Multi-omics insights into the positive role of strigolactone perception in barley drought response.

BACKGROUND: Drought is a major environmental stress that affects crop productivity worldwide. Although previous research demonstrated links between strigolactones (SLs) and drought, here we used barley (Hordeum vulgare) SL-insensitive mutant hvd14 (dwarf14) to scrutinize the SL-dependent mechanisms associated with water deficit response. RESULTS: We have employed a combination of transcriptomics, proteomics, phytohormonomics analyses, and physiological data to unravel differences between wild-type and hvd14 plants under drought. Our research revealed that drought sensitivity of hvd14 is related to weaker induction of abscisic acid-responsive genes/proteins, lower jasmonic acid content, higher reactive oxygen species content, and lower wax biosynthetic and deposition mechanisms than wild-type plants. In addition, we identified a set of transcription factors (TFs) that are exclusively drought-induced in the wild-type barley. CONCLUSIONS: Critically, we resolved a comprehensive series of interactions between the drought-induced barley transcriptome and proteome responses, allowing us to understand the profound effects of SLs in alleviating water-limiting conditions. Several new avenues have opened for developing barley more resilient to drought through the information provided. Moreover, our study contributes to a better understanding of the complex interplay between genes, proteins, and hormones in response to drought, and underscores the importance of a multidisciplinary approach to studying plant stress response mechanisms.

Persisting Effects in Daphnia magna Following an Acute Exposure to Flowback and Produced Waters from the Montney Formation.

Hydraulic fracturing extracts oil and gas through the injection of water and proppants into subterranean formations. These injected fluids mix with the host rock formation and return to the surface as a complex wastewater containing salts, metals, and organic compounds, termed flowback and produced water (FPW). Previous research indicates that FPW is toxic to Daphnia magna (D. magna), impairing reproduction, molting, and maturation time; however, recovery from FPW has not been extensively studied. Species unable to recover have drastic impacts on populations on the ecological scale; thus, this study sought to understand if recovery from an acute 48 h FPW exposure was possible in the freshwater invertebrate, D. magna by using a combination of physiological and molecular analyses. FPW (0.75%) reduced reproduction by 30% and survivorship to 32% compared to controls. System-level quantitative proteomic analyses demonstrate extensive perturbation of metabolism and protein transport in both 0.25 and 0.75% FPW treatments after a 48 h FPW exposure. Collectively, our data indicate that D. magna are unable to recover from acute 48 h exposures to ≥0.25% FPW, as evidence of toxicity persists for at least 19 days post-exposure. This study highlights the importance of considering persisting effects following FPW remediation when modeling potential spill scenarios.

DomainViz: intuitive visualization of consensus domain distributions across groups of proteins.

The prediction of functional domains is typically among the first steps towards understanding the function of new proteins and protein families. There are numerous databases of annotated protein domains that permit researchers to identify domains on individual proteins of interest. However, it is necessary to perform high-throughput domain searches to gain evolutionary insight into the functions of proteins and protein families. Unfortunately, at present, it is difficult to search for, and visualize domain conservation across multiple proteins and/or multiple groups of proteins in an intuitive manner. Here we present DomainViz, a new web-server that streamlines the identification and visualization of domains across multiple protein sequences. Currently, DomainViz uses the well-established PFAM and Prosite databases for domain searching and assembles intuitive, publication-ready 'monument valley' plots (mv-plots) that display the extent of domain conservation along two dimensions: positionality and frequency of occurrence in the input protein sequences. In addition, DomainViz produces a conventional domain-ordering figure. DomainViz can be used to explore the conservation of domains within a single protein family, across multiple families, and across families from different species to support studies into protein function and evolution. The web-server is publicly available at: https://uhrigprotools.biology.ualberta.ca/domainviz.

Closing the protein gap in plant chronobiology.

