Congruity of the Polymorphisms in the Expressed and Noncoding Parts of the Gli-B1 Locus in Common Wheat

: The previously deﬁned pairs of primers GliB1.1 and GliB1.2 were found to produce three and four principal variants, respectively, of PCR sequence length for the γ -gliadin pseudogene in 46 Triticum aestivum L. cultivars from 15 countries carrying 19 known alleles at the Gli-B1 locus. A congruity was established between this polymorphism, allelic sets of the Gli-B1 -produced gliadins (especially of the electrophoretic mobility in acid gels of the encoded γ -gliadin) and the presence in the wheat genotype of the Gli-B5b + Rg-1 allelic combination. Six di ﬀ erent alleles at the Gli-B1 locus encoding an identical γ -gliadin produced a PCR sequence of about 400 bp (GliB1.1). Nine Gli-B1d -carrying genotypes from four countries produced an identical sequence of about 409 bp (GliB1.2), while three cultivars with Gli-B1h and four with Gli-B1b produced three and two speciﬁc sequences, respectively, of slightly di ﬀ erent length. Allele Gli-B1j might be the result of recombination between coding and noncoding DNA sequences within the Gli-B1 locus. These observations imply that genetic diversity of the agriculturally important region of chromosome 1B marked by variants of the Gli-B1 locus is rather limited among common wheat cultivars of the 20th century, speciﬁcally to eight principal versions. These might have been incorporated into common wheat from diverged genotypes of diploid donor(s), and, due to the scarcity of recombination, subsequently maintained relatively intact. As well as its evolutionary signiﬁcance, this information is of potential use in wheat breeding and we consider it likely that novel variants of this region will be found in hitherto unstudied germplasm.


Introduction
The gliadins are major wheat endosperm storage proteins [1]; the characterisation of the structure of their encoding loci and of their allelic variation is an important field of study due to (i) their being an integral constituent of gluten, the composition of which is one of the principal determinants of end-use quality; (ii) the presence within some of their amino acid sequences of epitopes that are toxic Names of cultivars, country of origin, and year of registration and spring/winter habit were taken from the internet site wheatpedigree.net.

Alleles at the Gli-B1 Locus and PCR Sequence Length Polymorphism in the Set of Cultivars Studied
Alleles at the Gli-B1 locus were distinguished through the application of current knowledge of the gliadin polypeptide blocks encoded by each allele and observed in acid polyacrylamide gel electrophoregrams (APAGE) of gliadin proteins of single seeds [7].
DNA was extracted from dry single seeds using the CTAB method [18]. PCR amplification was performed in the thermocycler Analytik Jena (Flex Cycler, Germany) using the GliB1.1 and GliB1.2 primers [16]. Products of amplification for each cultivar were fractionated in 7% polyacrylamide gels (280 V) for two hours and stained as proposed by the Silver sequence TM DNA Sequencing System Technical Manual (Promega, Madison, WI, USA; https://www.promega.de/-/media/files/resources/ protocols/technical-manuals/0/silver-sequence-dna-sequencing-system-protocol.pdf). The program GelAnalyzer 19.1 was used for the calculation of the approximate length of the products of amplification by comparing them with the set of standard markers pUc19/Mspl containing fragments of known length. Each fine difference in the length of the PCR sequence was registered in neighbouring lanes of a slab gel and confirmed in different runs.

