2.1. Label-Free Shotgun Proteomics Data Analysis
Seedlings of lowland Nipponbare rice plants, which is considered to be an abiotic stress-sensitive genotype [
11], were grown for four weeks followed by exposure to three separate abiotic stresses; drought, salt, and temperature stress treatments. Plants were exposed to three different levels of water deficit by reducing the soil field capacity of watering (FC) to 70%, 50%, and 30%, referred to hereafter as mild, moderate and severe drought. The concentration of sodium chloride in the soil was increased to 50 mM for another set of rice plants, creating salt stress, and the last group of plants was treated with unfavorable temperature conditions by increasing the day-time temperature to 33 °C and decreasing the night-time temperature to 18 °C. Leaf samples from stressed and control plants were taken for proteomics analysis, to characterize the shared and unique proteome responses of the Nipponbare genotype to individual abiotic stresses.
Table 1 shows the number of reproducibly identified proteins and peptides from the analysis of each set of plants including control and all five stress treatments. A non-redundant total of 3997 proteins were identified across all treatments and groups when all six sample data sets were combined. The highest number of reproducibly identified proteins was 1732 in control conditions, while the lowest number was 1013 reproducibly identified proteins in moderate drought stress.
Table 2 shows the number of differentially abundant proteins (DAPs) identified as being statistically significantly changed in abundance for each of the stress treatments. Increasing the drought stress severity from mild to severe increased the number of differentially abundant proteins. However, the increase in DAPs between mild and moderate drought was much greater than that which occurred between moderate and severe drought. Both moderate and severe drought cause more than 20% of the reproducibly identified proteins to be changed in abundance, which represents a greater systemwide disturbance than that seen for the other stress conditions. Details of all proteins and DAPs identified are provided in
Supplementary Data File S1.
2.2. Analyzing the Effect of Divergent Drought Stress Levels on Leaf Proteome Quantification
A total of 1092 proteins were changed in abundance in response to various drought stress levels in Nipponbare plants, but less than 10% of these (106) were changed in abundance in response to all three drought conditions, indicative of a shared stress response (
Figure 1). Moreover, the greatest number of commonly altered proteins was observed between moderate and severe drought (31 proteins increased in abundance and 83 proteins decreased in abundance). The 52% overlap between differentially abundant proteins at these two drought stress levels illustrates that Nipponbare plants sensed and responded to moderate and severe drought stresses quite similarly.
Figure 1 also demonstrates the number of unique and shared differentially abundant proteins seen in diverse drought conditions, with the highest number of distinct proteins increased in abundance in response to severe drought with 51 proteins, while in comparison the greatest number of unique proteins decreased in abundance was 127 proteins altered in response to moderate drought. The pie charts in
Figure 1 show that the total number of differentially abundant proteins identified uniquely in response to a single stress was approximately 10% higher than the total number of differentially abundant proteins identified in more than one condition.
Almost all of the proteins increased or decreased in abundance in more than one drought stress condition were similarly increased or decreased in all stress conditions where they were observed, except for two proteins with very unusual patterns of abundance (
Table 3). C
2H
2-type domain-containing protein (Q8GS72) was decreased in abundance in response to severe drought, but increased in abundance in response to mild drought stress. Putative glycine-rich protein (Q6Z142) was decreased in abundance in moderate drought, but increased in abundance at mild drought stress level. C
2H
2 zinc finger proteins have been shown to participate as a pivotal regulator of Reactive Oxygen Species (ROS) signaling in the signal transduction of water stress [
14], and oxidative stress, in rice [
15]. It has been demonstrated previously that repressing the expression of zinc finger proteins under drought stress was linked to enhancing tolerance to drought stress through a complex regulatory network in plants [
16]. This suggests that Nipponbare plants sensed the severe drought condition as stress and responded to it by decreasing the abundance of C
2H
2-type domain-containing protein.
Glycine-rich RNA-binding proteins (GRPs), such as putative glycine-rich protein (Q6Z142), have been implicated in the responses of plants to environmental stresses including dehydration, and it has been determined that activation of GRP-related genes has a negative impact on plant response to water deficiency stress [
17]. This agrees with our finding that putative glycine-rich protein is increased in case of mild drought stress conditions, but then decreased in abundance under moderate drought stress.
