Breeding and genetic bases of winter wheat ear traits

Abstract

Wheat breeding is one of the key areas of agricultural research aimed at ensuring sustainable food production in the face of global climate change and the growing world population. Winter wheat holds a significant place in global cereal production due to its high yield and grain quality characteristics. One of the critical aspects affecting the productivity and adaptability of winter wheat is the characteristics of the spike and yield. Research focused on understanding the breeding and genetic basis of spike traits in winter wheat allows the identification of genetic mechanisms governing these traits and facilitates the development of new varieties with improved characteristics. In recent years, significant progress in this field has been achieved through the application of modern molecular genetic technologies, such as marker-assisted selection (MAS) and high-throughput sequencing. MAS significantly accelerates and improves the accuracy of the breeding process by utilizing molecular markers associated with desirable traits. Its application in winter wheat breeding allows for the effective selection of plants with optimal spike traits, ultimately leading to the creation of high-yielding and stress-resistant varieties. High-throughput sequencing («next-generation sequencing») enables the rapid and accurate sequencing of large amounts of DNA and RNA, the study of plant genomes, and the identification of genes and molecular markers associated with important agronomic traits of the wheat spike. This significantly accelerates the breeding process through precise and large-scale genetic data analysis. This study systematically analyzes the genetic and breeding basis of spike traits in winter wheat, such as the number of grains and the weight of 1000 grains, using marker-assisted selection (MAS) and high-throughput sequencing to analyze genetic variability and identify molecular markers. Special attention is given to the role of microRNAs in regulating grain development and degradome sequencing (as one of the methods of high-throughput sequencing) for identifying genes targeted by microRNAs and annotating their functions. For the research consisted of the winter wheat varieties Mexican Large Spike (MLS), Bainong 4199 (BN419), and an F2 population (145 plants) created by hybridizing Mexican Large Spike × Bainong 4199. The research results showed that the parent forms significantly differed in spike traits – the number of grains per spike and the weight of 1000 grains, which corresponds to the principle of parental selection when creating a population for QTL mapping. The traits exhibited a normal distribution and bidirectional segregation, indicating quantitative inheritance. According to the results of the correlation analysis, the number of spikes and the weight of 1000 grains had a highly significant positive correlation with a correlation coefficient of 0.953. A total of 143 pairs of polymorphic markers with distinct differences were tested for polymorphism using 300 pairs of SSR primers. The 143 pairs of polymorphic markers were genotyped, and genetic maps were constructed for the 145 individual plants of the F2 population. To construct the linkage map, 145 loci of polymorphic SSR markers were used, covering 19 wheat chromosomes, with a total length of 3128.17 cM. The average distance between markers was 25.23 cM, the maximum distance was 113.85 cM, and the minimum distance was 3.57 cM. The genetic density between some markers exceeded 50 cM, which was mainly due to the low density of molecular markers. The distribution of 145 polymorphic marker loci across chromosome groups A, B, and D was uneven. There were 77 marker loci present in chromosome group A, 41 in group B, and only 24 in group D, accounting for 54.22%, 28.87%, and 16.9% of the total number of marker loci, respectively. The SSR markers exhibited the highest polymorphism in chromosome group B and the lowest polymorphism in chromosome group D. Determination and analysis of epistatic QTL loci for number of grains per ear and thousand grain weight showed that nine epistatic QTL loci associated with number of grains per ear and thousand grain weight were identified, which were distributed on chromosomes 1B, 2B, 2D, 3B and 6B, including four associated QTL loci on chromosome 3B, two associated QTL loci on chromosome 6B. and one associated locus each on chromosomes 1B, 2B and 2D. This allows explaining 4.922%~21.1044% of phenotypic variation for the number of grains per ear. QGNS~1B and QGNS~3B2 had large genetic effects and were the main loci explaining 21.1044% and 15.8886% of phenotypic variation. One additive QTL locus controlling thousand grain weight was found at QTGW~3B, which explained 11.4727% of the phenotypic variation. The effect values for QGNS~2B, QGNS~2D, and QGNS~3B1 were less than zero, so the genetic effect of epistatic QTLs belonging to the recombinant type was greater than that of epistatic QTLs belonging to the parental type, which explained 41.91% of the total phenotypic variation. Three epistatic QTL loci associated with thousand grain weight were found, mainly on chromosomes 3B and 5A, among which the genetic effect values of QTGW~3B1 and QTGW~3B2 loci were greater than zero and belonged to the paternal type of epistatic QTL loci, and the values of the genetic effect of QTGW~5A loci were less than zero, which belonged to the recombinant type of epistatic QTL loci and accounted for 17.4% of the total variation in the phenotype. The effect values at the QGNS~3B2, QGNS~6B1 and QGNS~6B2 loci were greater than zero for the parental type epistatic QTLs, indicating that the parental type epistatic loci had a higher effect than the recombinant type epistatic loci and accounted for 37.23% of the total phenotypic variation. The effect of epistasis on phenotypic variations of grain size and thousand grain weight was significant in relation to the effect on phenotypic variability