Transgenerational Paternal Inheritance of TaCKX GFMs Expression Patterns Indicate a Way to Select Wheat Lines with Better Parameters for Yield-Related Traits

Abstract

Members of the TaCKX gene family (GFMs) encode the cytokinin oxygenase/dehydrogenase enzyme (CKX), which irreversibly degrades cytokinins in the organs of wheat plants; therefore, these genes perform a key role in the regulation of yield-related traits. The purpose of the investigation was to determine how expression patterns of these genes, together with the transcription factor-encoding gene TaNAC2-5A, and yield-related traits are inherited to apply this knowledge to speed up breeding processes. The traits were tested in 7 days after pollination (DAP) spikes and seedling roots of maternal and paternal parents and their F₂ progeny. The expression levels of most of them and the yield were inherited in F₂ from the paternal parent. Some pairs or groups of genes cooperated, and some showed opposite functions. Models of up- or down-regulation of TaCKX GFMs and TaNAC2-5A in low-yielding maternal plants crossed with higher-yielding paternal plants and their high-yielding F₂ progeny reproduced gene expression and yield of the paternal parent. The correlation coefficients between TaCKX GFMs, TaNAC2-5A, and yield-related traits in high-yielding F₂ progeny indicated which of these genes were specifically correlated with individual yield-related traits. The most common was expressed in 7 DAP spikes TaCKX2.1, which positively correlated with grain number, grain yield, spike number, and spike length, and seedling root mass. The expression levels of TaCKX1 or TaNAC2-5A in the seedling roots were negatively correlated with these traits. In contrast, the thousand grain weight (TGW) was negatively regulated by TaCKX2.2.2, TaCKX2.1, and TaCKX10 in 7 DAP spikes but positively correlated with TaCKX10 and TaNAC2-5A in seedling roots. Transmission of TaCKX GFMs and TaNAC2-5A expression patterns and yield-related traits from parents to the F₂ generation indicate their paternal imprinting. These newly shown data of nonmendelian epigenetic inheritance shed new light on crossing strategies to obtain a high-yielding F₂ generation.

1. Introduction

Bread wheat (Triticum aestivum) is the most important cereal crop in the temperate climate and provides a staple food for more than a third of the world’s population [1]. It belongs to the Triticeae tribe, which also includes barley, rye, and triticale. Among the species, wheat has the largest and most complex hexaploid genome (2n = 6x = 42), which consists of the three homoeologous subgenomes A, B, and D. Each gene present as homologues A, B, and D could retain its original function or, as a result of independent evolution, develop heterogeneous expression, and/or one or two copies may be silenced or deleted [2,3].

Cytokinins (CKs) perform a basic role in the growth, development, and productivity of any plant species, including wheat [4]. Their content in developing spikes of wheat is correlated with grain yield, grain number and weight, TGW, chlorophyll content in flag leaves, and seedling root weight [5,6,7,8,9]. There are two types of cytokinins, isoprenoid and aromatic. The most important and widely occurring are isoprenoids, cis-zeatin (cZ), trans-zeatin (tZ), isopentenyl adenine (iP), and dihydrozeatin (DZ), and fewer aromatic forms, e.g., benzylaminopurine (BA). The content of active forms in plant tissues and organs depends on metabolic processes, such as biosynthesis, degradation, inactivation, and reactivation. Active forms and their ribosides can also be transported throughout the plant [10]. Enzymes for all metabolic processes are encoded by members of the gene family (GFMs) [11,12,13,14,15]. One of the most important processes is the irreversible degradation of cytokinins by TaCKX GFMs. There are 13 basic TaCKX GFMs, 11 of which have homoeologs in subgenomes A, B, and D. Two of them, TaCKX2.2.2 and TaCKX2.2.3, are located only in the D subgenome [11]. The CKX GFMs encode the cytokinin oxidase/dehydrogenase enzyme. Their role in the regulation of yield-related traits in wheat [5,6,7,11] and other species were already shown [4,16,17,18]. Decreased expression of the TaCKX1 or TaCKX2 genes significantly influenced TGW, seed number, and chlorophyl content in flag leaves, and the effect was dependent on the silent gene and/or genotype [5,6,7]. The TaCKX4 copy number affected grain yield and chlorophyl content in flag leaves [19]. Haplotype variants of TaCKX6-D1 (actually TaCKX2.2.1-3D) were associated with TGW [20], and the allelic variant of TaCKX6a02 (annotated as TACKX2.1) influenced grain size, filling rate, and weight [21]. In addition to yield, some CKX GFMs influence other pleiotropic traits, including root growth, nutrient accumulation, and abiotic stress responses [18,22,23]. These genes could also be regulated at the transcriptional level by transcription factors (TFs), especially those belonging to the NAC family; however, knowledge about their function is still very limited [24]. A promising NAC-encoding candidate with a role in yield-related traits is TaNAC2-5A [25,26]. As reported, overexpression of the gene delayed leaf senescence and increased nitrate uptake and concentration, root growth, and grain yield under field conditions. It is interesting to note that in a controlled environment, TaNAC2-5A was negatively correlated with the activity of the CKX enzyme in seedling roots and the number of tillers [8].

Crossbreeding and selection are basic steps in crop improvement, and the only potential limitation is too narrow genetic variability. Therefore, it is important to know how yield-related traits are inherited. Moreover, traits, including stably integrated transgenes and edited genes, are inherited according to Mendelian rules [27,28,29,30]. An exception to Mendel’s principles that encompass both groups of genes is epigenetic inheritance [31] or the polygenic nature of genomic architecture for the linked traits, which can be regulated by transcription factors [32,33,34] or other gene regulatory networks [35].

The pattern of gene expression could be considered a trait with its own type of inheritance.

This concept, reviewed by Yoo et al. [2], called parental expression additivity, is defined as the arithmetic average of the expression of the parental genes. Expression additivity of parental genes is observed in the offspring of the diploid species. The deviation of additivity called parental non-additive expression is mainly found in the offspring of polyploid species. A bias when the expression of the offspring is similar to that of one of the parents is called expression-level dominance. If total offspring expression is lower or higher than in both parents, the phenomenon is called transgressive expression, and when the contribution of the parental homeologs to the total gene expression is unequal, it is named homeolog expression bias. All this deviation from additivity can be explained as a result of different factors, such as the influence of one of the parental genomes, epigenetic regulation, balance of gene dosage, and cis- and/or trans-regulatory elements [2].

Expression-level dominance, which is of uniparental origin, is also called genomic imprinting [36,37]. This phenomenon of epigenetic origin is the result of the asymmetries of DNA and histone methylation between maternal and paternal plants. Male and female genotypes are multicellular in origin; therefore, primary gene imprinting can occur in egg cells, central cells, and sperm and, subsequently, in the triploid endosperm or, less frequently, in the diploid embryo [37]. Generally, conservation of imprinting is limited across other crops; however, these genes that show conserved imprinting in cereals showed positive selection and were suggested to perform a dose-dependent function in the regulation of seed development [38]. However, as recently documented by Rodrigues et al. [39], most genes imprinted in the endosperm of seeds were imprinted across cultivars, extending their functions to chromatin and transcriptional regulation, development, and signaling. Only 4% to 11% of the imprinted genes showed divergent imprinting.

Imprinted gene expression affects mostly single genes or groups of genes. Most of them are maternally expressed and inherited [36,40,41]. The best recognized is the maternal effect of genes during embryo development [42]. Early development in Arabidopsis is coordinated by the supply of auxin from the mother integuments of the ovule, which is required for the correct embryo development of embryos [40]. The genomic imprinting of the cereal endosperm influences the timing of endosperm cellularization [43]. An example of imprinted maternally expressed genes in cereals is a polycomb group, which is important for the cellularization of endosperm in rice [44]. Reciprocal crosses between tetraploid and hexaploid wheats showed that imprinted genes were identified in endosperm and embryo tissue, supporting the predominant maternal effect on early grain development [45]. Paternally expressed imprinted genes were associated with hybrid seed lethality in Capsella [46]. In maize, the Dosage-effect defective1 (ded1) locus that contributes to seed size was found to be paternally imprinted [47]. The gene encodes a transcription factor that is specifically expressed during early endosperm development. There is also evidence that small RNAs might determine the paternal methylome by silencing transposons [48]. In addition to these reports, it is very difficult to find examples of paternally inherited genes, especially in cereals. Many studies have indicated dynamic changes in the epigenetic state, including DNA methylation, chromatin modifications, and small RNAs, which are observed during the reproductive development of plants [49,50]. The spatiotemporal pattern of gene expression, imprinting, and seed development in Arabidopsis endosperm is predominantly regulated by small maternal RNAs; however, they also originate from the paternal genome and the seed coat [51]. As reported by Tuteja et al. [52], imprinted paternally expressed genes, but not maternally expressed genes, in Arabidopsis evolve under positive Darwinian selection. These genes were involved in seed development processes, such as auxin biosynthesis and epigenetic regulation. Imprinted paternally expressed genes are mainly associated with hypomethylated maternal DNA alleles, which can be repressed by small genic RNAs and rarer with transposable elements [49,53]. Epigenetic changes can be developmentally regulated (developmental epigenetics). The state in which changes in DNA methylation are stable between generations and heritable is called transgenerational epigenetics [54].

