS from Ae. tauschii, specific sedimentation values, which are highly correlated with dough strength, were assessed as an index of gluten quality by dividing the SDS sedimentation volume (mL) by protein content (%), because protein content is reported to be highly correlated with sedimentation volume (Moonen et al., 1982; Tanaka and Tsujimoto, 2012).

Data were tested for normality and homogeneity of variance before analysis using the Shapiro–Wilk Test and Levene's Test, respectively. Analysis of variance (ANOVA) was performed for dough strength and protein content of the MSD lines using the GenStat Software program (18th edition). The least significant difference (0.05) was used for mean separation. Duncan's Multiple Range Test was used to compare the mean dough strength of the different HMW-GS combinations using SPSS Software (version 25.0.1). ANOVA for the field experiment was performed using Plant Breeding Tools v. 1.4 software (International Rice Research Institute, http://bbi.irri.org/products). Regression analysis was conducted using Microsoft Excel 2019.

At Glu-A1 locus, 307 lines possessed subunit 2^* inherited from N61, and 85 lines had a null allele derived from the LDN genome in SHW (Table 1). At Glu-B1 locus, 288 lines had the subunit pair 7+8 from N61, and 104 lines possessed the subunit pair 6+8 from LDN. As for the Glu-D1 locus, 289 lines inherited the 2.2+12 subunit pair from N61, whereas 103 lines inherited their Glu-D1 subunit pair from SHW harboring Ae. tauschii.

Given the nature of the MSD development method, the expected segregation ratio was 75% from N61 and 25% from SHW. Our result for the frequency of HMW-GS pairs at Glu-A1, Glu-B1, and Glu-D1 loci in the MSD panel fitted to the Mendelian expected ratio of 3:1. This indicates that no special selection occurred for any of the alleles during the development of the MSD panel and that no identical alleles with that of N61 (2.2+12) have been found in Ae. tauschii.

In the 103 lines harboring HMW-GS at Glu-D1 derived from Ae. tauschii, three subunits were x-type (2.1^t, 2^t, and 5^t) and three were y-type (10^t, 12^t, and 12.1^t) (Figure 1). Subunit 12.1^t had slightly smaller molecular weight and faster mobility than subunit 12 in N61, whereas subunit 2.1^t had slower mobility than subunit 2 in N61. The HMW-GS were found in the form of five different haplotypes, 2^t+10^t, 2^t+12^t, 5^t+10^t, 2.1^t+12^t, and 2^t +12.1^t. The most frequent pair of HMW-GS was 2^t+12^t (42 lines), followed by 5^t+10^t (30 lines) (Table 2). The subunit pair 2^t+12.1^t was found in 21 lines, 2^t+10^t in nine lines, and 2.1^t+12^t in only one line.

In all, 16 combinations of HMW-GS at the three Glu loci were distinguished in the 103 MSD lines. The most frequent combination was 2^*, 7+8, 2^t+12^t, which was observed in 25 lines, followed by 2^*, 7+8, 2^t+12.1^t, and 2^*, 7+8, 5^t+10^t which were observed in 18 and 17 lines, respectively (Table 2).

The MSD lines used in this study originated from three lineages of Ae. tauschii. The 60 MSD lines from TauL2 Ae. tauschii contained all combinations of HMW-GS except 2.1^t+12^t (Figure 2). The most common subunit pair was 2^t+12^t, which was found in 27 MSD lines in TauL2; the subunit pair 2^t+12.1^t was exclusively found in 21 lines from TauL2; and subunit pairs 5^t+10^t and 2^t+10^t were found in seven and five lines, respectively. All 15 lines of the TauL3 lineage contained the subunit pair 2^t+12^t. In TauL1, the 28 lines possessed three pairs of HMW-GS (5^t+10^t, 2^t+10^t, and 2.1^t+12^t); the most common subunit pair was 5^t+10^t (23 lines), followed by the subunit pair 2^t+10^t (four lines), and the subunit pair 2.1^t+12^t was found in only a single line from TauL1. TauL1 had no lines that possessed subunit pair 2^t+12^t, although it is the most abundant subunit in MSD lines (Table 2).

