Abstract

   Stress events have transgenerational effects on plant growth and
   development. In Mediterranean regions, water-deficit and heat (WH)
   stress is a frequent issue that negatively affects crop yield and
   quality. Nitrogen (N) is an essential plant macronutrient and often a
   yield-limiting factor for crops. Here, the response of durum wheat
   seedlings to N starvation under the transgenerational effects of WH
   stress was investigated in two genotypes. Both genotypes showed a
   significant reduction in seedling height, leaf number, shoot and root
   weight (fresh and dry), primary root length, and chlorophyll content
   under N starvation stress. However, in the WH stress-tolerant genotype,
   the percentage reduction of most traits was lower in progeny from the
   stressed parents than progeny from the control parents. Small RNA
   sequencing identified 1534 microRNAs in different treatment groups.
   Differentially expressed microRNAs (DEMs) were characterized subject to
   N starvation, parental stress and genotype factors, with their target
   genes identified in silico. GO and KEGG enrichment analyses revealed
   the biological functions, associated with DEM-target modules in stress
   adaptation processes, that could contribute to the phenotypic
   differences observed between the two genotypes. The study provides the
   first evidence of the transgenerational effects of WH stress on the N
   starvation response in durum wheat.

   Keywords: nitrogen starvation, water-deficit and heat stress,
   transgenerational effects, cross stress tolerance, microRNAs, crop
   improvement

1. Introduction

   Environmental stresses such as water deficiency, extreme temperatures,
   and soil nutrient deficiency present significant challenges to the
   development and production of crops. Durum wheat (Triticum turgidum L.
   ssp. durum) is a tetraploid wheat species (2n = 4x = 28, AABB) mainly
   grown in the Mediterranean basin, North America and the Australian
   wheat belt [[28]1,[29]2]. Compared with hexaploid wheat (bread wheat),
   durum wheat has higher grain protein content, strong yellow
   pigmentation, harder kernels, and a unique nutty flavor [[30]3,[31]4].
   With its significant agronomic value, excellent grain quality, and
   versatile end use, durum wheat is considered as a staple crop in
   Mediterranean regions.

   Grown under rain-fed conditions, durum wheat is often exposed to
   frequent episodes of water-deficit and heat stress [[32]5,[33]6]. In
   the field, high temperatures start to occur while soil water supply
   declines gradually during reproductive stages (e.g., flowering and
   grain filling) [[34]4,[35]7]. A significant number of studies have
   investigated the impact of the independent and combined effects of
   water-deficit and heat stress on plant growth, grain productivity, and
   grain quality in wheat [[36]4,[37]5,[38]6,[39]7,[40]8,[41]9].
   Water-deficit and heat stress have significant impacts on
   photosynthetic activities, transpiration efficiency, cellular osmotic
   homeostasis, nutrient uptake, and metabolite production
   [[42]1,[43]4,[44]8,[45]10]. Such changes affect the reproductive
   processes in wheat, ultimately leading to changes in yield components
   (e.g., grain number, spikelet number, grain weight) and grain quality
   traits (e.g., protein content, starch content, antioxidant levels)
   [[46]1,[47]4,[48]7,[49]10].

   Recent evidence suggests that water-deficit and heat stress could
   affect the stress response systems in the following generation through
   changes at the physiological, phenotypical and epigenetic level
   [[50]11,[51]12,[52]13,[53]14,[54]15,[55]16,[56]17]. A few studies have
   demonstrated the adaptive value of the transgenerational influence of
   the same stress type in the offspring [[57]13]. More interestingly,
   stress priming of one abiotic stress could have a beneficial impact on
   the occurrence of a different stress through synergistic stress
   signaling pathways [[58]18,[59]19,[60]20,[61]21]. For example, terminal
   drought stress applied in bread wheat from the reproductive stage until
   maturity improved the tolerance against salt stress in the next
   generation, mainly through changes in osmolyte accumulation, water
   relations modulation and lipid peroxidation [[62]21]. As crops grown
   under field conditions are often exposed to multiple stressors
   (simultaneously, sequentially, or across multiple generations),
   investigation towards such phenomena (cross-stress tolerance or
   cross-stress effects) would be very beneficial for providing new
   strategies in crop breeding practices. However, it remains unknown how
   parental water-deficit and heat stress affect progeny performance under
   a different stress (e.g., nitrogen stress) in durum wheat.