Several TaCKX GFMs and TaNAC2-5A (NAC2) were previously selected as important regulators of yield-related traits. To determine how the expression patterns of selected genes are inherited in the developing spikes and seedling roots of the parents and the F₂ generation, we used a reciprocal crossing strategy. The research hypothesis assumed that knowledge of inheritance of gene expression patterns that regulated yield-related traits indicated the way of selection of genotypes in wheat breeding. There is a research gap in documenting the inheritance of expression patterns for yield-related genes. We found that most of the genes in the F₂ generation were expressed in a pater-of-origin-specific manner, which shed new light on the ways of selecting wheat lines and the breeding strategy.

2. Results

2.1. Reciprocal Crosses Indicate That the Expression Patterns of Most of the TaCKX GFM and Yield-Related Traits Are Mainly Inherited from the Male Parent

Relative values (related to the female parent = 1.0) of the expression profiles of TaCKX GFM and NAC2 in 7 DAP spikes, seedling roots, and phenotypic traits in the female parent, male parent, and their six F₂ progeny from one reciprocal cross of S12B × S6C (C1) and S6C × S12B (C2) are presented in Figure 1. The same data obtained in reciprocal crosses of D16 × KOH7 (C3) and KOH7 × D16 (C4); D19 × D16 (C5) and D16 × D19 (C6); D19 × KOH7 (C7) and KOH7 × D19 (C8) are visualized in Figure S1. S6C, the paternal parent of the S12B × S6C cross (C1), showed higher expression of TaCKX1 and NAC2 and lower expression of TaCKX5 and TaCKX10 in spikes than the maternal parent (S12B). The expression of TaCKX1, NAC2, TaCKX5, and TaCKX11 in the spikes of the F₂ progeny of this cross was higher (Figure 1A). In the reverse cross (S6C × S12B), when S12B was a paternal component (Figure 1B), TaCKX1 and NAC2 were expressed at low levels, TaCKX5 and 10 were highly expressed in spikes compared to the maternal parent (S6C), and in F_2, TaCKX1 and NAC2 were expressed at low levels, and TaCKX9, 10, 11, and 5 were upregulated. In seedling roots, the paternal component of S12B × S6C (Figure 1A) showed high expression of TaCKX5 and NAC2 and low expression of TaCKX10 and 11 in the parental parent compared to the maternal parent. In the roots of the F₂ progeny of this cross, TaCKX5 and NAC2 were highly expressed, and TaCKX11 was downregulated. The expression data in the parents of the reverse cross, in which S12B was the paternal component, were opposite. Their F₂ progeny showed a strong upregulation of TaCKX8, 10, and 11 and a downregulation of TaCKX1, TaCKX3, and NAC2 in seedling roots. The total grain yield and the number of seeds in F₂ of S12B × S6C were low, similar to the male parent (Figure 1A). The root weight in the F₂ progeny of the same cross (S12B × S6C) was lower than that in the parents. The same yield components in the opposite cross (S6C × S12B) were high in one F₂ sibling, comparable to the maternal parent in two progeny and lower than in the parents in three of them (Figure 1B). Interestingly, the root mass in F₂ was higher than that in both parents.

As presented in

Table 1. TaCKX GFMs and NAC2 with high (↑), very high (↑↑), low (↓) or very low (↓↓) expression levels in 7 DAP spikes and seedling roots of the maternal parent (M), the paternal parent (P) and their F₂ progeny from cresses of D16 × KOH7 (C3) and KOH7 × D16 (C4); D19 × D16 (C5) and D16 × D19 (C6); D19 × KOH7 (C7) and KOH7 × D19 (C8), and high (↑) and low (↓) parameters of yield and root mass. Character colours indicate similar patterns of gene expression and yield-related traits in F₂ and paternal parent (red) or in F₂ and maternal parent (green).

	M	P	F2
	C1 = S12B × S6C
CKX expression 7 DAP	CKX1↓	CKX1↑	CKX1, 11↑↑
	NAC2↓	NAC2↑	CKX5, NAC2↑
	CKX5, 9↑	CKX5, 9↓
CKX expression root	CKX5, NAC2↓↓	CKX5↑↑	CKX5↑↑↑, NAC2↑
	CKX3↓	CKX3↑, NAC2↑
	CKX11, 10↑	CKX11, 10↓	CKX11↓
yield-related traits	yield↑	yield↓	yield↓↓
	CKX act. spike=	CKX act. spike=	CKX act. spike↓↓
	root=↓	root=↑	root=↓
	CKX act. root↓↓	CKX act. root↑↑	CKX act. root↑↑
	C2 = S6C × S12B
CKX expression 7 DAP	CKX5, 9↓	CKX5, 9↑	CKX9, 10, 11, 5↑
	CKX1, 2.1, NAC2↑	CKX1,  2.1, NAC2↓	CKX1, NAC2↓
CKX expression root	CKX10, 11↓↓	CKX10, 11↑↑	CKX5, 8, 10, 11↑↑
	CKX1, 3, 5, NAC2↑	CKX1, 3, 5, NAC2↓	CKX1, 3, NAC2↓
yield-related traits	yield↓	yield↑	yield=↓
	CKX act. spike=	CKX act. spike=	CKX act. Spike=
	root↑	root=↓	root↑↑
	CKX act. root↑	CKX act. root↓	CKX act. root=
	C3 = D16 × KOH7
CKX expression 7 DAP	CKX11↓↓	CKX11↑↑	CKX5, 9↑↑
	CKX2.1, 2.2.2↓	CKX2.1, 2.2.2↑
	CKX10↑	CKX10↓
CKX expression root	NAC2↓↓	NAC2↑↑	CKX3↓
	CKX1, 5, 10, 11↓↓	CKX1, 5, 10, 11↑↑	CKX1, 8, 11, NAC2↑↑
yield-related traits	yield↑	yield↓	yield↓
	CKX act. spike=↑	CKX act. spike=↓	CKX act. spike↓↓
	root=↑	root=↓	root=
			semi-empty spikes↑↑↑
	C4 = KOH7 × D16
CKX expression 7 DAP	CKX9, 10↓	CKX9, 10↑	CKX5, 9↑↑
	CKX11↑↑	CKX11↓↓	CKX11, NAC2↓↓
	CKX2.1, 2.2.2↑	CKX2.1, 2.2.2↓	CKX2.1, 2.2.2↓
CKX expression root	CKX3, 8↓	CKX3, 8↑	CKX8↑↑
	CKX1, 5, 10, 11↑	CKX1, 5, 10, 11↓	CKX3,10↓
	NAC2↑↑	NAC2↓↓
yield-related traits	yield↓	yield↑	yield↑
	CKX act. spike=↓	CKX act. spike=↑	CKX act. spike=
	root↓	root↑	root=↓
	C5 = D19 × D16
CKX expression 7 DAP	CKX2.2.2↑	CKX2.2.2↓	CKX2.2.2↓
	CKX10↓	CKX10↑	CKX9↑
CKX expression root	CKX 5, 8, NAC2↑	CKX5, 8, NAC2↓↓	CKX3, 5, 8, 11↓
	CKX1, 3, 10, 11↑	CKX1, 3, 10, 11↓	(CKX10, NAC2↑)
yield-related traits	yield↑	yield↓	yield↓↓
	CKX act. spike=↓	CKX act. spike=↑	CKX act. spike=↑
	root=	root=	root↓
	semi-empty spikes↓	semi-empty spikes↑	semi-empty spikes↑↑↑
	C6 = D16 × D19
CKX expression 7 DAP	CKX2.2.2, 5, 9↓	CKX2.2.2, 5, 9↑
	CKX10↑	CKX10↓	CKX1, 10, 11, NAC2↓
CKX expression root	CKX5, 8, NAC2↓↓	CKX5, 8, NAC2↑↑	CKX1, NAC2↑↑
	CKX1, 3, 10, 11↓	CKX1,  3,10, 11↑	CKX8, 10↑
yield-related traits	yield↓	yield↑↑	yield↑
	CKX act. spike=↑	CKX act. spike=↓	CKX act. spike=
	root=	root=	root↓
	C7 = D19 × KOH7
CKX expression 7 DAP	CKX2.1, 11, NAC2↓	CKX2.1, 11, NAC2↑	CKX10↑
	CKX9↑	CKX9↓	CKX2.2.2, 9↓
			NAC2↓
CKX expression root	NAC2↓↓	NAC2↑↑
	CKX1, 10, 11↓	CKX1, 10, 11↑	CKX11↑
	CKX3, 5, 8↑	CKX3, 5, 8↓	CKX3, 5, 8, 10↓
yield-related traits	yield↑	yield↓	yield↓
	CKX act. spike=	CKX act. spike=	CKX act. spike↑
	root=↑	root=↓	root=
		semi-empty spikes↑	semi-empty spikes↑↑
	C8 = KOH7 × D19
CKX expression 7 DAP	CKX5, 9, 10↓	CKX5, 9, 10↑	CKX9, 10↑
	CKX11↑	CKX11↓	CKX11, NAC2↓
	CKX2.1, NAC2=↑	CKX2.1, NAC2=↓	CKX2.1, 2.2.2↓
CKX expression root	CKX8↓↓	CKX8↑↑	CKX3, 8↑↑
	CKX3, 5↓	CKX3, 5↑	CKX1, 10, NAC2↓
	NAC2↑↑	NAC2↓↓
yield-related traits	yield↓	yield↑↑	yield↑↑
	CKX act. spike=	CKX act. spike=	CKX act. spike↑
	root=↓	root=↑	root=