Highly significant (P < 0.001) differences for dough strength were found among the 103 MSD lines that derived their HMW-GS from Ae. tauschii (Table 3 and Figure 3). Moreover, dough strength for MSD lines showed a normal distribution according to the Shapiro–Wilk normality test (P < 0.05). The variation in dough strength among MSD lines ranged from weak to strong (0.232–0.732 mL/%). In comparison with N61, 3.9% of MSD lines (4 lines, viz. MSD272, MSD363, MSD219, and MSD61) (Supplementary Table 2) showed dough strength significantly higher than N61. A total of 42 MSD lines (40.8%) showed dough strength comparable with that of N61, whereas 55.3% of MSD lines (57 lines) showed significantly lower dough strength than N61.

MSD lines with subunit pair 2^t+12^t exhibited the widest variation in dough strength (0.31–0.68 mL/%) followed by those that carried subunit pairs 5^t+10^t, 2^t+12.1^t, and 2^t+10^t (Figure 4).

The mean dough strength of the recurrent parent N61 was significantly higher than the mean dough strength for lines possessing subunit pair 5^t+10^t which exhibited the lower dough strength average among all subunits derived from Ae. tauschii. The dough strength of the single line that carried subunit pair 2.1^t+12^t was notably higher than the means of the other HMW-GS pairs derived from Ae. tauschii (2^t+12.1^t, 2^t+10^t, 2^t+12^t, and 5^t+10^t), and was also higher than that of the recurrent parent N61 although this difference was not significant. There was variation inside each subunit pair where some lines were comparable, and others were significantly lower or higher than N61. For example, although the means dough strength for the subunits 2^t+12.1^t and 2^t+12^t was comparable to N61, there were two lines from both subunits that were significantly higher than N61 and 13 and 14 comparable, six and 24 lines were significantly lower than N61 respectively (Figure 4).

To explore the impact of the Glu-D1 locus on dough strength and exclude the effects of Glu-A1 and Glu-B1 loci, we calculated the mean dough strength of the lines that had the same subunits at Glu-A1 and Glu-B1 loci but different subunits at Glu-D1 locus (Figure 5). Accordingly, lines were divided into four groups: (i) Group 1 consisted of lines with subunit combination of 2^* and 7+8 at Glu-A and Glu-B loci, respectively, and four subunit pairs at Glu-D1 locus: 2^t+12.1^t, 2^t+12^t, 2^t+10^t, and 5^t+10^t; (ii) Group 2 comprised of lines with subunit combination of 2^* and 6+8 at Glu-A and Glu-B loci, respectively, together with four subunit pairs at Glu-D1 locus: 2^t+12.1^t, 2^t+12^t, 2^t+10^t, and 5^t+10^t; (iii) Group 3 consisted of lines showing null subunits and 7+8 at Glu-A and Glu-B loci, respectively, and five different subunit pairs at Glu-D1 locus: 2.1^t+12^t, 2^t+12^t, 2^t+12.1^t, 2^t+10^t, and 5^t+10^t; (iv) Group 4 consisted of subunits null and 6+8 at Glu-A and Glu-B locus, respectively, and three different subunit pairs at Glu-D1 locus: 2^t+12^t, 2^t+12.1^t, and 2^t+10^t (Figure 5).