Our modern understanding of diel cell regulation in plants stems from foundational work in the late 1990s that analysed the dynamics of selected genes and mutants in Arabidopsis thaliana. The subsequent rise of transcriptomics technologies such as microarrays and RNA sequencing has substantially increased our understanding of anticipatory (circadian) and reactive (light- or dark-triggered) diel events in plants. However, it is also becoming clear that gene expression data fail to capture critical events in diel regulation that can only be explained by studying protein-level dynamics. Over the past decade, mass spectrometry technologies and quantitative proteomic workflows have significantly advanced, finally allowing scientists to characterise diel protein regulation at high throughput. Initial proteomic investigations suggest that the diel transcriptome and proteome generally lack synchrony and that the timing of daily regulatory events in plants is impacted by multiple levels of protein regulation (e.g., post-translational modifications [PTMs] and protein-protein interactions [PPIs]). Here, we highlight and summarise how the use of quantitative proteomics to elucidate diel plant cell regulation has advanced our understanding of these processes. We argue that this new understanding, coupled with the extraordinary developments in mass spectrometry technologies, demands greater focus on protein-level regulation of, and by, the circadian clock. This includes hitherto unexplored diel dynamics of protein turnover, PTMs, protein subcellular localisation and PPIs that can be masked by simple transcript- and protein-level changes. Finally, we propose new directions for how the latest advancements in quantitative proteomics can be utilised to answer outstanding questions in plant chronobiology.

Phosphate and phosphite have a differential impact on the proteome and phosphoproteome of Arabidopsis suspension cell cultures.

Phosphorus absorbed in the form of phosphate (H2 PO4- ) is an essential but limiting macronutrient for plant growth and agricultural productivity. A comprehensive understanding of how plants respond to phosphate starvation is essential for the development of more phosphate-efficient crops. Here we employed label-free proteomics and phosphoproteomics to quantify protein-level responses to 48 h of phosphate versus phosphite (H2 PO3- ) resupply to phosphate-deprived Arabidopsis thaliana suspension cells. Phosphite is similarly sensed, taken up and transported by plant cells as phosphate, but cannot be metabolized or used as a nutrient. Phosphite is thus a useful tool for differentiating between non-specific processes related to phosphate sensing and transport and specific responses to phosphorus nutrition. We found that responses to phosphate versus phosphite resupply occurred mainly at the level of protein phosphorylation, complemented by limited changes in protein abundance, primarily in protein translation, phosphate transport and scavenging, and central metabolism proteins. Altered phosphorylation of proteins involved in core processes such as translation, RNA splicing and kinase signaling was especially important. We also found differential phosphorylation in response to phosphate and phosphite in 69 proteins, including splicing factors, translation factors, the PHT1;4 phosphate transporter and the HAT1 histone acetyltransferase - potential phospho-switches signaling changes in phosphorus nutrition. Our study illuminates several new aspects of the phosphate starvation response and identifies important targets for further investigation and potential crop improvement.

BoxCar and Library-Free Data-Independent Acquisition Substantially Improve the Depth, Range, and Completeness of Label-Free Quantitative Proteomics.

Data-dependent acquisition (DDA) methods are the current standard for quantitative proteomics in many biological systems. However, DDA preferentially measures highly abundant proteins and generates data that is plagued with missing values, requiring extensive imputation. Here, we demonstrate that library-free BoxCarDIA acquisition, combining MS1-level BoxCar acquisition with MS2-level data-independent acquisition (DIA) analysis, outperforms conventional DDA and other library-free DIA (directDIA) approaches. Using a combination of low- (HeLa cells) and high- (Arabidopsis thaliana cell culture) dynamic range sample types, we demonstrate that BoxCarDIA can achieve a 40% increase in protein quantification over DDA without offline fractionation or an increase in mass-spectrometer acquisition time. Further, we provide empirical evidence for substantial gains in dynamic range sampling that translates to deeper quantification of low-abundance protein classes under-represented in DDA and directDIA data. Unlike both DDA and directDIA, our new BoxCarDIA method does not require full MS1 scans while offering reproducible protein quantification between replicate injections and providing more robust biological inferences. Overall, our results advance the BoxCarDIA technique and establish it as the new method of choice for label-free quantitative proteomics across diverse sample types.