Bold—cross number and parents.

by colours, most of the expression patterns tested for TaCKX GFM and NAC2 are inherited from the male parent (red). For example, up-regulated in 7 DAP spikes of the paternal parent TaCKX1 and NAC2 compared to the maternal parent is up-regulated in F₂ as well. The upregulated TaCKX5 and NAC2 and downregulated TaCKX11 in seedling roots of the paternal parents are similarly expressed in an F₂. To summarize, the expression levels of all tested TaCKX GFMs and NAC2 in 7 DAP spikes, in addition to being represented in different crosses, showed similar expression patterns to the paternal parents and were independent of the cross path. Among the TaCKX GFMs in 7 DAP spikes, which showed the paternal expression patterns were TaCKX1, 2.1, 2.2.2, 5, 9, 10, and 11. In the seedling roots, there were TaCKX5, 11, NAC2; TaCKX10, 11, NAC2; TaCKX1, 11, NAC2; TaCKX10; 1, 3, 5, 8, 10, 11, NAC2; TaCKX1, 8, 10, NAC2; and 3, 5, 8, 11; NAC2 (all tested but represented in different crosses). The only exceptions are TaCKX5 in 7 DAP spikes of S12B × S6C (C1) and TaCKX3 in seedling roots of KOH7 × D16 (C4), whose expression level is similar to that of the maternal parent (green).

Yield-related traits are represented by total grain yield and root mass (Table 1). Interestingly, grain yield in 7 out of 8 crosses is inherited from the paternal parent. The exceptions are the F₂ progeny of S6C × S12B (C2), which show very large differences in yield, exceeding parental data. The root mass in F₂ was lower than that in both parents or higher than that in both parents. In the first case, the root mass in the paternal parent was higher than that in the maternal parent, and in the second, the root mass in the paternal parent was lower than that in the maternal parent.

The results of crossing the low-yielding maternal parent with the higher-yielding paternal parent and their accompanying up- or down-regulated TaCKX GFMs and NAC2 in F₂ generations are presented in Figure 2.

Depending on the crosses, downregulation of TaCKX5 with TaCKX9 and upregulation of NAC2 in spikes of the low-yielding maternal parent and the opposite regulation of these genes in spikes of the higher-yielding paternal parent resulted in high-yielding F₂, characterized, as in the paternal component, by a higher expression level of TaCKX9 and a lower expression level of NAC2. Upregulation of TaCKX2.1 and 11 in spikes of the maternal parent and downregulation of these genes in the paternal parent were associated with downregulation of TaCKX11 in high-yielding F₂. Similarly, the upregulation of TaCKX2.2.2 and the downregulation of TaCKX10 in the spikes of the low-yielding maternal parent and opposite regulation of these genes, and the yield in the paternal parent, resulted in the downregulation of TaCKX2.2.2 and the upregulation of TaCKX10 in the spikes of the high-yielding F₂. The expression of TaCKX3 in seedling roots of the high-yielding paternal parent and F₂ was upregulated. However, TaCKX8 expression was upregulated, and NAC2 was downregulated in the same organ of the paternal parent, but these genes were up- or down-regulated in F₂, depending on the cross.

2.2. Cooperating and Opposite-Functioning Genes

TaCKX5 with TaCKX9 (yellow) and TaCKX2.1 with TaCKX11 (green) showed coordinated up- or downregulation in 7 DAP spikes of the paternal parent of C1, C2, C6, and C8 crosses; and C3, C4, C7, and C8 crosses, respectively (

Table 2. Coordinated expression of TaCKX GFMs and NAC2 genes in 7 DAP spikes and seedling roots of the paternal parent (P) of four reciprocal crosses.

Cross	P Spike P Expr. +/P expr. −	P Root P Expr. +/P Expr. −	F2 Spike/Root (Yield, Root Mass in F2)
C1	1, NAC2/5, 9	3, 5, NAC2/10, 11	NAC2+/NAC2+ y−, r = −, As−−
C2	5, 9/1, NAC2	10, 11/3, 5, NAC2	NAC2−/NAC2− y = −, r+, As =
C3	2.1, 11, 2.2.2/10	NAC2, 1, 5, 10, 11/3	10−, 11+/10+,11+, NAC2+ y−, r=, As−, s-e+++
C4	9, 10/2.1, 11, 2.2.2	3, 8/1, 5, 10, 11, NAC2	10+, 11−/10−, 11−, NAC2− y+, r = −, As=
C5	2.2.2/10	?/3, 5, 8, NAC2, 1, 10, 11	10−/10−, NAC2− y−, r−, As = +, s-e+++
C6	2.2.2, 5, 9/10	3, 5, 8, NAC2, 1, 10, 11/none	5+, 10−/ 5+, 10+, NAC2+ y+, r−, As=
C7	2.1, 11, NAC2/9	1, 10, 11, NAC2/3, 5, 8	11+, NAC2+/11+, NAC2+ y−, r=, As+, s-e+++
C8	5, 9, 10/2.1, 11, NAC2	3, 5, 8/NAC2	5+, NAC2−/5+, NAC2− y++, r=, As+

P—paternal parent; expr.+—upregulated; expr.—downregulated; 1, 3, 5, 9…-TaCKX GFMs, As—CKX activity spike; s-e+++—high number of semi-empty spikes, y—yield, r—root mass.

). Higher expression of TaCKX5 and 9 in this parent was associated with a higher yield in F₂. However, a higher coordinated expression of TaCKX2.1 with TaCKX11 in the paternal parent determined a lower yield in F₂, and, in contrast, a lower expression of these two genes in the paternal parent was associated with a higher yield.

In the 7 DAP spikes, the paternal parent of the C3, C4, C5, and C6 crosses, TaCKX2.2.2, showed opposite expression to TaCKX10 (blue), and upregulation of the first and downregulation of the second were associated with lower yield (but not in C6). In the paternal parent of the C1, C2, C7, and C8 crosses, NAC2 was oppositely expressed to TaCKX9; in these crosses, a high yield was observed when TaCKX9 was upregulated and NAC2 was downregulated, and vice versa (only in C7 and C8). Furthermore, upregulated TaCKX5 and downregulated TaCKX1 were associated with high root mass in C2 and conversely in the reverse cross (C1).

Among the TaCKX genes coordinately expressed in the paternal seedling roots were TaCKX3, 5, and 8 (green) in C5, C6, C7, and C8 crosses; TaCKX3 and 8 (green) in C1 and C2 crosses; TaCKX10, 11, and 1 (yellow), and NAC2 in C3 to C7 crosses; and TaCKX10 and 11 (yellow) in C1 and C2 crosses. However, in one reciprocal cross, C3 and C4, the expression of TaCKX3 and TaCKX5 was opposite, and in the case of upregulation of TaCKX3 and downregulation of TaCKX5, the grain yield in F₂ was higher.