2.3. Shared Functional Proteome Response under Three Drought Stress Levels
Differentially abundant proteins identified under drought stress conditions belonged to many different functional classifications, which provides insights into water deficit proteome response mechanisms. Changes in abundance of proteins in many different functional categories reflects the diversity and versatility of plant responses when exposed to different levels of water deficit.
Figure 2A displays the gene ontology (GO) analysis of the 106 proteins that were changed in abundance under all three drought stress conditions. This reveals that the proteins commonly changed in response to drought stress were mostly related to photosynthesis, translation, and transport functions. Proteins in the photosynthesis functional category were mostly increased in abundance while the majority of proteins in the translation and transport categories were decreased in abundance. Growth was the only functional category in which all shared proteins were decreased in abundance.
The proportional abundance of proteins increased and decreased commonly between combinations of two or three drought stress levels is represented in
Figure 2B. Clearly, the greatest effect of drought stress occurred in proteins increased in abundance in the photosynthesis functional category. In terms of proteins decreased in abundance in response to combinations of stresses, the highest proportion of proteins was those related to the transport function, with proteins in the translation functional category the next most abundant. These results reflect the generalised effects of drought stress on Nipponbare plants with respect to stress level.
2.4. Stressor-Specific Proteome Response under Individual Drought Stress Levels
The results of proteomic studies can reflect different stress-coping strategies depending on the given stress treatment. Analysis of the GO functional categorization of 378 proteins specifically expressed under each drought stress condition is shown in
Figure 3 (152 proteins in severe drought, 149 proteins in moderate drought and 77 proteins in mild drought). Similar to the shared proteins, photosynthesis, transport, and translation were the main biological functions affected by individual drought stresses. Proteins in the growth functional category were all decreased in abundance (
Figure 3A). One such protein that was significantly decreased in abundance only in plants subjected to severe drought stress was putative acyl-CoA dehydrogenase (Q6ZDX3). Increasing acyl-CoA dehydrogenase accumulation under stress treatment has been reported to be a molecular response of plants which enables them to grow more rapidly [
18], so decreasing the protein abundance under severe drought reflects the negative impact on plant growth of severe drought conditions. Except for growth, all six other biological functions responded to severe drought stress level by both increasing and decreasing the abundance of related proteins (
Figure 3B).
Analysis of biological functional categories of proteins changed in response to drought stress, either uniquely or shared between different conditions, indicated that severe drought stress appeared to cause the most obvious changes in the expressed proteome of Nipponbare plants, irrespective of whether proteins decreased or increased in abundance.
2.5. Proteome Quantification Comparison under Various Abiotic Stresses
The specific protein expression patterns observed under different stress treatments reflect the evolution of diverse plant stress tolerance mechanisms. Moreover, despite various stress-independent commonalities when plants are exposed to a particular abiotic stress, stress-specific signatures can also be detected in plants.
Prior to examining the common proteome response when plants were exposed to drought, salt, and temperature stresses, we needed to choose the best candidate for subsequent comparisons from among the different drought stress levels. Considering that the overlapped proteins differentially abundant under all three different drought stress levels gave us a small number of proteins (
Figure 1) for further comparative proteome analysis with salt and temperature treatments, we decided to focus on common proteins expressed under two drought stress conditions. This subset of differentially abundant proteins is sufficient to achieve a similar effect size to the other stresses, and facilitate statistical analyses. Hence, common proteins differentially altered in abundance in both moderate and severe drought stresses (53 proteins increased in abundance and 167 proteins decreased in abundance) were considered as proteins affected by drought stress in Nipponbare plants, and included in subsequent comparative analyses with salt and temperature stress.
Figure 4 shows that approximately 75% of 551 total differentially abundant proteins were stressor-specific proteins and only less than 5% (24 proteins) belonged to shared responses among all abiotic stresses. Approximately half of the proteins changed in abundance by only one stress treatment were those induced in response to salt stress treatment, with 156 proteins decreased and 46 proteins increased in abundance.