Diurnal dynamics of the Arabidopsis rosette proteome and phosphoproteome.

Plant growth depends on the diurnal regulation of cellular processes, but it is not well understood if and how transcriptional regulation controls diurnal fluctuations at the protein level. Here, we report a high-resolution Arabidopsis thaliana (Arabidopsis) leaf rosette proteome acquired over a 12 hr light:12 hr dark diurnal cycle and the phosphoproteome immediately before and after the light-to-dark and dark-to-light transitions. We quantified nearly 5,000 proteins and 800 phosphoproteins, of which 288 fluctuated in their abundance and 226 fluctuated in their phosphorylation status. Of the phosphoproteins, 60% were quantified for changes in protein abundance. This revealed six proteins involved in nitrogen and hormone metabolism that had concurrent changes in both protein abundance and phosphorylation status. The diurnal proteome and phosphoproteome changes involve proteins in key cellular processes, including protein translation, light perception, photosynthesis, metabolism and transport. The phosphoproteome at the light-dark transitions revealed the dynamics at phosphorylation sites in either anticipation of or response to a change in light regime. Phosphorylation site motif analyses implicate casein kinase II and calcium/calmodulin-dependent kinases among the primary light-dark transition kinases. The comparative analysis of the diurnal proteome and diurnal and circadian transcriptome established how mRNA and protein accumulation intersect in leaves during the diurnal cycle of the plant.

Diurnal changes in concerted plant protein phosphorylation and acetylation in Arabidopsis organs and seedlings.

Protein phosphorylation and acetylation are the two most abundant post-translational modifications (PTMs) that regulate protein functions in eukaryotes. In plants, these PTMs have been investigated individually; however, their co-occurrence and dynamics on proteins is currently unknown. Using Arabidopsis thaliana, we quantified changes in protein phosphorylation, acetylation and protein abundance in leaf rosettes, roots, flowers, siliques and seedlings at the end of day (ED) and at the end of night (EN). This identified 2549 phosphorylated and 909 acetylated proteins, of which 1724 phosphorylated and 536 acetylated proteins were also quantified for changes in PTM abundance between ED and EN. Using a sequential dual-PTM workflow, we identified significant PTM changes and intersections in these organs and plant developmental stages. In particular, cellular process-, pathway- and protein-level analyses reveal that the phosphoproteome and acetylome predominantly intersect at the pathway- and cellular process-level at ED versus EN. We found 134 proteins involved in core plant cell processes, such as light harvesting and photosynthesis, translation, metabolism and cellular transport, that were both phosphorylated and acetylated. Our results establish connections between PTM motifs, PTM catalyzing enzymes and putative substrate networks. We also identified PTM motifs for further characterization of the regulatory mechanisms that control cellular processes during the diurnal cycle in different Arabidopsis organs and seedlings. The sequential dual-PTM analysis expands our understanding of diurnal plant cell regulation by PTMs and provides a useful resource for future analyses, while emphasizing the importance of analyzing multiple PTMs simultaneously to elucidate when, where and how they are involved in plant cell regulation.

Activation of Mitochondrial Protein Phosphatase SLP2 by MIA40 Regulates Seed Germination.