In three reciprocal crosses, NAC2 was downregulated in paternal roots (C2, C4, and C8), and in two of them (C2 and C8), NAC2 was downregulated in paternal spikes as well. This negative regulation of NAC2 occurred in the F₂ progeny, which was accompanied by a higher yield and a higher or similar to the parents’ mass of the seedling roots. In contrast, in another way crosses, when expression of NAC2 was increased in paternal roots (C1, C3, and C7) and was upregulated in paternal spikes, the same was observed in F₂ progeny, characterized by lower yield and lower or similar to the parent mass of the roots.

The higher yield in F₂ has been associated with the same or higher CKX activity, as in the paternal parent, in 7 DAP spikes. A higher number of semi-empty spikes, which occurred in low-yielding F₂ of the C3, C5, and C7 crosses, was accompanied by downregulated TaCKX10 and/or upregulated TaCKX11 in 7 DAP spikes and upregulated TaCKX10, 11, and NAC2 or downregulated TaCKX10 and NAC2 in seedling roots.

2.3. The Correlation Coefficients between TaCKX GFMs and NAC2 Expression, CKX Activity, and Yield-Related Traits Were Significant for Both Parents or the Maternal or Paternal Parent Separately

The correlation coefficients between TaCKX GFMs and NAC2 expression, CKX activity, and yield-related traits in reciprocal crosses were analyzed separately for the maternal parent and F₂, paternal parent, and F₂ for each cross (Table S1).

2.4. Correlations between TaCKX GFM and NAC2 Expression and Yield-Related Traits in the Group of Maternal Plants, and F₂ and Paternal Plants, and F₂ of Reciprocal Crosses

Seed number and spike number were positively correlated (

Table 3. Correlations between TaCKX GFM and NAC2 expression in 7 DAP spikes or seedling roots, and yield-related traits in the group of maternal plants and F₂, and paternal plants and F₂ of reciprocal crosses.

		7 DAP Spike	Seedling Root	Yield-Related Traits	F2 Phenotype
		Seed Number
C1 S12B × S6C	M + F2	CKX1−, CKX5−	CKX11+	spike number+	yield−−, CKX act. −−, root=−, CKX act. root++
	P + F2	nc	CKX11+	spike number+
C2 S6C × S12B	M + F2	nc	nc	CKX act. root−, spike number++	yield−, CKX act.=, root+, CKX act. root=
	P + F2	nc	CKX3+	CKX act. root−, spike number+
C3 D16 × KOH7	M + F2	CKX2.1+,CKX2.2.2+	CKX3+, CKX8+,CKX11−	plant height+, spike number+	yield−, CKX act. −−, root=−, semi-empty spikes++
	P + F2	CKX2.1+, CKX2.2.2+	CKX1−, CKX3++, CKX8+, CKX11−	plant height+, spike number+
C4 KOH7 × D16	M + F2	nc	nc	plant height+, spike number+	yield+, CKX act.=, root=
	P + F2	nc	nc	plant height+, spike number+
C5 D19 × D16	M + F2	CKX1−	CKX3+, CKX5+, NAC2+	plant height+, spike number+	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	CKX3+, NAC2+	plant height+ spike number+
C6 D16 × D19	M + F2	nc	nc	spike number+	yield+, CKX act.=, root=−
	P + F2	nc	nc	spike number+
C7 D19 × KOH7	M + F2	nc	CKX1−	spike number++	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	CKX1−	spike number++
C8 KOH7 × D19	M + F2	nc	CKX1−	empty spikes−, spike number+	yield++, CKX act.=+, root=
	P + F2	CKX2.1+	CKX1−	empty spikes−, spike number+
		Seed yield
C1 S12B × S6C	M + F2	CKX11−, NAC2−	CKX11+	spike number+, seed number++	yield−, CKX act. −−, root=−, CKX act. root++
	P + F2	nc	CKX11+	spike number+, seed number++
C2 S6C × S12B	M + F2	nc	CKX10+	spike number+, seed number++	yield−, CKX act.=, root+, CKX act. root=
	P + F2	nc	CKX10+, CKX3+	spike number+, seed number++
C3 D16 × KOH7	M + F2	CKX2.1+, CKX2.2.2+	CKX3+, CKX8+	plant height+, spike number+, seed number++	yield−, CKX act. −−, root=−, semi-empty spikes++
	P + F2	CKX2.1+, CKX2.2.2+	CKX3+, CKX8+	plant height+, spike number+, seed number++
C4 KOH7 × D16	M + F2	nc	nc	plant height+, spike number+, seed number++	yield+, CKX act.=, root=
	P + F2	nc	nc	plant height+, spike number+, seed number++
C5 D19 × D16	M + F2	CKX1−	CKX3+, CKX5+, NAC2+	plant height+, spike number+, seed number++	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	CKX3+, NAC2+	plant height+, spike number+, seed number++
C6 D16 × D19	M + F2	CKX2.1+	nc	plant height+, spike number+, seed number++	yield+, CKX act.=, root=−
	P + F2	CKX1−	nc	plant height+, spike number+, seed number++
C7 D19 × KOH7	M + F2	nc	nc	plant height+, spike number++, seed number++	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	nc	plant height+, spike number++, seed number++
C8 KOH7 × D19	M + F2	nc	CKX1−	spike number+, seed number++	yield++, CKX act.=+, root=
	P + F2	nc	CKX1−	spike number+, seed number++
		Spike number
C1 S12B × S6C	M + F2	CKX2.1+, CKX2.2.2+, CKX11−	CKX11+	plant height+	yield−, CKX act.−−, root=−, CKX act. root++
	P + F2	CKX2.1+, CKX2.2.2+	nc	CKX act.+, plant height+
C2 S6C × S12B	M + F2	CKX2.2.2+	nc	nc	yield−, CKX act.=, root+, CKX act. root=
	P + F2	nc	NAC2−	nc
C3 D16 × KOH7	M + F2	nc	CKX8+	nc	yield−, CKX act. −−, root=−, semi-empty spikes++
	P + F2	nc	CKX8+	nc
C4 KOH7 × D16	M + F2	nc	CKX5+	nc	yield+, CKX act.=, root=
	P + F2	nc	CKX5+	nc
C5 D19 × D16	M + F2	nc	CKX1−, CKX5+, NAC2+	nc	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	CKX1−, CKX5+, NAC2+	nc
C6 D16 × D19	M + F2	nc	nc	nc	yield+, CKX act.=, root=−
	P + F2	CKX2.1+	nc	nc
C7 D19 × KOH7	M + F2	nc	CKX1−	nc	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	CKX1−, CKX8−	nc
C8 KOH7 × D19	M + F2	nc	CKX1−, CKX5−, CKX8−	nc	yield++, CKX act.=+, root=
	P + F2	nc	CKX1−, CKX5−	nc
		TGW
C1 S12B × S6C	M + F2	CKX2.1+, CKX10+, NAC2+	nc	nc	yield−, CKX act. −−, root=−, CKX act. root++
	P + F2	CKX2.2.2+, NA2C+	nc	nc
C2 S6C × S12B	M + F2	nc	nc	CKX act. root++, seed number−	yield−, CKX act.=, root+, CKX act. root=
	P + F2	nc	nc	CKX act. root++, seed number−
C3 D16 × KOH7	M + F2	nc	nc	plant height+, yield+, semi-empty−, seed number+	yield−, CKX act. −−, root=−, semi-empty spikes++
	P + F2	nc	nc	plant height+, yield+
C4 KOH7 × D16	M + F2	CKX2.2.2−	CKX10+, NAC2+	CKX act.−	yield+, CKX act.=, root=
	P + F2	CKX2.2.2−	CKX10+,NAC2+	CKX act.−
C5 D19 × D16	M + F2	nc	NAC2+	yield+, seed number+	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	NAC2+	yield+
C6 D16 × D19	M + F2	nc	nc	plant height+	yield+, CKX act.=, root=−
	P + F2	nc	nc	plant height+
C7 D19 × KOH7	M + F2	CKX2.1−, CKX2.2.2−, CKX9−	nc	spike length−	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	CKX2.1−,CKX2.2.2−, CKX5+	nc	nc
C8 KOH7 × D19	M + F2	CKX2.1−, CKX2.2.2−	nc	nc	yield++, CKX act.=+, root=
	P + F2	CKX2.1−, CKX2.2.2−!, CKX10−	nc	nc
		Root mass
C1 S12B × S6C	M + F2	CKX5+,CKX11+, NAC2+	CKX1−, CKX3−, CKX10−, CKX11−,	seed number−	yield−, CKX act. −−, root=−, CKX act. root++
	P + F2	nc	NAC2−, CKX1−, CKX10−, CKX11−	seed number−
C2 S6C × S12B	M + F2	NAC2−	nc	nc	yield−, CKX act.=, root+, CKX act. root=
	P + F2	nc	nc	nc
C3 D16 × KOH7	M + F2	CKX1+, CKX11+	CKX5+	nc	yield−, CKX act.−−, root=−, semi-empty spikes++
	P + F2	CKX1+, CKX11+, NAC2+	CKX5+, CKX8−	nc
C4 KOH7 × D16	M + F2	nc	CKX3+, CKX8+, NAC2−	CKX act.+	yield+, CKX act.=, root=
	P + F2	CKX2.1+, CKX2.2.2+, CKX10+	CKX3+, CKX8+, NAC2−	nc
C5 D19 × D16	M + F2	NAC2−	CKX1−, CKX11+	nc	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	CKX2.1+	nc	nc
C6 D16 × D19	M + F2	CKX2.1+, CKX2.2.2+	nc	yield+	yield+, CKX act.=, root=−
	P + F2	CKX2.2.2+	nc	plant height+
C7 D19 × KOH7	M + F2	nc	CKX1−	nc	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	NAC2−	CKX1−	spike length+
C8 KOH7 × D19	M + F2	CKX10−	CKX3−	nc	yield++, CKX act.=+, root=
	P + F2	nc	CKX3−	yield+
		Plant height
C1 S12B × S6C	M + F2	nc	nc	nc	yield−, CKX act. −−, root=−, CKX act.root++
	P + F2	nc	nc	nc
C2 S6C × S12B	M + F2	CKX11+	CKX1−	nc	yield−, CKX act.=, root+, CKX act.root=
	P + F2	nc	CKX5−, CKX11+	nc
C3 D16 × KOH7	M + F2	nc	CKX11−	nc	yield−, CKX act.−−, root=−, semi-empty spikes++
	P + F2	nc	CKX11−	nc
C4 KOH7 × D16	M + F2	nc	nc	nc	yield+, CKX act.=, root=
	P + F2	CKX2.2.2−	NAC2+	nc
C5 D19 × D16	M + F2	CKX1−	CKX1−, NAC2+	nc	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	CKX1−, NAC2+	nc
C6 D16 × D19	M + F2	CKX1−, CKX5−	CKX5−	nc	yield+, CKX act.=, root=−
	P + F2	CKX5−	CKX5−	nc
C7 D19 × KOH7	M + F2	nc	nc	nc	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	nc	nc
C8 KOH7 × D19	M + F2	CKX2.2.2−, CKX9−	CKX11+	nc	yield++, CKX act.=+, root=
	P + F2	CKX9−	nc	nc
		Spike length
C1 S12B × S6C	M + F2	NAC2−	nc	semi-empty spikes−, seed number+, yield+	yield−, CKX act.−−, root=−, CKX act. root++
	P + F2	NAC2−	nc	spike number+, seed number+, yield+
C2 S6C × S12B	M + F2	CKX9+, CKX11+	nc	plant height+	yield−, CKX act.=, root+, CKX act. root=
	P + F2	nc	nc	nc
C3 D16 × KOH7	M + F2	CKX2.1+	CKX3+, CKX11−	plant height+, seed number+, yield+	yield−, CKX act.−−, root=−, semi-empty spikes++
	P + F2	CKX2.1+, CKX2.2.2+	CKX1−, CKX3++, CKX11−	plant height+, seed number+, yield+
C4 KOH7 × D16	M + F2	CKX1+, CKX10+	nc	plant height+, seed number+, yield+	yield+, CKX act.=, root=
	P + F2	nc	nc	seed number+, yield+
C5 D19 × D16	M + F2	CKX5−, CKX11+, NAC2−	CKX11+	nc	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	nc	seed number+, yield+
C6 D16 × D19	M + F2	nc	nc	spike number+, seed number+, yield+	yield+, CKX act.=, root=−
	P + F2	nc	nc	seed number+, yield+
C7 D19 × KOH7	M + F2	CKX1+, CKX9+	nc	seed number+	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	CKX9+, NAC2−	nc	nc
C8 KOH7 × D19	M + F2	nc	nc	nc	yield++, CKX act.=+, root=
	P + F2	CKX2.1+	nc	seed number+
		Semi-empty spikes
C1 S12B × S6C	M + F2	CKX5+	CKX5−, CKX10−	nc	yield−, CKX act. −−, root=−, CKX act. root++
	P + F2	CKX5+!, CKX9+, CKX11+	nc	nc
C2 S6C × S12B	M + F2	CKX9+	nc	nc	yield−, CKX act.=, root+, CKX act. root=
	P + F2	CKX9+	CKX8−	empty spikes+
C3 D16 × KOH7	M + F2	CKX9+	nc	nc	yield−, CKX act.−−, root=−, semi-empty spikes++
	P + F2	CKX11−	nc	nc
C4 KOH7 × D16	M + F2	nc	nc	spike number+	yield+, CKX act.=, root=
	P + F2	nc	nc	spike number+
C5 D19 × D16	M + F2	nc	nc	nc	yield−−, CKX act.+, root=−, semi-empty spikes++
	P + F2	CKX2.1−	nc	nc
C6 D16 × D19	M + F2	CKX5+	CKX5+	spike number+	yield+, CKX act.=, root=−
	P + F2	CKX5+	nc	seed number+
C7 D19 × KOH7	M + F2	CKX10+	nc	nc	yield−, CKX act.+, root=−, semi-empty spikes++
	P + F2	nc	nc	nc
C8 KOH7 × D19	M + F2	nc	CKX5−, CKX8−,NAC2−	nc	yield++, CKX act.=+, root=
	P + F2	CKX10+	CKX5−, NAC2−	nc