   Soil N availability has been a major limiting factor in wheat
   production. Nitrogen deficiency negatively affects the grain yield as
   well as grain quality in cereal crops through its impact on the
   nutrient uptake, photosynthesis rate, respiration efficiency, and
   enzyme activities
   [[63]22,[64]23,[65]24,[66]25,[67]26,[68]27,[69]28,[70]29]. N-stressed
   crop plants often have chlorotic leaves, less fertile tillers, shorter
   plant height and slow growth [[71]27,[72]28,[73]29]. Changes in root
   architecture, root length, and root biomass are also known
   morphological responses to N starvation [[74]28,[75]30]. Several
   studies in wheat have investigated the molecular networks controlling
   the N stress response and N use efficiency through high-throughput
   approaches [[76]28,[77]31,[78]32]. Specifically, a transcriptomic study
   in durum wheat has identified 4626 differentially expressed genes
   (DEGs) in response to N starvation at the grain filling stage [[79]31].
   The majority of the DEGs were nitrate or ammonium transporters,
   transcription factors, protein kinases and other genes involved in N
   assimilation. Furthermore, stress-responsive microRNAs (miRNAs) have
   also gained increasing attention for their essential roles in
   regulating plant adaptive responses to nutrient deprivation
   [[80]23,[81]25,[82]33,[83]34,[84]35].

   As an essential type of epigenetic regulator, miRNAs fine-tune the
   expression of their protein-coding target genes through
   post-transcriptional gene silencing
   [[85]36,[86]37,[87]38,[88]39,[89]40,[90]41,[91]42]. In crops, miRNAs
   can rapidly respond to various environmental and developmental cues,
   playing important roles in plant growth, reproductive development and
   stress adaptation [[92]36,[93]43,[94]44,[95]45,[96]46,[97]47]. In
   particular, studies in durum wheat have discovered a significant number
   of miRNAs that play central roles in water-deficit stress, heat stress,
   and N-stress response networks
   [[98]9,[99]12,[100]23,[101]25,[102]48,[103]49,[104]50,[105]51,[106]52,[
   107]53,[108]54]. Our previous research had shown that the parental
   water-deficit stress had a significant impact on the durum miRNA
   transcriptome in the progeny, contributing to the differences in stress
   response and crop performance when the next generation was exposed to
   water-deficit stress [[109]11]. We have also demonstrated that the
   combination of parental water-deficit stress and heat stress
   significantly affected progeny germination traits and seedling vigor
   through changes in the miRNAome [[110]12]. However, it is unknown how
   the miRNA-regulated N stress response networks are affected by the
   transgenerational effects of water-deficit and heat stress.

   In this study, we characterized the morphological and physiological
   changes of durum wheat seedlings in response to N starvation under the
   transgenerational effects of water-deficit stress and heat (WH) stress
   in a WH stress-tolerant and WH stress-sensitive genotype. Using the
   small RNA sequencing approach, a systematic analysis of the miRNA
   expression profile subject to the progeny treatment factor, parental
   treatment factor, and genotype factor was performed on a genome-wide
   scale. To our knowledge, this is the first description of the
   transgenerational cross-stress effects in durum wheat. The results
   provide new insights to researchers and breeding programs addressing
   stress-tolerance improvement in cereal crops.

2. Results

2.1. Seedling Performance of Two Durum Wheat Genotypes under the Effects of
Parental Water-Deficit and Heat Stress and Progeny N Starvation Stress

   Two Australian durum wheat genotypes were used in this study. DBA
   Aurora is tolerant to water-deficit and heat (WH) stress and L6 (a
   University of Adelaide breeding line) is sensitive to WH stress
   [[111]4]. To study the transgenerational effects of parental stress
   treatment, seeds of the two genotypes were collected from control (CG)
   and WH-stressed parents in a previous experiment [[112]4]. There were
   four seed groups: AuCG (seeds from DBA Aurora parents treated with the
   control condition), AuWH (seeds from DBA Aurora parents treated with
   water-deficit and heat stress), L6CG (seeds from L6 parents treated
   with the control condition) and L6WH (seeds from L6 parents treated
   with water-deficit and heat stress). To study the response of progeny
   to N starvation stress, two treatment groups—control (C) and N
   starvation (N)—were set up for each seed group. Therefore, the current
   study had eight treatment groups in total: AuCG_C (DBA Aurora control
   parents, progeny treated with the control), AuCG_N (DBA Aurora control
   parents, progeny treated with N starvation), AuWH_C (DBA Aurora
   water-deficit and heat stress parents, progeny treated with the
   control), AuWH_N (DBA Aurora water-deficit and heat stress parents,
   progeny treated with N starvation), L6CG_C (L6 control parents, progeny
   treated with the control), L6CG_N (L6 control parents, progeny treated
   with N starvation), L6WH_C (L6 water-deficit and heat stress parents,
   progeny treated with the control), and L6WH_N (L6 water-deficit and
   heat stress parents, progeny treated with N starvation).