Reversible protein phosphorylation catalyzed by protein kinases and phosphatases represents the most prolific and well-characterized posttranslational modification known. Here, we demonstrate that Arabidopsis (Arabidopsis thaliana) Shewanella-like protein phosphatase 2 (AtSLP2) is a bona fide Ser/Thr protein phosphatase that is targeted to the mitochondrial intermembrane space (IMS) where it interacts with the mitochondrial oxidoreductase import and assembly protein 40 (AtMIA40), forming a protein complex. Interaction with AtMIA40 is necessary for the phosphatase activity of AtSLP2 and is dependent on the formation of disulfide bridges on AtSLP2. Furthermore, by utilizing atslp2 null mutant, AtSLP2 complemented and AtSLP2 overexpressing plants, we identify a function for the AtSLP2-AtMIA40 complex in negatively regulating gibberellic acid-related processes during seed germination. Results presented here characterize a mitochondrial IMS-localized protein phosphatase identified in photosynthetic eukaryotes as well as a protein phosphatase target of the highly conserved eukaryotic MIA40 IMS oxidoreductase.

Rhizobiales-like Phosphatase 2 from Arabidopsis thaliana Is a Novel Phospho-tyrosine-specific Phospho-protein Phosphatase (PPP) Family Protein Phosphatase.

Cellular signaling through protein tyrosine phosphorylation is well established in mammalian cells. Although lacking the classic tyrosine kinases present in humans, plants have a tyrosine phospho-proteome that rivals human cells. Here we report a novel plant tyrosine phosphatase from Arabidopsis thaliana (AtRLPH2) that, surprisingly, has the sequence hallmarks of a phospho-serine/threonine phosphatase belonging to the PPP family. Rhizobiales/Rhodobacterales/Rhodospirillaceae-like phosphatases (RLPHs) are conserved in plants and several other eukaryotes, but not in animals. We demonstrate that AtRLPH2 is localized to the plant cell cytosol, is resistant to the classic serine/threonine phosphatase inhibitors okadaic acid and microcystin, but is inhibited by the tyrosine phosphatase inhibitor orthovanadate and is particularly sensitive to inhibition by the adenylates, ATP and ADP. AtRLPH2 displays remarkable selectivity toward tyrosine-phosphorylated peptides versus serine/threonine phospho-peptides and readily dephosphorylates a classic tyrosine phosphatase protein substrate, suggesting that in vivo it is a tyrosine phosphatase. To date, only one other tyrosine phosphatase is known in plants; thus AtRLPH2 represents one of the missing pieces in the plant tyrosine phosphatase repertoire and supports the concept of protein tyrosine phosphorylation as a key regulatory event in plants.

Genome-scale analysis of regulatory protein acetylation enzymes from photosynthetic eukaryotes.

BACKGROUND: Reversible protein acetylation occurring on Lys-Ne has emerged as a key regulatory post-translational modification in eukaryotes. It is mediated by two groups of enzymes: lysine acetyltransferases (KATs) and lysine deacetylases (KDACs) that catalyze the addition and removal of acetyl groups from target proteins. Estimates indicate that protein acetylation is second to protein phosphorylation in abundance, with thousands of acetylated sites now identified in different subcellular compartments. Considering the important regulatory role of protein phosphorylation, elucidating the diversity of KATs and KDACs across photosynthetic eukaryotes is essential in furthering our understanding of the impact of reversible protein acetylation on plant cell processes. RESULTS: We report a genome-scale analysis of lysine acetyltransferase (KAT)- and lysine deacetylase (KDAC)-families from 53 photosynthetic eukaryotes. KAT and KDAC orthologs were identified in sequenced genomes ranging from glaucophytes and algae to land plants and then analyzed for evolutionary relationships. Based on consensus molecular phylogenetic and subcellular localization data we found new sub-classes of enzymes in established KAT- and KDAC-families. Specifically, we identified a non-photosynthetic origin of the HD-tuin family KDACs, a new monocot-specific Class I HDA-family sub-class, and a phylogenetically distinct Class II algal/heterokont sub-class which maintains an ankyrin domain not conserved in land plant Class II KDACs. Protein structure analysis showed that HDA- and SRT-KDACs exist as bare catalytic subunits with highly conserved median protein length, while all KATs maintained auxiliary domains, with CBP- and TAFII250-KATs displaying protein domain gain and loss over the course of photosynthetic eukaryote evolution in addition to variable protein length. Lastly, promoter element enrichment analyses across species revealed conserved cis-regulatory sequences that support KAT and KDAC involvement in the regulation of plant development, cold/drought stress response, as well as cellular processes such as the circadian clock. CONCLUSIONS: Our results reveal new evolutionary, structural, and biological insights into the KAT- and KDAC-families of photosynthetic eukaryotes, including evolutionary parallels to protein kinases and protein phosphatases. Further, we provide a comprehensive annotation framework through our extensive phylogenetic analysis, from which future research investigating aspects of protein acetylation in plants can use to position new findings in a broader context.