of the number of grains in the ear. Thus, the factors that influence the results of QTL localisation are the choice and size of the population, the type and number of molecular markers, environmental conditions, statistical methods and the density of markers for genetic linkage mapping. The above factors can be considered in combination for more accurate identification of QTL loci associated with target traits The results showed that compared to 14 DAP, 7DAP has a broader mRNA-mediated regulation of gene expression. Moreover, more novel microRNAs were expressed in 14 DAP than in 7 DAP. 31 new microRNAs were expressed in two libraries: 1 specific one in the 7 DAP library and 30 specific ones in the 14 DAP library. Most of the new microRNAs had relatively low expression levels. The most common novel microRNAs were novel-m0064_5p, novel-m0492_5p, and novel-m0661_5p. Non-coding RNAs were identified, and after removing annotated RNAs, the number of unannotated sequences was 2,448,113 (17.46%) in the 7th DAP library and 3,456,209 (33.3%) in the 14 DAP pure read libraries. Of these unannotated sequences, 235.8 79 (10.64%) are reads in 7 DAP libraries and 1 458 680 (43.06%) are reads in 14 DAP libraries, which perfectly corresponds to the shotgun whole-genome sequences (WGS). The microRNA sequencing samples were used to generate degradome libraries, which were created by ligating polyA-enriched RNAs with a designated RNA adapter containing the 3'-site of Mme I. The degradome library was created and then subjected to degradome sequencing. A total of 10,620,205 net reads were obtained, including 4,612,965 unique reads. 2,776,485 (60.19%) unique reads were aligned to the reference genome. Using BLASTN searches in GenBank and Rfam databases, structural RNAs (rRNA, tRNA, miRNA and mRNA) were removed and the remaining reads were used to identify potential microRNA targets. Total RNA lengths ranged from 18 to 30 nt in both libraries, with the majority of them being between 21 and 24 nt, with the most common lengths being 21 nt in the 7 DAP library and 24 nt in the 14 DAP library. The length distributions of pure strand and microRNAs were different in the two libraries: the distribution of mRNA lengths was 61% and 73.89% between 20~24 nt, respectively, and the percentage of mRNAs with 24 nt sequence (7 DAP was 26.79%, 14 DAP was 40.42%) was higher than that of other sequences. In the library, 7 DAP mRNA with 23 nt was second only to the 24 nt mRNA sequence, and in the library, 14 DAP mRNA with 21 nt. The comparison of unannotated RNA sequences that perfectly matched the reference genome in miRBase22.0 based on the perfect match criterion resulted in 89 known wheat microRNAs in the two RNA libraries, of which 46 were expressed in the 7 DAP library and 87 in the 14 DAP library. In total, 16 known microRNAs from other plant species were identified in the available two RNA libraries, including 13 expressed in 7 DAP libraries and 16 expressed in 14 DAP libraries. The base distribution at each position of known small RNAs was calculated, and the results showed that the first base at the 5'-end of microRNAs is most commonly A (adenine) and least commonly G (guanine). This is inconsistent with the existing confirmed microRNA-specific sequences, where the first base at the 5'-end is most commonly U (uracil) and least commonly G (guanine). It was shown that the expression of known microRNAs, including ata-miR9863a-3p, osa-miR396e-5p, tae-miR9670-3 and tae-miR7757-5p, was the most active, whereas some other known microRNAs such as stae-miR9672b, tae-miR9662a-3p, tae-miR167c-5p ta, e-miR156, tae-miR9777 and tae-miR9669-5p were moderately upregulated. Most of the widely expressed microRNAs are known microRNAs from the wheat microRNA database and are conserved in plant species, such as miR156 and miR396e-5p. For microRNA identification, 19 known and 20 novel microRNAs were differentially expressed (r < 0.05) between 7 and 14 DAP grains, among which 18 known microRNAs and 13 novel microRNAs were up-regulated in 14 DAP grains. Potential microRNA targets were predicted computationally using software and, starting from 184 total microRNA sequences, a set of 810 potential targets for 25 novel microRNAs, 13 conserved microRNAs and 40 known microRNAs were predicted. Among 88 DEmiRNAs, 477 potential targets for 147 miRNAs were identified. Among all 810 miRNA targets, 174 (21.48%) could be annotated in the COG database, 605 (74.69%) in the GO database, and 176 (21.73%) in the KEGG database. However, out of 477 DE miRNA targets, only 84 were functionally annotated by COG. The results showed that the 23 predicted microRNA targets include 3 known and 4 novel microRNAs. In addition, most single microRNAs potentially regulate multiple targets, while some single microRNAs act on only one target gene. Of the 16 targets, 13 have functional annotations. Tae-miR160 targets five genes, including Traes_1AL_147CF243C, Traes_1BL_54CD82AC3, Traes_7AL_E3ADC8C38, Traes_7BL_18D335F08 and Traes_7DL_5. 5ADB3528; tae-miR1119 has three targets, including Traes_6BS_04A3400AF, Traes_6BS_222CE7DA and Traes_6BS_4D03398A8; Novel-miR0012 targets Traes_3DS_2F5F2C276, Traes_3B_8824DBB56 and Traes_3AS_7EEF1386F, and Novel-miR0011 targets Traes_3DS_C6D17D438 and Traes_3B_E7D2E8720. Novel-miR0036, novel-miR0075 and tae -miR1 56 target Traes_7AS_2084DE83B, Traes_5AL_147EA9565 and Traes_6BS_542961EA4, respectively. Some of the targets were found to be involved in various biological processes, such as mitotic cell cycle (GO:0000278), nucleation (GO:0000280), cell morphogenesis (GO:0000902), seed development (GO:0048316), embryo development (GO:0009790) (4 targets) For 2 microRNAs), meristem initiation (GO:0010014), leaf development (GO:0048440), cell differentiation (GO:0030154), cell division (GO:0051301), starch metabolism (GO:0005982) and regulation of cell division (GO:0051302). This confirms that microRNAs play an important role in the regulation of grain development.