Allelic Variation
In 46 cultivars (wheat genotypes) studied, 19 different alleles at the Gli-B1 locus (including the allele Gli-B1l represented by the 1BL.1RS translocation) were distinguished (Table 1). Gliadin protein electrophoregrams of some of the wheat genotypes studied, and blocks of gliadin electrophoretic bands controlled by each of 18 alleles (all except Gli-B1l), are shown in Figure 1 and Figure S1.  1 In bold, the cultivars whose gliadin electrophoregrams are shown in the Figure 1. 2 The approximate length (bp) of PCR product in cultivars with different alleles at the Gli-B1 locus. Three or more single seeds for each grain sample giving a definite PCR product were analysed and no intra-varietal non-uniformity was noted for any of the cultivars studied. 3 The name was written on the label. As the genotype of this sample s seeds was different from that of the cultivar with this name studied earlier (Metakovsky et al. 2018), the allele at the Gli-B1 locus was identified in this sample using gliadin electrophoresis (data not shown). 4 For the cultivar Newcaster, only a very weak PCR product was obtained; therefore, its length was not definitively established. Alleles at the Gli-B1 locus differed, in particular, in the electrophoretic mobility (EM) in APAGE of their encoded major γ-gliadin polypeptide. There were six main variants of the EM of this γ-gliadin (for example, from the slowest to the fastest, Figure 1a,d,e,k,m,n). Alleles at the Gli-B1 locus differed, in particular, in the electrophoretic mobility (EM) in APAGE of their encoded major γ-gliadin polypeptide. There were six main variants of the EM of this γ-gliadin (for example, from the slowest to the fastest, Figure 1a,d,e,k,m,n).
Two variants (I and II) of relatively slow EM for this γ-gliadin were observed in several alleles per variant. Thus, variant I (the slowest γ-gliadin) occurred in alleles Gli-B1i (Figure 1a), o (Figure 1b), k, m, p and r; these alleles, with their γ-gliadin of apparently identical EM, differed among them in the composition of the encoded ω-gliadins ( Figure S1). Variant II occurred in alleles Gli-B1j (Figure 1c), s (Figure 1d), n and q.
Alleles of variant III encoded this γ-gliadin of intermediate EM (Figure 1e-j), although the γ-gliadin encoded by the allele Gli-B1f had a slightly higher EM compared with that of other alleles of the group (for example, Figure 1h,i; Figure S1). In contrast, alleles Gli-B1b and Gli-B1c differed considerably from each other in their encoded blocks and from those produced by other alleles of variant group III, Gli-B1e, g, and f ( Figure S1). Therefore, alleles Gli-B1b and Gli-B1c were considered as two more independent variants of the Gli-B1 locus.

PCR Product Length Polymorphism
At least seven principal variants of the PCR product length were revealed when the 46 selected cultivars from around the world were tested in our study, three variants with the GliB1.1 primers ( Figure 2, zone A) and four with the GliB1.2 set ( Figure 3, zone A). The approximate length of these products varied from 369 bp (GliB1.1) to 415 bp (GliB1.2) ( Table 1).
Agronomy 2020, 10, x FOR PEER REVIEW 6 of 14 Two variants (I and II) of relatively slow EM for this γ-gliadin were observed in several alleles per variant. Thus, variant I (the slowest γ-gliadin) occurred in alleles Gli-B1i (Figure 1a), o ( Figure  1b), k, m, p and r; these alleles, with their γ-gliadin of apparently identical EM, differed among them in the composition of the encoded ω-gliadins ( Figure S1). Variant II occurred in alleles Gli-B1j ( Figure  1c), s (Figure 1d), n and q.
Alleles of variant III encoded this γ-gliadin of intermediate EM (Figure 1e,f,g,h,i,j), although the γ-gliadin encoded by the allele Gli-B1f had a slightly higher EM compared with that of other alleles of the group (for example, Figure 1h,i; Figure S1). In contrast, alleles Gli-B1b and Gli-B1c differed considerably from each other in their encoded blocks and from those produced by other alleles of variant group III, Gli-B1e, g, and f ( Figure S1). Therefore, alleles Gli-B1b and Gli-B1c were considered as two more independent variants of the Gli-B1 locus.

PCR Product Length Polymorphism
At least seven principal variants of the PCR product length were revealed when the 46 selected cultivars from around the world were tested in our study, three variants with the GliB1.1 primers ( Figure 2, zone A) and four with the GliB1.2 set ( Figure 3, zone A). The approximate length of these products varied from 369 bp (GliB1.1) to 415 bp (GliB1.2) ( Table 1).    Two variants (I and II) of relatively slow EM for this γ-gliadin were observed in several alleles per variant. Thus, variant I (the slowest γ-gliadin) occurred in alleles Gli-B1i (Figure 1a), o ( Figure  1b), k, m, p and r; these alleles, with their γ-gliadin of apparently identical EM, differed among them in the composition of the encoded ω-gliadins ( Figure S1). Variant II occurred in alleles Gli-B1j ( Figure  1c), s (Figure 1d), n and q.
Alleles of variant III encoded this γ-gliadin of intermediate EM (Figure 1e,f,g,h,i,j), although the γ-gliadin encoded by the allele Gli-B1f had a slightly higher EM compared with that of other alleles of the group (for example, Figure 1h,i; Figure S1). In contrast, alleles Gli-B1b and Gli-B1c differed considerably from each other in their encoded blocks and from those produced by other alleles of variant group III, Gli-B1e, g, and f ( Figure S1). Therefore, alleles Gli-B1b and Gli-B1c were considered as two more independent variants of the Gli-B1 locus.