We found that MSD lines with the same subunit combination at Glu-A1 and Glu-B1 loci showed a wide variation in their dough strength values when the subunit pair at Glu-D1 was different. In Group 1, N61 and the MSD lines that had subunit pair 2^t+12.1^t at Glu-D1 locus had significantly higher mean dough strength of than lines with subunit pairs 2^t+10^t and 5^t+10^t (Figure 5). In Group 2, no significant differences were observed in the dough strength of the MSD lines. In Group 3, N61 and two MSD lines that had subunit pairs 2.1^t+12^t and 2^t+10^t at Glu-D1 locus had significantly higher mean dough strength than those with subunit pairs 2^t+12.1^t, 2^t+12^t, and 5^t+10^t. In Group 4, no significant differences were observed in the dough strength of the MSD lines, but all had significantly lower dough strength than that of N61. When results in Group 1 were considered in detail, two MSD lines with the subunit combination of 2^*, 7+8, 2^t+12.1^t showed significantly higher dough strength than that of N61, whereas 12 lines were comparable and 4 lines were significantly lower than N61 respectively. Interestingly, two lines out of 4 lines that were significantly lower than N61 were developed from the same Ae. tauschii accession of the superior lines (MSD272 and MSD219) of the same combination (2^*, 7+8, 2^t+12.1^t). For the subunit combination of 2^*, 7+8, 2^t+12^t, two lines were significantly higher than N61, 10 were comparable and 13 lines were significantly lower than N61 respectively. Also, out of the 13 lines that were significantly lower than N61 in this combination there was three and two lines were developed from the same Ae. tauschii accession of the superior lines MSD61 and MSD363 respectively. On the other hand, all lines with the subunit combinations of null, 7+8, 5^t+10^t and null, 7+8, 2^t+12^tshowed significantly lower dough strength than N61 (Figure 5).

To explore the impact of introgressions from Ae. tauschii on the yield and the quality, we performed a regression analysis for protein content and grain yield/plant (Figure 6). Results showed no relationship between the two traits (r = 0.046; P < 0.6438). The grain yield/plant in all MSD lines did not differ from N61 (according to the least significant difference of 0.05) (Table 3). For the protein content, 69 lines were significantly higher (Supplementary Table 2), 20 lines were significantly lower, and 15 were comparable to N61. Thus, we separated the MSD lines into three categories (A), (B), and (C) based on significant differences in protein content.

In this study, we reported a large variation in HMW-GS and flour quality in MSD lines derived from 43 Ae. tauschii accessions in the background of the Japanese wheat cultivar N61.

Our investigation at Glu-D1 locus revealed five types of HMW-GS pairs derived from Ae. tauschii. Sixteen subunit combinations at Glu-1 loci were distinguished in the 103 MSD lines. This result is considered representative of the wide diversity in Ae. tauschii across their natural habitat.

Our result on the frequency of HMW-GS at Glu-A1, Glu-B1, and Glu-D1 loci in the MSD panel (Table 1) indicated that the SHW genome was successfully introgressed into the recurrent parent (N61) through self-pollinating and backcrossing, and that no identical alleles with that of N61 (2.2+12) have been found in Ae. tauschii. The subunit 2.2 (x-type subunit) has not been observed in any previous study of Ae. tauschii, which supports the hypothesis of Payne et al. (1983) that this subunit arose within hexaploid wheat by a rare unequal crossing-over with another HMW-GS gene. Since there is a tight link between x-type and y-type subunits in bread wheat (Lawrence and Shepherd, 1980), there is no possibility that some lines can simultaneously derive subunit 12 from N61 (which is a y-type subunit) and x-type subunits from Ae. tauschii. This result agrees with Gorafi et al. (2018) who analyzed the crossing over status of MSD individuals and found the result consistent with the expected ratio after one backcross event.