All correlation coefficients ≥0.60; bold—significant correlation coefficients; nc—no correlation; +—positive correlation; =+—low positive correlation; ++—strong positive correlation; −—negative correlation; =−—low negative correlation; −−—strong negative correlation.

and Table S1); however, each of these yield-related traits was correlated with different TaCKX GFMs.

2.4.1. Seed Number

The decrease in seed number in maternal plants and their F₂ (M and F₂) of the C1 cross (S12B × S6C) was strongly negatively correlated with upregulated TaCKX1 and TaCKX5 in spikes and positively correlated with the downregulated TaCKX11 in the seedling roots of F₂. There was no significant correlation between the expression of TaCKX GFM and the yield-related traits in the groups of paternal plants (P) and F₂ in the same cross, and M and F₂, and P and F₂ in the reverse, C2 cross. The F₂ progeny in this reverse cross showed a similar yield and greater root mass compared to the parents.

In both M and F₂, and P and F₂ of C3, the decrease in seed number was strongly positively correlated with TaCKX2.1 and TaCKX2.2.2 in spikes and positively correlated with TaCKX3 and TaCKX8 but negatively correlated with TaCKX11 in seedling roots. These correlations were not significant in the reverse, C4 cross, in which the M (KOH7) and F₂ plants showed higher yields.

The decrease in seed number in M × F₂ of C5 was negatively correlated with TaCKX1 in spikes and positively correlated with downregulated TaCKX3 and 5, and upregulated NAC2 in seedling roots. There were also positive correlations of TaCKX3 in roots between P and F₂ of the same cross. These correlations were not significant in the reverse C6 cross; however, F₂ of this cross was characterized by higher yield and similar root mass than in the parents.

The increase in seed number was strongly positively correlated with downregulated TaCKX2.1 in spikes and negatively correlated with downregulated TaCKX1 in the seedling roots only in F₂ progeny of a C8 cross, 14K (in the case of P × F₂ only for TaCKX2.1).