2.6. Proteins Induced in Common under Abiotic Stresses
Figure 4 and
Table 4 show that 24 proteins were changed in abundance in response to all three abiotic stresses, with 20 proteins decreased in abundance in response to all three stresses and 3 increased in abundance in response to all three stresses; Epimerase domain-containing protein (Q2QSR7), L-ascorbate peroxidase 8 (Q69SV0), and Phosphoglucomutase (Q9AUQ4). The only protein with changes in abundance occurring in contrasting directions in response to different stresses was Peptidyl-prolyl cis-trans isomerase (Q69WA8), with a similar increase in abundance of 1.99-fold change under both temperature and salt stresses, and significant decrease in abundance under drought conditions (
Table 4). Interestingly, the level of protein abundance for Epimerase domain-containing protein (Q2QSR7) and L-ascorbate peroxidase 8 (Q69SV0) proteins increased further with decreasing the level of soil water, while the trend for phosphoglucomutase, along with that of (Q69WA8), was in the opposite direction, with protein abundance decreasing when the drought stress was increased from moderate to severe. The unusual behavior of these proteins in Nipponbare plants, where they are altered in abundance in subtly different ways in response to different individual abiotic stresses, makes them attractive candidates for future detailed studies.
Among the 20 proteins that were reduced in abundance in response to all three abiotic stresses in all plants, Glycine-rich RNA-binding protein (Q6ASX7) displayed the largest variation in expression between drought stress and other abiotic stresses. Other examples of variation in expression level under different stresses were Fasciclin-like arabinogalactan protein 15 (Q7XIM4) and Calmodulin-binding protein 60 C (Q7XRM0) which decreased in abundance under all abiotic stresses, but with a much greater change in abundance in response to salinity, and Putative 5’-3’ exoribonuclease (Q5N739) which decreased much more under temperature stress in comparison to the other two stresses. Investigating the metabolic activities of each of these proteins could help us to understand the mechanistic reasons for differences in abundance change of each in response to different individual abiotic stress.
A total of 106 proteins were changed in abundance under any combination of two stress conditions (
Figure 4). This included both similar and antithetic reactions of plants to different stresses; 6 of these showed abundance changes in opposite direction between drought, and each of salt or temperature stresses. Specifically, UDP-glucose 6-dehydrogenase 3 (Q9AUV6) was changed in abundance in response to drought and temperature stresses, however, its abundance decreased under drought stress and increased in response to temperature extreme. Notably, five proteins responded to both drought and salt stresses in common but in opposite manner. Two of them were chloroplastic proteins, namely Glutamate-1-semialdehyde 2,1-aminomutase (Q6YZE2) and Uroporphyrinogen decarboxylase 2 (Q10LR9), which increased in abundance under salt stress but decreased in abundance in case of water deficiency. Three other proteins with different cellular localizations showed an opposite pattern by decreasing in abundance in response to salinity and increasing in abundance in response to drought stress, including Peroxiredoxin Q (P0C5D5), actin (Q67G20), and aquaporin PIP2-7 (Q651D5).
2.7. Determining the Molecular Functions of Stress-Specific and Non-Specific Proteins under Individual Abiotic Stresses
Analysis of proteins at the organelle level is a useful approach for understanding cell behavior under abiotic stress conditions. Abiotic stress alters interactions between organelles in plant cells, and this subsequently changes the regulation and secretion of proteins in cellular organelles and compartments. Therefore, to better understand the biological function of stress-specific and shared differentially abundant proteins under unfavourable conditions, we focused on the molecular functions of major subcellular organelles where functions are typically affected under abiotic stress, including the nucleus, cytosol, plasma membrane, cell wall, mitochondria, cytosol, vacuole, chloroplast and endoplasmic reticulum (
Figure 5).
To withstand environmental stresses, plants have evolved interconnected regulatory pathways that enable them to respond and adapt to their environments promptly. Plant stress proteomics is a dynamic discipline aimed at the study of plant acclimation or tolerance regulations and mechanisms exposed to stress. However, the differential expression proteomics approach itself, based on protein identification and quantitation, can sometimes not give sufficient information on protein function since a certain protein can be responsible for diverse functions depending on its subcellular localization.