PCR Product Length Polymorphism
At least seven principal variants of the PCR product length were revealed when the 46 selected cultivars from around the world were tested in our study, three variants with the GliB1.1 primers ( Figure 2, zone A) and four with the GliB1.2 set ( Figure 3, zone A). The approximate length of these products varied from 369 bp (GliB1.1) to 415 bp (GliB1.2) ( Table 1).

The Correspondence between PCR Product Polymorphism and Allelic Variation at the Gli-B1 Locus
The alleles tested in our work fell into two contrasting groups, in accordance with previous observations [17]. One group (alleles Gli-B1b,i,j,k,m,n,o,p,q,r,s) produced a PCR sequence with the GliB1.1 primers (Figure 2), while the other (Gli-B1a,c,d,e,f,g,h) gave a product with the GliB1.2 primers ( Figure 3). Generally, the alleles encoding the slow-moving γ-gliadins (groups I and II) and the allele Gli-B1b produced PCR sequences with the primers GliB1.1, whereas the other alleles did so with the GliB1.2 (Table 1). Hence, variants of the Gli-B1 locus encoding slow-moving γ-gliadins were accessible for the GliB1.1 primers and faster moving γ-gliadins were accessible for the GliB1.2 primers. As expected, the cultivar Cartaya carrying the 1BL.1RS translocation (the Gli-B1l allele) did not give a PCR product with either pair of primers (Table 1).
A strict congruency was discovered among the EM variants of the γ-gliadin and the PCR sequence length variants of a γ-gliadin pseudogene tested. For example, the lowest EM of the Gli-B1 encoded γ-gliadin of variant group I corresponded to the PCR (GliB1.1) sequence of approximately 400 bp. Alleles of group II (except Gli-B1j) generated the PCR (GliB1.1) sequence of 372 bp. Alleles Gli-B1e and Gli-B1g, differing in the presence of one minor ω-gliadin, as well as the allele Gli-B1f encoding a γ-gliadin of slightly higher EM (Figure 1h,i; Figure S1), each produced the PCR (GliB1.2) sequence of 397 bp. The alleles Gli-B1d and Gli-B1a encoding different fast-moving γ-gliadins gave PCR (GliB1.2) products of 409 bp and 415 bp, respectively (Table 1; Figure 3).
The allele Gli-B1b (Figure 1e-g) produced a unique PCR sequence, the shortest observed amongst all alleles studied, of 369 bp (Table 1), and its controlled block differed strongly from those encoded by alleles of group III ( Figure S1). Therefore, we consider the apparent equivalence of the EM of the encoded γ-gliadins of the group III and the Gli-B1b as being coincidental.
The allele Gli-B1c, while differing significantly from the other alleles of group III in its block s composition (the major ω-gliadin was absent and the major γ-gliadin was weaker) (compare Figure 1h-j), gave a PCR (GliB1.2) sequence identical in length to that of the other alleles of the group (Table 1).

The Probable Recombinational Nature of the Allele Gli-B1j
The allele Gli-B1j occurred in only one cultivar, Cluj-650, in more than 1000 common wheat cultivars studied to date (Table S1 in [7]). The EM of the γ-gliadin encoded by this allele was similar to those controlled by the group II alleles, for example, Gli-B1s (Figure 1c,d, respectively). Nevertheless, the PCR sequence produced by this cultivar was of approximately 400 bp in length, which is characteristic of the group I alleles encoding the slowest γ-gliadin ( Figure 2, lanes 8-14; Table 1).
We suggest that the allele Gli-B1j might be the product of a rare intralocus recombination event. As a result, the Gli-B1 locus could have generated this allele by obtaining its γ-gliadin coding DNA (tested with protein electrophoresis) from some allele belonging to group II (any of Gli-B1s, q or n) and its noncoding DNA (tested with the use of the GliB1.1 primers) from some allele of group I. The alternative mutational transition of the EM of the γ-gliadin from group I into the other known group II polypeptides seems much less probable. In addition, the composition (set of gliadin polypeptides) of the Gli-B1j encoded block was similar to that encoded by the group II alleles, not to those of group I ( Figure S1).