2.4.2. Spike Number

The decrease in the number of spikes in M × F₂ and P × F₂ of the C1 cross was strongly positively correlated with TaCKX2.1 and TaCKX2.2.2 in spikes and positively correlated with TaCKX11 in seedling roots (only in M × F₂). There was also a significant and positive correlation between the expression of TaCKX2.2.2 and the number of spikes in the reverse C2 cross of M and F₂. Furthermore, in the same cross, the spike number was negatively correlated with downregulated NAC2, but only in the P × F₂ group. There were no significant correlations between TaCKX GFM expression and spike number in M × F₂ and P × F₂ spikes of C3 and C4. However, there was a strong and positive correlation of the spike number with TaCKX8 in the roots of C3 and TaCKX5 in the roots of C4.

The decreased spike number in M × F₂ and P × F₂ of C5 was not correlated with any TaCKX expressed in the spikes but was negatively correlated with TaCKX1, positively correlated with TaCKX5, and positively correlated with NAC2 expressed in the seedling roots. Conversely, in reverse C6 cross, there was a positive correlation of the spike number with TaCKX2.1 in a P × F₂, which resulted in a higher yield phenotype in the F₂.

The spike number in C7 and C8 crosses was not correlated with the level of expression of any gene tested in the spikes; however, it was negatively correlated with the expression of TaCKX1, 5, and 8 in seedling roots.

2.4.3. TGW

TGW was positively correlated with TaCKX2.1, 10, and NAC2 in spikes of M and F₂ of C1 and with TaCKX2.2.2 of P and F₂ of the same cross. There was no correlation in F₂ between the expression of the genes tested and TGW in spikes of the C2 and roots of the C1 and C2 crosses. There were no correlations between the TGW and TaCKX genes in the spikes and roots of C3. However, there was a negative correlation of this trait with TaCKX2.2.2 in spikes and positive correlations with TaCKX10 and NAC2 in roots of the reciprocal C4 cross. Positive correlations of NAC2 with TGW were also observed in the roots of C5 but not in those of C6. Negative correlations of TaCKX2.1 and 2.2.2 in spikes with TGW were observed in both reciprocal crosses, C7 and C8. Additionally, TaCKX9 was negatively correlated with the trait in M + F_2, TaCKX5 was positively correlated with the trait in P + F₂ of C7, and TaCKX10 was negatively correlated with TGW in P + F₂ of C8. There was no correlation between TGW and any gene expression in roots.

2.4.4. Root Mass

The mass is positively correlated with the expression of TaCKX5, 11, and NAC2 in spikes of M + F₂ of C1 and negatively correlated with NAC2 in M + F₂ of C2. A positive correlation between root mass and TaCKX11, and NAC2 was also visible in P + F₂ of C3, and a negative correlation between trait and NAC2 expression was also observed in spikes of M + F₂ of C5 and P + F₂ of C7. Furthermore, in the C1 cross, this trait was negatively correlated with the expression of TaCKX1, 3, 10, and NAC2 in roots of M + F₂ and with TaCKX11 in roots of P + F₂. There were also positive correlations between root mass and TaCKX1 (M + P of C3), root mass and TaCKX2.1 (P + F₂ of C4; P + F₂ of C₅; M + F₂ of C6), and root mass and TaCKX2.2.2 (P + F₂ of C4; M + F₂ and P + F₂ of C6). The expression of another gene, TaCKX10, in spikes, was positively correlated with root mass in P + F₂ of C4 but negatively correlated in M + F₂ of C8. Correlations between root mass and gene expression tested in roots were dependent on the parent and cross. There were negative correlations with TaCKX1 in 5 out of 16 combinations tested, negative correlations with TaCKX3 in 3 combinations, but positive correlations in two combinations, positive correlations with TaCKX5 in two combinations, and single positive or negative correlations with TaCKX8, 11, and NAC2. The root mass in single combinations was positively correlated with the yield (twice), height of the plant (once), and length of the spike (once), and negatively correlated with seed number of seeds (twice).

2.4.5. Semi-Empty Spikes

The number of semi-empty spikes was positively correlated with the expression of TaCKX9 in the P + F₂ C1, C2, and M + F₂ C3 crosses, positively correlated with TaCKX5 in the M + F₂, and P + F₂ C1 and C6 crosses, and positively correlated with TaCKX10 in the M + F₂ C7 and P + F₂ C8 crosses, all expressed in 7 DAP spikes. The negative correlation between the number of semi-empty spikes and TaCKX2.1 was in P + F₂ of C5, and between the same trait and TaCKX11 was in P + F₂ of C3. In seedling roots of various crosses, this trait was mainly negatively correlated with TaCKX5, 8, 10, and NAC2.

Generally, negative correlations between the expression of TaCKX2.1, 2.2.2, and 10 in spikes and TGW, seed number, seed yield, and spike number were correlated with higher yield, and positive correlations were correlated with lower yield. On the other hand, positive correlations between the expression of these genes and root mass determine a higher yield in F₂. Higher yield in F₂ is also associated with balanced CKX enzyme activity in spikes and seedling roots.

A summary of the regulation of yield-related traits by TaCKX GFMs and NAC2 in the high-yielding progeny of F₂ is presented in Figure 3.

3. Discussion

Common wheat is a very important cereal crop for feeding the world’s population; therefore, continued improvement of the yield of this species is significant. CKX GFMs have already been documented to perform a pivotal role in determining yield-related traits in many plant species, including wheat [4,12]. The genes are tissue-specific; they encode cytokinin oxidase/dehydrogenase, the enzyme that irreversibly degrades cytokinins. We have already characterized the role of TaCKX1 and TaCKX2 in the regulation of yield traits in awnless and owned-spike cultivars [5,6,7]. The range of natural variation in the expression levels of most TaCKX genes among breeding lines and cultivars was very high, indicating the possibility of selecting beneficial genotypes for breeding purposes [8]. Therefore, we were interested in how the expression of these genes is inherited.

3.1. The Expression Patterns of Most TaCKX GFMs and TaNAC2-5A Are Mainly Inherited from the Paternal Parent

Comparison of the expression patterns of most of the TaCKX GFMs and yield-related traits between parents and F₂ progeny in all reciprocal crosses tested indicated their inheritance from the paternal parent. This rule includes expression patterns in both tissues tested, 7 DAP spikes, and seedling roots, and all TaCKX GFMs and TaNAC2-5A tested were represented in different crosses. The exception was TaCKX5 expressed in 7 DAP spikes, and TaCKX3 expressed in seedling roots, for which the expression level in single crosses was inherited from the maternal parent. Furthermore, high or low yield was predominantly inherited from the paternal parent, and root mass was inherited from both parents or in one reciprocal cross from the maternal parent. We have not found such examples of inheritance in the literature; however, some deviations from parental additivity of expression in polyploid plants were described [2]. An example of such non-additive gene expression takes place when the gene expression level in progeny is higher than that of one parent. The expression level dominance of one parent, also called genomic imprinting, is epigenetic in origin and was investigated primarily at the molecular level in plants and animals [36,37]. The main regulators of gene imprinting are DNA and histone methylation asymmetries between parental genomes. Most of the imprinted genes in the endosperm of grains of different rice cultivars are imprinted across cultivars, and their functions are associated with the regulation of transcription, development, and signaling [39]. Imprinting might affect a single gene or a group of genes. Genes that showed conserved imprinting in cereals have been shown to reveal positive selection and were suggested to regulate seed development in a dose-dependent manner [38]. The only example of a paternally imprinted locus in maize is ded1, which encodes a transcription factor specifically expressed during early embryo development and activates early embryo genes that contribute to grain set and weight [47]. To our knowledge, there are no examples of paternally inherited expression patterns. According to Arabidopsis research, imprinted paternally expressed genes during seed development are mainly related to hypomethylated maternal alleles, repressed by small RNAs or less frequently with transposable elements [48,49]. Contrary to developmental epigenetics, in the case of transgenerational epigenetics, these epigenetic changes do not reset between generations, and this type of inheritance is more related to plants than animals (heritable changes in DNA methylation) [54]. Therefore, we suggest that this paternal inheritance of selected TaCKX GFMs is an effect of transgenerational epigenetic changes, not reset between generations. These heritable epigenetic changes might be effects of DNA methylation, repression of maternal alleles by small RNAs, transposable elements, or, most likely, transcription factors. From our in silico analysis and expression analysis (Iqbal et al., not published yet), several NAC transcription factors appear to strongly regulate the expression of TaCKX GFMs and TaIPT GFMs, influencing yield-related traits [24].