In our study, we used SDS-PAGE analysis, the traditional standard for distinguishing HMW-GS; thus, the identification of the five subunits that derived from Ae. tauschii was based on previously reported electrophoresis mobility results. For instance, William et al. (1993), Rasheed et al. (2012), and Tariq et al. (2018) reported similar SDS-PAGE mobility of 5^t+10^t and 2^t+12^t derived from Ae. tauschii accessions. Yan et al. (2003) observed similar SDS-PAGE mobility of 2^t+10^t and 2^t+12^t derived from Ae. tauschii accessions. Similarly, the subunit pair 2.1^t+12^t has been identified according to its SDS-PAGE mobility (Pflüger et al., 2001; Rasheed et al., 2012; Tariq et al., 2018). Our finding was in agreement with Lagudah et al. (1987), who documented the presence of subunit pairs 5^t+10^t and 2^t+12^t, which were equivalent to the SDS-PAGE bands of common wheat, and they also found subunit 2.1^t in conjunction with subunit 10.1^t. Gianibelli et al. (2001) reported for the first time the presence of subunit 2^t+12.1^t by SDS-PAGE in Ae. tauschii accessions. Our results contrasted with the findings of Delorean et al. (2021), who documented the absence of the 5+10 wheat haplotype in 273 sequenced Ae. tauschii accessions at both the molecular level and also by SDS-PAGE mobility.

The most frequent subunit pair in our study was 2^t+12^t (40.8%), which is the most abundant pair in common wheat as well (Payne, 1987), followed by subunit pair 5^t+10^t (29.1%) (Table 2). The same pattern of frequency for these two pairs (2^t+12^t and 5^t+10^t) has been observed in primary SHW derived from 52 Ae. tauschii accessions (Tariq et al., 2018). Rasheed et al. (2012) reported a lower frequency for the subunit pair 2^t+12^t and a higher frequency for 5^t+10^t in 95 selected synthetic lines developed by the International Maize and Wheat Improvement Center. Subunit pairs 2^t+10^t and 2.1^t+12^t were found in 9 lines (8.7%) and only one line (1%), respectively. In contrast to our result, the subunit pair 2^t+10^t has been found at the higher frequency of 12.63% in 198 Ae. tauschii accessions (Yan et al., 2003), and 2.1^t+12^t has been reported in six accessions in a 92 accessions of Ae. tauschii accessions (Pflüger et al., 2001) and at a frequency of 16.8% in a 95 synthetic hexaploid accessions (Rasheed et al., 2012).

TauL2 exhibited the widest diversity at the Glu-D1 locus, compared to TauL1 and TauL3. MSD lines that belonged to TauL2 contained all types of HMW-GS derived from Ae. tauschii except the subunit 2.1^t+12^t. All MSD lines belonging to TauL2 originated from Iran. This diversity at Glu-D1 alleles in TauL2 matched well with literature that considers Iran to be the center of genetic variation of Ae. tauschii (Dudnikov and Goncharov, 1993). Also, Delorean et al. (2021) evaluated the Glu-D1 diversity relative to the geographic origin of Ae. tauschii accessions and found that the greatest concentration of haplotype diversity was located along the shores of the Caspian Sea in Iran. Similarly, Lagudah and Halloran (1988) reported that the northeastern region of Iran exhibited a wide diversity of the Glu-D1 subunits. Therefore, identifying geographical areas where the progenitor species of existing SHW were collected would assist in guiding future collection missions (Ogbonnaya et al., 2013).

Subunit 12.1^t exclusively belongs to the Ae. tauschii genome and does not exist at the Glu-D1 locus in common wheat (Tahernezhad et al., 2013). In our study, we found the subunit pair 2^t+12.1^t exclusive to TauL2, which might indicate that this subunit pair has a unique origin, but further studies are needed to confirm this. All lines that belonged to TauL3, which originated from Georgia, carried subunit pair 2^t+12^t and were genetically similar to TauL2 (Matsuoka et al., 2013). Delorean et al. (2021) studied gene-level phylogeny at Glu-D1 for 273 sequenced Ae. tauschii accessions and showed that a unique group of Glu-D1 alleles belonging to Lineage 3 accessions was found within a narrow clade with Lineage 2.