Variation in PCR Product Length among Cultivars Having Identical Alleles at the Gli-B1 Locus
As a rule, the same allele occurring in different wheat genotypes produced a PCR sequence of the same apparent length. For example, nine genotypes with the allele Gli-B1d produced PCR sequences (GliB1.2) of approximately 409 bp (Table 1; Figure 3, lanes 6, 7, 8).
However, some genotypes with identical alleles at Gli-B1 might provide PCR sequences differing slightly in length, with the allele Gli-B1h being the most interesting of this kind of observation. The Gli-B1h encoded γ-gliadin has its own unique EM in different cultivars (data not shown), but this allele produced, among three cultivars studied, three PCR (GliB1.2) sequences of slightly different length (approximately, 400-403 bp) that were not, however, produced by any other allelic variant of the locus. The shortest sequence was reproducibly produced by this allele in cultivar Krasnodonka and the longest in cultivar Caia, with the length of the PCR product of the cultivar Ardec probably intermediate between these two variants ( Figure 3, lanes 2, 3, 5; 12, 14, 13; and 20, 22, 21, respectively). Two different grain samples of Krasnodonka were used with identical results (lanes 11, 12).
However, some genotypes with identical alleles at Gli-B1 might provide PCR sequences differing slightly in length, with the allele Gli-B1h being the most interesting of this kind of observation. The Gli-B1h encoded γ-gliadin has its own unique EM in different cultivars (data not shown), but this allele produced, among three cultivars studied, three PCR (GliB1.2) sequences of slightly different length (approximately, 400-403 bp) that were not, however, produced by any other allelic variant of the locus. The shortest sequence was reproducibly produced by this allele in cultivar Krasnodonka and the longest in cultivar Caia, with the length of the PCR product of the cultivar Ardec probably intermediate between these two variants ( Figure 3, lanes 2, 3, 5; 12, 14, 13; and 20, 22, 21, respectively). Two different grain samples of Krasnodonka were used with identical results (lanes 11,12).
The cases of a small difference of the length of the PCR product among cultivars with the same allele help to confirm the congruity between the EM of the encoded γ-gliadin and the particular length of the PCR product.

Polymorphism of the PCR Products in "Zone B" of the Electrophoregram
In addition to the polymorphism detected in "zone A" of the DNA electrophoregram, variation was observed among PCR-derived DNA fragments of higher length ("zone B"). As a rule, the DNA sequence in this zone was about 100 bp longer compared to zone A in the same genotype. For example, the approximate length of the PCR product in zone B was about 497 bp and 495 bp for the alleles of groups I and III, respectively. Although the nature of the PCR products in zone B remains to be studied, the polymorphism in this zone is reasonably reproducible and may contribute to the polymorphism of the noncoding sequences of the Gli-B1 locus.
For example, this polymorphism reinforced the difference between some wheat genotypes carrying different alleles at the Gli-B1 locus, but with PCR products (in zone A) of apparently identical length. The most striking difference in zone B of two DNA electrophoregrams identical in zone A A final example was given by the allele Gli-B1c, which, when present in the cultivar Diego ( Figure 3, lane 17), might generate a longer PCR product compared with the other cultivars carrying this allele, Prinqual and Siete-Cerros-66 ( Figure 3, lanes 18, 19).
The cases of a small difference of the length of the PCR product among cultivars with the same allele help to confirm the congruity between the EM of the encoded γ-gliadin and the particular length of the PCR product.