3.2. Cooperation of TaCKX GFMs and TaNAC2-5A in the Determination of Yield-Related Traits

The coordinated high or low level of expression of a few groups of genes in the paternal parent positively or negatively regulates higher or lower yield. In two reciprocal crosses, where both TaCKX5 and TACKX9 showed high expression in 7 DAP spikes of the paternal parent, the yield in the F₂ progeny was high and vice versa. In others, the high yield in the F₂ progeny was determined by a low level of expression of TaCKX2.1 and TaCKX11 in spikes of the paternal parent and high levels of their expression in the maternal parent. The level of expression of TaNAC2-5A in the paternal parent and/or F₂ was in opposition to TaCKX5 and 9; however, it was in agreement with TaCKX11 and TaCKX2.1, suggesting their role in the regulation of transcription of these genes. In fact, it was proven by correlation analysis of its expression with yield-related traits [8]. Opposite cooperation of some of the genes in paternal spikes, which resulted in high or low yield in F₂, has also been observed. The high level of expression of TaCKX2.2.2 and the low level of expression of TaCKX10 predominantly resulted in the low yield in F₂ progeny and vice versa.

Such common rules of gene expression in the paternal parent associated with yield in the paternal parent and F₂ progeny were also observed in the seedling roots. The high level of expression of TaCKX3 and TaCKX8 in the paternal parents of three reciprocal crosses resulted in high yield in the F₂ progeny and vice versa. The expression of TaNAC2-5A in spikes and seedling roots of high-yielding paternal parents and F₂ progeny showed a predominantly low expression level and inversely. These principles of paternal inheritance of selected TaCKX GFMs and TaNAC2-5A expression associated with high yield could be directly involved as molecular markers in high-yielding wheat breeding.

3.3. Regulation of Yield-Related Traits by TaCKX GFMs and TaNAC2-5A in the F₂ Generation

The grain number, grain yield, spike number, and TGW were strongly positively correlated with TaCKX2.1 and TaCKX2.2.2 independent of the parent; however, only in the crosses resulted in decreased yield. In contrast, negative correlations were observed between TaCKX2.1, TaCKX2.2.2, and TGW in a reciprocal cross of C7/C8 and TaCKX2.2.2 in a one-way cross (C4), in which the F₂ progeny had a higher yield. All these correlations prove our earlier observations [6,7]. Modified wheat lines with 60% decreased expression of TaCKX2.2.2, and a slight decrease in the TaCKX2.2.1 and 2.1 genes exhibit a significantly higher TGW and slightly increased yield [7]. Interestingly, this result was observed in cultivars and breeding lines that represent awnless spikes. In the owned-spike cultivar, silencing of the TaCKX2 genes co-expressed with other TaCKX resulted in decreased yield; however, TGW was at the same level as in non-silent plants [5]. Furthermore, a strong feedback mechanism for regulation of the expression of TaCKX2 and TaCKX1 genes was observed in both awnless and owned-spike cultivars [5,6,7]. Silencing of TaCKX2 genes upregulated the expression of TaCKX1 and vice versa. This feedback mechanism could explain the observed positive correlations of the TaCKX2 genes with yield-related traits in low-yielding F₂ progeny and negative correlations in high-yielding F₂ progeny. A similar mechanism is visible when we analyze individual traits in high-yielding F₂, such as grain number, grain yield, and spike number. These traits are promoted by up-regulated in 7 DAP spikes TaCKX2.1 and down-regulated TaCKX1. Silencing of HvCKX1 in barley, which is an ortholog of TaCKX1, decreased CKX enzyme activity and led to increased seedling root mass and higher plant productivity [55]; however, knock-out of this gene caused a significant decrease in CKX enzyme activity but no changes in grain yield were observed [56]. These differences might be explained by differences in the level of decreased gene expression, which variously coordinate the expression of other genes, regulate phytohormone levels, and determine particular phenotypes, as was already documented in wheat [5,7].

The association of TaCKX2 genes with yield-related traits has also been reported in different wheat cultivars or genotypes. Zhang et al. [20] showed that TaCKX6 (renamed by Chen et al. [11] TaCKX2.2.1-3D), which is an ortholog of rice OsCKX2 associated with grain number [57], is related to grain weight. Another allele of CKX2, TaCKX6a02 [21], annotated as TACKX2.1 [58], significantly correlated with grain size, weight, and grain filling rate. Wheat plants with silenced by RNAi expression of TaCKX2.2.1-3A (originally TaCKX2.4) showed a strong correlation with the number of grains per spike implied by more filled florets [59]. Since TaCKX2.2.1-3D was associated with grain weight, these differences in functions between TaCKX2.2.1-3A and TaCKX2.2.1-3D were interpreted as subgenome-dependent.

Grain number was also negatively correlated with TaCKX1 and TaCKX5, and grain yield was negatively correlated with TaCKX1, predominantly for M and F₂, which were characterized by decreased grain number and lower yield. This observation is also in agreement with previous research. The silencing of TaCKX1 caused an increase in spike number and grain number but a decrease in TGW because this trait is opposite to grain number [6]. The low-yield progeny of F₂ showed positive correlations between the expression of TaCKX11, 3, 5, and 8, and NAC2 in the seedling roots and the grain number, the spike number and the grain yield; however, the higher-yield progeny of F₂ displayed a negative correlation between TaCKX1 and these yield-related traits in the seedling roots of some crosses. In summary, high-yielding F₂ was the result of upregulation of TaCKX2.1 in spikes and downregulation of TaCKX1 in seedling roots. As documented earlier, TaCKX11, 5, 8 and TaNAC2-5A are expressed in all organs, and their expression is correlated with the expression of spike-specific TaCKX2 and TaCKX1 [8,58].

Rice OsCKX11 is an orthologue of wheat TaCKX11 and is highly expressed in the roots, leaves, and panicles. The gene was shown to coordinate the simultaneous regulation of leaf senescence and grain number by the relationship of source and sink [60]. Since TaCKX11 is expressed in seedling roots and highly expressed in leaves, inflorescences, and 0, 7, and 14 DAP spikes, it could perform a similar function. This is partly proven by silencing of the TaCKX2 genes in awnless spikes of cv. Kontesa, which resulted in significant upregulation of TaCKX11 and growth of TGW, and chlorophyll content in flag leaves [7]. In contrast, TaCKX11 is significantly negatively regulated by TaCKX1, resulting in a higher spike number and grain number [6]. Its orthologue in rice, OsCKX11, was found to regulate leaf senescence and grain number by the coordinated source and sink relationship [60].

Based on a summary of the regulation of yield-related traits in high-yielding F₂, it is possible to identify singular genes or groups of genes that are up- or down-regulated in 7 DAP spike or seedling roots and specifically regulate yield-related traits. Upregulated in spikes TaCKX2.1 and downregulated in seedling roots TaCKX1 were found to determine grain number, grain yield, spike number, and spike length. Furthermore, upregulated in 7 DAP spikes TaCKX10 and downregulated TaNAC2-5A, together with others, depending on cross, control spike length, semi-empty spikes, root mass, and increased grain yield. As discussed above, high TGW is in contrary to high grain number and partly grain yield and was strongly determined by downregulated TaCKX2.2.2 together with TaCKX2.1 in 7 DAP spikes and upregulated TaCKX10 and NAC2 in seedling roots. The upregulated in seedling roots TaNAC2-5A participates in the determination of TGW and plant height, and the downregulation of TaNAC2-5A in seedling roots controls the development of semi-empty spikes and root mass.

In previous research, TaNAC2-5A has been documented as a gene encoding a nitrate-inducible wheat transcription factor. Overexpression of the gene improved root growth, grain yield, and grain nitrate concentration [25]. This is in agreement with our observations of growth of TGW but not enhanced roots. The increase was argued to be the consequence of regulation of nitrate concentration and its remobilization in developing grains by direct binding of the TaNAC2-5A protein to the promoter of the nitrate transporter, TaNRT2.5-3A and positive regulation of its expression [61]. The expression of TaNAC2-5A is coregulated by expressed in 7 DAP spikes TaCKX2 genes and expressed in 7 DAP spikes and seedling roots TaCKX1 gene [5,7]. Independent of awnless or awned-spike genotype, downregulation of TaCKX2 genes by RNAi significantly increased TaNAC2-5A expression, resulting in higher chlorophyll content in flag leaves and delayed leaf senescence. As discussed above, the strong feedback mechanism between the TaCKX2 and TaCKX1 genes implies that downregulation of TaCKX1 resulted in opposite results. Similar to our observations in wheat, an ortholog of TaNAC2-5A in rice, OsNAC2, was described as a negative regulator of crown root number and root length [62]. Its expression was positively correlated with cytokinin synthesis genes, OsIPT3, 5, the gene determining the formation of active cytokinins, OsLOG3, and negatively correlated with OsCKX4 and 5. The authors concluded that OsNAC2 stimulated cytokinin accumulation by suppressing CKX expression and stimulating IPT expression by binding the OsNAC protein to the promoters of these genes. Therefore, OsNAC2 functions as an integrator of cytokinin and auxin signals that regulate root growth. In our experiments, orthologous to OsCKX4, TaCKX4 was not tested due to its weak expression in roots. However, the up-regulated expression of highly specific in seedling roots TaCKX3 and TaCKX8 [8,58] was antagonistically regulated by TaNAC2-5A in these organs, positively influencing seedling growth. Furthermore, our in silico analysis of TaNACs with TaIPTs and TaCKXs showed that the same NAC proteins might join promotor sites of cytokinin synthesis and cytokinin degradation genes in wheat (Iqbal et al., not published yet).