   To evaluate seedling performance, eight morphological and physiological
   traits were measured at the three-week stage: seedling height, leaf
   number, shoot fresh weight, shoot dry weight, root fresh weight, root
   dry weight, primary root length, and chlorophyll content ([113]Table 1
   and [114]Table 2). The growth and development of durum wheat seedlings
   were significantly affected by N deficiency. For progeny groups with
   the same parental origin, all eight traits showed a significant
   reduction under N starvation stress when compared with their control
   (i.e., AuCG_N vs. AuCG_C, AuWH_N vs. AuWH_C, L6CG_N vs. L6CG_C, and
   L6WH_N vs. L6WH_C). However, the % reduction of each trait varied
   across progeny groups in the two genotypes. For example, for plant
   height, progeny from the WH parents appeared to have a lower percentage
   reduction in response to N starvation when compared with the progeny
   from the control parents in both genotypes. In DBA Aurora, the
   percentage reduction of plant height was 32.8% between AuWH_N vs.
   AuWH_C, while the percentage reduction was 34.5% between AuCG_N vs.
   AuCG_C ([115]Table 1). In L6, the percentage reduction of plant height
   was 34.3% between L6WH_N vs. L6WH_C, while the percentage reduction was
   35.9% between L6CG_N vs. 6CG_C ([116]Table 2). A similar pattern was
   also observed for the primary root length (11.3% (AuWH_N vs. AuWH_C)
   and 13.8% (AuCG_N vs. AuCG_C) in DBA Aurora; 13.8% (L6WH_N vs. L6WH_C)
   and 14.2% (L6CG_N vs. 6CG_C) in L6).

Table 1.

   Seedling performance traits measured in DBA Aurora. The treatment
   groups are: AuCG_C (DBA Aurora control parents, progeny treated with
   the control), AuCG_N (DBA Aurora control parents, progeny treated with
   N starvation), AuWH_C (DBA Aurora water-deficit and heat stress
   parents, progeny treated with the control), AuWH_N (DBA Aurora
   water-deficit and heat stress parents, progeny treated with N
   starvation). Results shown as mean ± SE (n = 6).
   Treatment Group Seedling Height (cm) Leaf
   Number Shoot Fresh Weight (g) Shoot Dry Weight (g) Root Fresh Weight
   (g) Root Dry Weight (g) Primary Root Length (cm) Chlorophyll Content
   (SPAD Units)
   AuCG_C 36.92 ± 0.77 5.42 ± 0.15 1.658 ± 0.058 0.211 ± 0.009 1.327 ±
   0.026 0.155 ± 0.004 27.62 ± 0.46 48.33 ± 0.70
   AuCG_N 24.17 ± 0.42 3.17 ± 0.25 0.605 ± 0.022 0.087 ± 0.003 0.779 ±
   0.018 0.094 ± 0.001 23.80 ± 0.34 38.38 ± 0.57
   % Reduction 34.5% 41.5% 63.5% 60.7% 41.3% 39.3% 13.8% 20.6%
   AuWH_C 35.93 ± 0.53 5.33 ± 0.17 1.630 ± 0.060 0.214 ± 0.006 1.328 ±
   0.027 0.154 ± 0.003 28.77 ± 0.34 49.18 ± 0.55
   AuWH_N 24.13 ± 0.50 3.25 ± 0.17 0.640 ± 0.019 0.089 ± 0.001 0.804 ±
   0.023 0.097 ± 0.002 25.52 ± 0.41 40.28 ± 0.55
   % Reduction 32.8% 39.1% 60.7% 58.4% 39.4% 37.1% 11.3% 18.1%
   F pr. Parent treatment 0.381 1.000 0.940 0.667 0.584 0.876 0.001 0.032
   F pr. Progeny treatment <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
   <0.001 <0.001
   F pr. Parent × Progeny treatment 0.413 0.663 0.481 0.412 0.613 0.447
   0.476 0.388
   l.s.d Parent treatment n.a ^1 n.a n.a n.a n.a n.a 0.813 1.242
   l.s.d Progeny treatment 1.184 0.393 0.092 0.011 0.049 0.005 0.813 1.242
   l.s.d Parent × Progeny treatment n.a n.a n.a n.a n.a n.a n.a n.a
   [117]Open in a new tab

   ^1 n.a, not applicable.

Table 2.