Polymorphism of the PCR Products in "Zone B" of the Electrophoregram
In addition to the polymorphism detected in "zone A" of the DNA electrophoregram, variation was observed among PCR-derived DNA fragments of higher length ("zone B"). As a rule, the DNA sequence in this zone was about 100 bp longer compared to zone A in the same genotype. For example, the approximate length of the PCR product in zone B was about 497 bp and 495 bp for the alleles of groups I and III, respectively. Although the nature of the PCR products in zone B remains to be studied, the polymorphism in this zone is reasonably reproducible and may contribute to the polymorphism of the noncoding sequences of the Gli-B1 locus.

The PCR Sequence Corresponds to Genotype, Not to the Label of the Grain Sample under Study
Four grain samples studied contained some seeds representing other genotypes than those identified previously in these cultivars. For example, the seeds of the grain sample labelled "Kavkaz" gave a PCR product (of length 369 bp), in spite of the fact that the Russian cultivar Kavkaz is one of the best known carriers and donors of the 1BL.1RS translocation and, therefore, lacking a true Gli-B1 locus from its genotype ( Table 1). Gliadin analysis (APAGE) of seeds of this sample showed that its genotype included allele Gli-B1b (data not shown), and therefore its labelling as "Kavkaz" was incorrect.
The genotype of only one of the two grain samples of the cultivar Argelato studied included the allele Gli-B1g characteristic of this cultivar. There were two more cases of discrepancy between genotypes of grains studied earlier [7] and in this work, Inia-66 and Libero. Thus, it is remarkable that the PCR product always corresponded to genotypes identified by APAGE in this work, not to the name of the cultivar written on the label ( Table 1). The presence of admixtures of alien genotypes in grain samples of wheat and errors in labelling was noted earlier in different wheat collections [19][20][21].
One seed of the Canadian cultivar Katepwa successfully produced PCR sequences with both pairs of primers, 409 bp (GliB1.2) and 372 bp (GliB1.1). We suggest that this seed was heterozygous at the Gli-B1 locus for allele Gli-B1d (produces the sequence of the 409 bp) characteristic for this cultivar (Table 1) and some allele of group II (for example, Gli-B1q, present in genotypes of some Canadian cultivars). It was thereby implied that this genotype produced PCR sequences from both homologous chromosomes. Heterozygous seeds originating from undetected cross-pollination may occur in wheat grain samples of different origin and are considered as foreign seeds or admixtures [21].