4. Materials and Methods

4.1. Plant Material

Five common wheat breeding lines and cultivars (Triticum aestivum L.), named S12B, S6C, D16, KOH7, and D19, which showed differences in the expression levels of TaCKX GFMs and TaNAC2-5A (NAC2) in 7 DAP spikes, seedling roots, and yield-related traits were selected for the study. They were used in four reciprocal crosses: (1) S12B × S6C and S6C × S12B (C1 and C2, respectively); (2) D16 × KOH7 and KOH7 × D16 (C3 and C4, respectively); (3) D19 × D16 and D16 × D19 (C5 and C6, respectively); and (4) D19 × KOH7 and KOH7 × D19 (C7 and C8, respectively) to obtain the F₁ and F₂ generations. The experimental tissue samples were collected from the parental lines and their F₂ progeny growing in a growth chamber during the same period.

Ten seeds of each genotype germinated in Petri dishes for five days at room temperature in the dark. Six out of ten seedlings from each Petri dish were replanted in pots with soil. The plants were grown in a growth chamber under controlled environmental conditions with 20 °C day/18 °C night temperatures and a 16 h light/8 h dark photoperiod. The light intensity was 350 μmol s⁻¹·m⁻². The plants were irrigated three times a week and fertilized once a week with Florovit according to the manufacturer’s instructions.

The following tissue samples in three biological replicates were collected: 5-day-old seedling roots, which were cut 0.5 cm from the root base before replanting in the pots, and first 7 DAP) spikes from the same plants grown in the growth chamber. All of these samples were collected at 9:00 a.m. The collected material was frozen in liquid nitrogen and kept at −80 °C until use.

4.2. Cross-Breeding

The maternal plant was deprived of its own anthers so that it would not self-fertilize, then pollinated by transferring three anthers from the paternal plant for each ovary of the maternal parent plant and placed in an isolator. The seeds were harvested.

4.3. RNA Extraction and cDNA Synthesis

Total RNA from 7 DAP spikes and roots from 5-day-old seedlings was extracted using TRI Reagent (Invitrogen, Lithuana) according to the manufacturer’s protocol. The concentration and purity of the isolated RNA were determined using a NanoDrop spectrophotometer (NanoDrop ND-1000, Thermi Fisher Scientific, Wilmington, DE, USA), and the integrity was checked on 1.5% (w/v) agarose gels. To remove residual DNA, RNA samples were treated with DNase I (Thermo Fisher Scientific, Lithuana). Each time, 1 μg of good quality RNA was used for cDNA synthesis using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific, Lithuana) following the manufacturer’s instructions. The cDNA was diluted 20 times prior to use in the RT-qPCR assays.

4.4. Quantitative RT-qPCR

RT-qPCR assays were performed for 10 genes: TaCKX1, TaCKX2.1, TaCKX2.2.2, TaCKX3, TaCKX5, TaCKX8, TaCKX9, TaCKX10, TaCKX11, and TaNAC2-5A. The sequences of the primers for each gene are shown in Table S2. All real-time reactions were performed on a Rotor-Gene Q (QIAGEN Hilden, Germany) thermal cycler using 1× HOT FIREPol EvaGreen qPCR Mix Plus (Solis BioDyne, Estonia), 0.2 μM of each primer and 4 μL of cDNA in a total volume of 10 μL. Each reaction was carried out in three biological and three technical replicates in the following temperature profile: initial denaturation and polymerase activation of 95 °C–12 min (95 °C–25 s, 62 °C–25 s, 72 °C–25 s) × 45 cycles, 72 °C–5 min, with melting curve at 72–99 °C 5 s per step. The expression of TaCKX genes was calculated according to the two standard curve method using ADP-ribosylation factor (Ref 2) as a normalizer. The relative expression for each TaCKX GFM and TaNAC2-5A was calculated in relation to the control female parents, set as 1.00.

4.5. Analysis of CKX Activity

CKX enzyme activity was performed in the same samples subjected to TaCKX gene expression analysis according to the procedure developed by Frebort et al. [63] and optimized for wheat tissues. The plant material was powdered with liquid nitrogen using a hand mortar and extracted with a 3-fold excess (v/w) of 0.2 M Tris–HCl buffer, pH 8.0, containing 1 mM phenylmethylsulfonyl fluoride (PMSF) and 0.3% Triton X-100 ((St. Louis, MO, USA). Plant samples were incubated in a reaction mixture consisting of 100 mM McIlvaine buffer, 0.25 mM of the electron acceptor dichlorophenolindophenol and 0.1 mM of substrate (N6-isopentenyl adenine). The volume of the enzyme sample used for the assay was adjusted based on the enzyme activity. The incubation temperature was 37 °C for 1–16 h. After incubation, the reaction was stopped by adding 0.3 mL of 40% trichloroacetic acid (TCA) and 0.2 mL of 2% 4-aminophenol (PAF). The product concentration was determined by scanning the absorption spectrum from 230 nm to 550 nm. The total protein concentration was estimated based on the standard curve of bovine serum albumin (BSA) according to the Bradford procedure [64].

4.6. Measurement of Yield-Related Traits

Morphometric measurement of yield-related traits of selected genotypes was performed. The described traits were plant height, spike number, semi-empty spike number, tiller number, spike length, grain yield, grain number, TGW, and 5-day seedling root weight.

4.7. Statistical Analysis

Statistical analysis was performed using Statistica 13 software (StatSoft). The normality of the data distribution was tested using the Shapiro–Wilk test. The significance of the changes was analyzed using ANOVA variance analysis and post hoc tests. The correlation coefficients were determined using parametric correlation matrices (Pearson’s test) or a nonparametric correlation (Spearman’s test).

5. Conclusions

We indicate, for the first time, that the pattern of expression of selected TaCKX GFMs and TaNAC2-5A, and grain yield in wheat, is paternally inherited by the F₂ generation. Pater-origin transmission of gene expression levels sheds new light on the method of parent selection and crossing to obtain high-yielding phenotypes. We also showed which genes cooperate together by upregulation or downregulation and which function in the opposite manner in establishing yield-related traits. This knowledge can be applied to select the desirable phenotype in F₂. For example, a high-yielding paternal parent with downregulated, compared to the maternal parent, expression of TaCKX2.1 and TaCKX11 in 7 DAP spikes and upregulated expression of TaCKX3 and TaCKX8 and downregulated TaNAC2-5A in seedling roots is expected to transmit this pattern of expression to F₂, which will result in a high yield. The main problem is the antagonistic expression patterns of genes for some important yield-related traits, such as grain number, grain yield, and spike number, to TGW, which is the result of the feedback mechanism of the regulation of expression between TaCKX1 and TaCKX2 genes and others. The expression analysis of TaNAC2-5A and the in silico analysis of TaNAC GFMs revealed that the encoded proteins participate in the regulation of transcription of selected TaCKX genes responsible for cytokinin degradation and TaIPT genes responsible for cytokinin biosynthesis. Therefore, TaNACs are important additional regulators of yield-related traits in wheat, which should be taken into consideration in wheat breeding.

6. Patents

Nadolska-Orczyk A., Szala K., Dmochowska-Boguta M., Orczyk W. Wzory ekspresji genów jako nowe markery molekularne produktywności zbóż oraz sposób przekazywania wysokiej produktywności I strategia selekcji wysokoplonujących odmian zbóż. (Patterns of gene expression as new molecular markers of cereal productivity and a way of transfer of high yield and the strategy for selecting high-yielding cereal varieties). Patent application filed with the Polish Patent Office (UP RP) 23 January 2023, nr P.443557.