Proteomic data can also be interrogated to facilitate identification of molecular and cellular mechanisms that are specific to certain abiotic stresses, or shared between two or more abiotic stress treatments. On the other hand, apart from the unique protein dynamic change mediated by different stress conditions, there can be various points of crosstalk between stress signaling pathways. Identification of crosstalk between signaling pathways has been crucial in strengthening our understanding of how plants regulate their responses to a particular stress condition. For example, aquaporin family proteins that are responsible for ion transport in the plasma membrane responded both negatively and positively to stress in this study. Specifically, aquaporin PIP 2-7 protein was increased in abundance under temperature and salt stress but decreased in abundance when coping with drought stress.
Chloroplasts play a key role in plant response to stress treatment, by adjustment of photosynthesis as a biological function. All individual abiotic stresses affected chloroplasts both negatively and positively, either uniquely or in common. For example, L-ascorbate peroxidase 8, which is a chloroplastic protein responsible for oxidative stress response, was reduced in abundance in response to all abiotic stresses. Some proteins related to different molecular mechanisms in chloroplasts showed opposing changes in abundance under drought and salt stresses, such as Peroxiredoxin, Uroporphyrinogen decarboxylase, and Glutamate-1-semialdehyde 2,1-aminomutase.
Abundance of DAPs in the nucleus were mainly decreased in response to abiotic stresses in both shared and stressor-specific protein categories. For instance, nuclear proteins related to major molecular functions such as gene dynamic change, signal transduction, and metabolic process, were decreased in abundance under all abiotic stresses. One example of differential effects of various abiotic stresses on plant cell organelle functions in our study was related to the heat shock protein family (HSPs), which are known as one of the major stress responses mostly involved in response to temperature stress. Although 24.1 kDa heat shock protein, which is localized in the mitochondria, was increased in abundance in response to both temperature and salt stresses, another member of this family, Heat shock 70 kDa protein BIP4, which occurs in the endoplasmic reticulum (ER), was increased in abundance only under salt stress.
The cytosol plays a significant role in plant function under stresses, not only by itself but also through interaction among different subcellular organelles. RNA binding protein and Phosphoglucomutase were two cytosol-localized proteins that were accumulated in common among all abiotic stresses. However, Aldo_ket_red domain-containing protein and Calcium-dependent kinase, which are related to photosynthesis and transport functions, respectively, decreased in abundance commonly under all individual abiotic stresses. Peptidyl-prolyl cis-trans isomerase, which performs a signaling molecular function in the cytosol, was detected as the only protein that increased in abundance under drought stress but decreased in abundance under both salt and temperature stress treatments.
2.8. Parallel Reaction Monitoring (PRM) Validation
A series of PRM experiments were performed to validate the results from label-free shotgun proteomics analysis of Nipponbare plants under various individual abiotic stresses. The PRM results indicated that the differential changes in abundance of four selected proteins measured by PRM, as shown in
Figure 6, agreed with the label-free shotgun proteomics results.
These included Betaine aldehyde dehydrogenase (O24174), Fasciclin-like arabinogalactan protein 15 (Q7XIM4), t-SNARE coiled-coil homology domain-containing (Q8H5R6), and Anamorsin homolog 1 (Q7XQ97). Betaine aldehyde dehydrogenase protein was significantly reduced in abundance under drought and temperature stresses, and was reduced in expression slightly after salt treatment but not to a statistically significant level (
Figure 6A). The Fasciclin-like arabinogalactan protein 15 (
Figure 6B) and t-SNARE coiled-coil homology domain-containing protein (
Figure 6C) both responded similarly to all abiotic stresses by decreasing in abundance significantly. Anamorsin homolog 1 responded to drought stresses by decreasing in abundance significantly, while under salt and temperature stress it was reduced in abundance, but not to a statistically significant level (
Figure 6D).
Table 5 shows for comparison purposes the reported fold-change of each of the four proteins from the label free quantitative proteomics data analysis.