Discussion
All the expressed genes located at a particular Gli locus are known to inherit virtually as a Mendelian unit, almost without recombination [6]. The rate of recombination between two of the most remote gliadin-coding genes harboured in a Gli-1 locus does not exceed 1% [22]. These data imply tight clustering of the expressed gliadin genes at this locus.
However, analysis of large DNA sequences of the Gli loci showed that the distance of 10 Kb between two gliadin genes, around ten times more than the length of a gliadin gene [4,13] itself, may be considered as unusually short and that the genes can be considered to be clustered [23]. In addition to noncoding gliadin sequences (pseudogenes, gene fragments) [14], transposable elements occupy most of the noncoding DNA between seed storage protein genes and occur in abundance at the distal end of the short arm of chromosome 1B [12,[23][24][25].
The reason for the ability of several expressed gliadin genes, in spite of being relatively remote within the Gli locus, to be transmitted through intervarietal crosses as a unit with little visible signs of recombination, remains unclear. In order to understand the organization and function of the entire Gli locus, study and use of the polymorphism of its noncoding DNA sequences is required.
Two pairs of primers, GliB1.1 and GliB1.2, were developed for the analysis of the γ-gliadin pseudogene located within the Gli-B1 locus [16]. Nineteen known allelic variants of the Gli-B1 locus (revealed using APAGE of gliadin) were tested in our work using these pairs of primers, and a strict correspondence was discovered between variants of the EM of the γ-gliadin encoded by different alleles at the Gli-B1 locus and variants of the length of the PCR-produced sequence of the γ-gliadin pseudogene tested. At least seven principal variants of the Gli-B1 locus were distinguished. These variants are represented by four single alleles, Gli-B1a, Gli-B1b, Gli-B1d, Gli-B1h, and three allelic groups I, II, and III (without Gli-B1b). The seven variants differed in both their coding sequences (firstly, the EM of the major γ-gliadin, and, secondly, the composition of the entire block) and noncoding sequences (the length of the PCR-produced γ-gliadin pseudogene sequence).
The difference between variants at the distal end of chromosome 1B marked by different Gli-B1 alleles may be reinforced by two more genetic loci situated distally from, but closely linked to [26], Gli-B1, namely, Gli-B5 (where allele Gli-B5b encodes two faint ω-gliadins) and Rg-1, responsible for red glume colour. Analysis of a collection of common wheat varieties showed that the presence, in a certain wheat genotype, of any allele of groups I and II was nearly always accompanied by alleles Gli-B5b (Figure 1 and Figure S1) and Rg-1, while other alleles at the Gli-B1 locus occurred in genotypes together with the alleles Gli-B5a (the absence of the two faint bands) and rg-1 (white glume colour) [27]. This pattern was confirmed in Italian [28] and Spanish [29] wheat cultivars, although in each germplasm collection studied, single recombinant genotypes might occur. Hence, alleles of groups I and II, and the alleles Gli-B1c and Gli-B1h, occurred in wheat genotypes mainly with the allelic combination Gli-B5b + Rg-1, while the presence of any other allele was accompanied by the combination Gli-B5a + rg-1. In the set of cultivars studied in the current work, the cultivar Krasnodonka ( Figure 1k) had a recombinant genotype: the "red-colour allele" at Gli-B1 together with alleles Gli-B5a and rg-1.
The allele Gli-B1c (Figure 1j) produced the corresponding PCR sequence similar in length to that of the group III alleles at the Gli-B1 locus (Table 1). However, this allele marks a recombined variant of the distal part of the chromosome 1B: it encodes the set of gliadins (block), especially of the ω-gliadins, different from that of the group III ( Figure S1), and differs from other alleles of this group in being linked to the "red-colour" allelic combination, Gli-B5b + Rg-1. Therefore, the allele Gli-B1c may be considered as an eighth variant relatively independent from the group III alleles. In general, each homologous variant of chromosome 1B in common wheat germplasm is marked by one of the eight distinct variants of the Gli-B1 locus.
Diploid donors of the genomes A, B and D might already have been polymorphic at the time of the origin of the tetraploid AABB and hexaploid AABBDD wheats, due to the differentiation of their genotypes at the diploid level [30]. We suggest that variants of the distal part of different homologous 1B chromosomes marked by each one of eight different variants of the Gli-B1 locus were introduced into common wheat by considerably diverged genotypes of the donor(s) of the B genome. Due to the scarcity of recombination, each of the homologous variants of this chromosome s region may maintain intact (preserved from mixing) the precise gene (allelic) combination that came into common wheat from a particular genotype of the diploid donor. It is important to note that there are many agriculturally valuable genes in the distal part of the chromosome arms 1S [8,11].
The differences between homologous variants of common wheat chromosomes, especially of the B genome, were registered using cytological approaches [31,32]. It was assumed that these differences might cause a reduction in the number of chiasmata and chromosome pairing in intervarietal F1 hybrids compared to that of their parental inbred lines [31,33,34]. For the short arm of chromosome 1B (with the Gli-B1 locus in its distal region) this reduction, in some crosses, might reach 40% [35].
The degree of differences between homologous chromosomes may influence the recombination rate observed in intervarietal crosses. For example, the genetic distance between Gli-A3 and Gli-A1 loci varied among intervarietal crosses from 12.6% to 35.3% [36] or from 7 cM to 27 cM [37]. The distance between the Nor-B2 and Gli-B2 loci was about 2.0 cM in a cross between the varieties Chinese Spring and Timstein, but 10.2 cM for Chinese Spring crossed to the variety Cheyenne [38]. Hence, the differences between the DNA sequences in Chinese Spring and Timstein might be greater than those between Chinese Spring and Cheyenne. The recombination frequency between Gli-B1 and Rg-1 loci varied from 0.0% (two different crosses) to 2.7% and 4.8% in two further crosses [26]. In each of these four crosses, one parent was the cultivar Salmone (Gli-B1s), and the frequency of recombination in each case might have depended upon the variant of the Gli-B1 locus present in the second parent (alleles Gli-B1m, e, b, b, respectively).
For chromosome 3B, 82% of crossovers occurred in the distal region of chromosome 3B, representing only 19% of the chromosomal length [39]. Its "cold" areas (of low recombination rate) had a peculiar composition of transposable elements: a higher level of retrotransposons compared to DNA transposons [40]. The reduction in recombination rate in regions with a high density of retrotransposons could be due to DNA sequence variation created by their activity [41]. The negative impact of retroelements on recombination could also be attributed to their ability to condense DNA and lock the regions, preventing their accessibility to the recombination machinery [40]. The "cold" areas carry fewer and less expressed genes [40] and more single nucleotide substitutions (SNP) [10].
It is not clear if Gli-B1 is analogous to the case of a genetic locus situated inside the "cold" region of chromosome arm 3BS. Nevertheless, eight variants of alleles at the Gli-B1 locus may serve as markers of the distal part of homologous, but diverged, 1B chromosomes.
The primary structure of the γ-gliadin polypeptides includes, besides other sequences, variable numbers of CAA-repeats and a small microsatellite encoding poly-glutamine [5]. The number of CAA-repeats in different γ-gliadin genes varied, in 170 wheat genotypes studied, from seven to 22 [13]. The visible differences in PCR sequence lengths produced by the γ-gliadin pseudogene GAG56B in several common wheat cultivars were caused by variation in the number of CAA-repeats in this sequence [15]. Our preliminary results showed that common wheat genotypes with alleles Gli-B1a and Gli-B1b produced, respectively, the longest (31 repeats) and the shortest (seven repeats) PCR sequence of GAG56B (L. Pascual, unpublished). There is no doubt that polymorphism in the length of the γ-gliadin pseudogene tested in our work with the GliB1.1 and GliB1.2 pairs of primers was mainly due to the number of CAA-repeats in its PCR sequence.
Microsatellite sequences might frequently change their length, with the mutation rate up to 2.4 × 10 −3 [42], or up to 5 × 10 −3 [43] per locus, per generation. In a series of microsatellite alleles, they differed from one another, as a rule, by two or three nucleotide pairs [44]. Hence, the difference in length of the microsatellite may cause the absence of exact identity between the PCR sequences produced by different cultivars having the same allele at the Gli-B1 locus (Gli-B1b, Gli-B1h or Gli-B1c) documented in our work.
The velocity of DNA divergence in noncoding regions was shown to be at least 3 to 4 times higher than in transcribed genes [45,46]. In the case of the Gli-B1 locus, however, the variation in the length of the γ-gliadin pseudogene PCR sequence correlated strictly with the polymorphism of the EM of the encoded γ-gliadin, which merits further study.