   Seedling performance traits measured in L6. The treatment groups are:
   L6CG_C (L6 control parents, progeny treated with the control), L6CG_N
   (L6 control parents, progeny treated with N starvation), L6WH_C (L6
   water-deficit and heat stress parents, progeny treated with the
   control), L6WH_N (L6 water-deficit and heat stress parents, progeny
   treated with N starvation). Results shown as mean ± SE (n = 6).
   Treatment Group Seedling Height (cm) Leaf
   Number Shoot Fresh Weight (g) Shoot Dry Weight (g) Root Fresh Weight
   (g) Root Dry Weight (g) Primary Root Length (cm) Chlorophyll Content
   (SPAD Units)
   L6CG_C 35.47 ± 0.45 5.00 ± 0.22 1.556 ± 0.040 0.205 ± 0.005 1.313 ±
   0.024 0.152 ± 0.002 27.13 ± 0.35 47.33 ± 0.73
   L6CG_N 22.73 ± 0.57 2.83 ± 0.17 0.562 ± 0.010 0.076 ± 0.002 0.755 ±
   0.019 0.090 ± 0.002 23.28 ± 0.37 35.97 ± 0.60
   % Reduction 35.9% 43.3% 63.9% 63.1% 42.5% 40.9% 14.2% 24.0%
   L6WH_C 34.42 ± 0.57 5.00 ± 0.18 1.502 ± 0.046 0.199 ± 0.006 1.303 ±
   0.021 0.152 ± 0.003 27.83 ± 0.31 46.28 ± 0.56
   L6WH_N 22.60 ± 0.63 2.75 ± 0.11 0.517 ± 0.011 0.072 ± 0.001 0.735 ±
   0.022 0.088 ± 0.002 24.00 ± 0.31 34.90 ± 0.62
   % Reduction 34.3% 45.0% 65.6% 63.9% 43.6% 41.7% 13.8% 24.6%
   F pr. Parent treatment 0.301 0.815 0.133 0.211 0.507 0.704 0.047 0.108
   F pr. Progeny treatment <0.001 <0.001 <0.001 <0.001 <0.001 <0.001
   <0.001 <0.001
   F pr. Parent × Progeny treatment 0.421 0.815 0.893 0.759 0.826 0.799
   0.980 0.990
   F pr. Parent treatment n.a ^1 n.a n.a n.a n.a n.a 0.697 n.a
   F pr. Progeny treatment 1.164 0.367 0.066 0.009 0.045 0.005 0.697 1.313
   F pr. Parent × Progeny treatment n.a n.a n.a n.a n.a n.a n.a n.a
   [118]Open in a new tab

   ^1 n.a, not applicable.

   For the other six traits (leaf number, shoot fresh weight, shoot dry
   weight, root fresh weight, root dry weight, primary root length, and
   chlorophyll content), a genotype-dependent pattern can be observed when
   it comes to the transgenerational effects of parental treatment. For
   DBA Aurora, the WH stress-tolerant variety, the parental exposure of WH
   helped to mitigate the negative effects of N starvation in the progeny
   (lower percentage reduction for the traits measured). For example, in
   DBA Aurora, the percentage reduction of chlorophyll content in progeny
   groups from the AuCG parents was 20.6%, while in the progeny from the
   AuWH parents the percentage reduction was 18.1% ([119]Table 1). In
   contrast, for the WH-sensitive genotype L6, the parental exposure of WH
   exacerbated the negative impacts of N starvation. For example, for
   shoot fresh weight, the percentage reduction in progeny from the L6CG
   parents was 63.9%, while in progeny groups from the L6WH parents, the
   reduction rate was 65.6% ([120]Table 2).

2.2. Durum Wheat MiRNA Expression Profile across Different Treatment Groups

   To investigate how miRNAs are involved in the N starvation response
   under the effects of transgenerational WH stress, eight sRNA libraries
   were constructed and sequenced from the eight treatment groups (AuCG_C,
   AuCG_N, AuWH_C, AuWH_N, L6CG_C, L6CG_N, L6WH_C, L6WH_N). In total, over
   143 million raw reads were generated with over 51 million reads being
   unique ([121]Table S1). After filtering and data processing, a total of
   102.05 million clear sRNA reads were obtained, of which over 45 million
   were unique sRNA reads ([122]Table S1). Through the bioinformatics
   pipeline, a total of 1534 miRNA-MIR entries (different combinations of
   mature miRNA products and MIR origins in the genome) were identified,
   with 190 being novel miRNAs ([123]Table S2). The identified miRNAs were
   grouped into five categories (group 1 to 5). The definition and
   selection criteria for each group were described previously [[124]12].
   Briefly, groups 1 to 4 contained different types of conserved miRNAs,
   and group 5 included all the novel miRNAs identified in this study. The
   conserved miRNAs belonged to 61 MIR families ([125]Table S2). Group 2
   (definition: sRNA reads can be mapped to miRNA references in the