The most common 2^t+12^t subunit pair was not found in TauL1. This may indicate that Ae. tauschii genotypes belonging to this TauL lineage may not have been involved in the evolution of common wheat. Indeed, it appears that TauL2 and TauL3 are closer to common wheat than TauL1 because they contain the most common HMW-GS allele prevalent in common wheat (2^t+12^t) (Matsuoka et al., 2013). Moreover, Delorean et al. (2021) demonstrated that the superior subunit pair 5+10 was found to be clustered very tightly with TauL3, whereas the wheat subunit pair 2+12 was found to be clustered with TauL2, indicating the contribution of these two lineages (TauL2 and TauL3) to the current wheat genome.

We evaluated the dough strength of the 103 MSD lines that derived their HMW-GS from Ae. tauschii to explore the effect of the wild gene in the background of N61. The significant genotypic differences (P < 0.001) observed in dough strength indicated high genetic diversity among the MSD lines. This variation has been attributed mainly to different introgression segments of Ae. tauschii in the MSD lines (Itam et al., 2021a). Itam et al. (2021b) and Elbashir et al. (2017a) also found high genetic diversity in the MSD lines for different traits and have attributed these variations to Ae. tauschii.

We calculated the mean dough strength for each of the five pairs of HMW-GS derived from Ae. tauschii. In our study, the subunit 2.1^t exhibited the strongest dough strength average in combination with subunit 12^t, and this pair was significantly higher than 5^t+10^t, 2^t+10^t, and 2^t+12^t. This indicates the positive impact of this subunit when combined with subunit 12^t; however, it has been reported to have a weak contribution to specific rheological characteristics when associated with subunit 10.1^t in SHW (Lagudah et al., 1987). The subunit pair 2^t+12.1^t also showed relatively strong dough strength. Two MSD lines possessing this subunit pair showed significantly stronger dough compared to N61, indicating the positive impact of this subunit pair on dough strength. To the best of our knowledge, this is the first time that the effect of subunit pair 2^t+12.1^t in Ae. tauschii on wheat quality has been studied in the background of a wheat cultivar.

Payne (1987) proved that the subunit pair 5+10 at Glu-D1 locus has the highest positive effect on dough strength and a higher Glu-1 score compared to 2+12. This explains its frequent association with dough characterized by stronger elasticity and superior end-use qualities for bread making. Results in our study showed that lines carrying subunit 5^t+10^t exhibited the lowest dough strength values among all subunits at Glu-D1 and were significantly lower than N61. For the first time, the impact on dough strength of subunit pair 5^t+10^t inherited from Ae. tauschii in the background of a wheat cultivar is reported as poor. Previous studies documented that the subunit pair 5^t+10^t, which has mobility in SDS-PAGE typical to bands of 5+10 in wheat derived from Ae. tauschii, is associated with good bread-making quality (Lagudah et al., 1987; Hsam et al., 2001). We suggest that the decreased dough strength values in this study of lines carrying subunit pair 5^t+10^t could be due to the lack of the extra cysteine in 1Dx5. The absence of an additional cysteine in subunit 1Dx5^t derived from Ae. tauschii has been reported in previous studies (Pflüger et al., 2001). It is also possible that the subunit pair (5^t+10^t) present in our materials might be a different pair of the same size—and therefore with the same mobility in SDS-PAGE—as the subunits 5+10 in T. aestivum. Delorean et al. (2021) used haplotype molecular sequence diversity and SDS-PAGE to explore whether the electrophoresis mobility would reflect the differences visible at the molecular level, and observed that in some cases genes showed a difference in SDS-PAGE mobility, but the alleles looked identical at the molecular level. Similarly, haplotypes were seen to be different at the molecular level but identical in SDS-PAGE mobility. Interestingly, the presence of an Ae. tauschii haplotype identical to that of wheat subunit pair 2+12 and the absence, even by SDS-PAGE mobility, of the exact wheat 5+10 haplotype in the Ae. tauschii accession has been documented (Delorean et al., 2021). Likewise, the difference in basic isoelectric value between subunit 5^t+10^t derived from Ae. tauschii and the same subunit pair in common wheat has been confirmed by a two-dimensional method (isoelectric focusing with SDS-PAGE), although these subunit pairs had identical electrophoretic mobility (Lagudah and Halloran, 1988). Mackie et al. (1996) reported that the subunits Dy10^t and Dy12^t from Ae. tauschii were more hydrophobic than those from T. aestivum. Yan et al. (2003) reported a difference in relative mobility by A-PAGE between subunits Dx5^t and Dy10^t from Ae. tauschii and those of common wheat, although they showed the same mobility under SDS-PAGE. Thus, the unexpectedly poor impact of subunit 5^t+10^t from Ae. tauschii on dough strength could be attributed to the lack of an extra cysteine or to difficulties in interpreting SDS-PAGE mobility. More confirmatory investigations are needed.