Conclusions
In this work, we compared nearly all allelic variants of the Gli-B1 locus revealed earlier in our study of gliadin genotypes in about 900 common wheat registered cultivars bred in 20th century. Analysis of allelic variation at the Gli-B1, Gli-B5, Rg-1 loci and at the γ-gliadin pseudogene located within the Gli-B1 locus permitted the identification of eight well-distinguished versions of the distal part of chromosome 1B in the germplasm studied. A correspondence was discovered between the length of the PCR sequences produced by different variants of the pseudogene, the variants present of the Gli-B1-controlled sets of gliadins (blocks), especially of the electrophoretic mobility of the encoded γ-gliadin, and the presence of the Gli-B5b + Rg-1 allelic combination present in a certain wheat genotype. We suggest that distinct versions of the distal part of different homologous 1B chromosomes (each of them marked by one of eight variants of the Gli-B1 locus), known to carry many agriculturally-important genes, were introduced into common wheat by diverged genotypes of the donor(s) of the B genome. Due to the scarcity of recombination, each of the homologous variants of this chromosome s region might maintain intact (preserving from mixing) the precise gene (allelic) combination that was incorporated into common wheat from a particular genotype of the diploid donor. Therefore, genetic polymorphism in this region of chromosome 1B in registered cultivars bred in the 20th century might be essentially limited to these eight principal variants.
As well as their evolutionary significance, these results are of potential use in wheat breeding in providing a simple means to select, via markers, for certain allelic combinations for this region across the generations of a genetic improvement programme. Furthermore, the results provide the knowledge base with which to compare the variation potentially present in hitherto unstudied collections of wheat germplasm, such as cultivars from other countries, landraces, and accessions of species related to common wheat, where we consider new and agronomically useful variants are likely to be found.