The same subunit may play varied roles on wheat quality in different pairs of HMW-GS (Zhao et al., 2020). Our result showed a wide variation even between the same HMW-GS pair. This variation in the same subunit pair, and even between sister lines with the same subunit pair, indicates that there might be other factors/genes that affect dough strength. It might be due to the different recombinant portions from SHW (introgressed segments) across the 21 chromosomes. The presence of different introgressed segments from SHW in MSD lines, including sister lines, has been documented (Itam et al., 2021a). These different genomic segments have been found to cause a variation in physio-agronomic traits between MSD lines (Itam et al., 2021a). In further study, a genome-wide association study will need to be performed to find factors/genes other than HMW-GS. Furthermore, LMW-GS and gliadin, which were not investigated in this study, are known to contribute markedly to flour quality, sometimes even more so than the HMW-GS (Pogna et al., 1982; Gupta et al., 1989). Therefore, revealing the allelic compositions of LMW-GS and gliadin is very important to better understand the observed variations in wheat quality.

Two lines (MSD61 and MSD363) (Supplementary Table 2 and Figure 4) that carried 2^t+12^t exhibited good dough strength despite the fact that this subunit pair is frequently associated with weak dough strength (Payne, 1987). This might be attributed to their high proportion of total HMW-GS at the Glu-D1 locus, as has been reported earlier (Horvat et al., 2006).

We observed that two lines that carried the null allele at the Glu-A1 locus showed dough strength values higher than the recurrent parent N61. This suggested that the introgression of the Glu-D1 locus from Ae. tauschii compensated for the negative impact of the null allele in these lines, where most of the studies reported a significant negative effect of null allele and its association with lower values of gluten strength (Ruiz and Carrillo, 1993; Raciti et al., 2003). Thus, those lines could be used in breeding programs to improve the quality and overcome the negative impact of the null allele.

Although the negative relationship between grain protein content and grain yield is well-known (Kibite and Evans, 1984; Cox et al., 1985; Gauer et al., 1992; Delzer et al., 1995; Marinciu and Suaulescu, 2008; Giancaspro et al., 2019; Taheri et al., 2021), our findings showed no relationship between the two traits in MSD lines. Moreover, most of the MSD lines had higher or lower protein content with comparable grain yield values to the recurrent parent. This may indicate that the increase or decrease in protein content that occurred due to the introgression of the D genome is independent of the grain yield in MSD lines. Also, it may have increased the variation in protein content to such an extent that it counteracts the generally known negative relationship between protein content and grain yield. Although our finding is based on a homogenous grain yield and protein content of five independent plants evaluated for one season under optimum condition, it is very promising and pave the way for more detailed investigation and validation. The regression analysis used was powerful and allowed the classification of the MSD lines in different groups considering their protein content and grain yield. Some lines in group C had a clear good comparable grain yield and high protein content compared to N61, these lines could be a target for more detailed analysis and evaluation to elucidate the basis of the positive or no correlation between the grain yield and protein content especially that breaking the negative relationship between these traits is an important aspect for wheat breeding to increase the grain yield and maintain the quality characteristics. The identified MSD lines could provide a valuable genetic resource for enhancing the end-use quality without any loss in productivity.