Evolution of bacterial-like phosphoprotein phosphatases in photosynthetic eukaryotes features ancestral mitochondrial or archaeal origin and possible lateral gene transfer.

Protein phosphorylation is a reversible regulatory process catalyzed by the opposing reactions of protein kinases and phosphatases, which are central to the proper functioning of the cell. Dysfunction of members in either the protein kinase or phosphatase family can have wide-ranging deleterious effects in both metazoans and plants alike. Previously, three bacterial-like phosphoprotein phosphatase classes were uncovered in eukaryotes and named according to the bacterial sequences with which they have the greatest similarity: Shewanella-like (SLP), Rhizobiales-like (RLPH), and ApaH-like (ALPH) phosphatases. Utilizing the wealth of data resulting from recently sequenced complete eukaryotic genomes, we conducted database searching by hidden Markov models, multiple sequence alignment, and phylogenetic tree inference with Bayesian and maximum likelihood methods to elucidate the pattern of evolution of eukaryotic bacterial-like phosphoprotein phosphatase sequences, which are predominantly distributed in photosynthetic eukaryotes. We uncovered a pattern of ancestral mitochondrial (SLP and RLPH) or archaeal (ALPH) gene entry into eukaryotes, supplemented by possible instances of lateral gene transfer between bacteria and eukaryotes. In addition to the previously known green algal and plant SLP1 and SLP2 protein forms, a more ancestral third form (SLP3) was found in green algae. Data from in silico subcellular localization predictions revealed class-specific differences in plants likely to result in distinct functions, and for SLP sequences, distinctive and possibly functionally significant differences between plants and nonphotosynthetic eukaryotes. Conserved carboxyl-terminal sequence motifs with class-specific patterns of residue substitutions, most prominent in photosynthetic organisms, raise the possibility of complex interactions with regulatory proteins.

Molecular mechanisms underlying the interaction of protein phosphatase-1c with ASPP proteins.

The serine/threonine PP-1c (protein phosphatase-1 catalytic subunit) is regulated by association with multiple regulatory subunits. Human ASPPs (apoptosis-stimulating proteins of p53) comprise three family members: ASPP1, ASPP2 and iASPP (inhibitory ASPP), which is uniquely overexpressed in many cancers. While ASPP2 and iASPP are known to bind PP-1c, we now identify novel and distinct molecular interactions that allow all three ASPPs to bind differentially to PP-1c isoforms and p53. iASPP lacks a PP-1c-binding RVXF motif; however, we show it interacts with PP-1c via a RARL sequence with a Kd value of 26 nM. Molecular modelling and mutagenesis of PP-1c-ASPP protein complexes identified two additional modes of interaction. First, two positively charged residues, Lys260 and Arg261 on PP-1c, interact with all ASPP family members. Secondly, the C-terminus of the PP-1c α, ß and γ isoforms contain a type-2 SH3 (Src homology 3) poly-proline motif (PxxPxR), which binds directly to the SH3 domains of ASPP1, ASPP2 and iASPP. In PP-1cγ this comprises residues 309-314 (PVTPPR). When the Px(T)PxR motif is deleted or mutated via insertion of a phosphorylation site mimic (T311D), PP-1c fails to bind to all three ASPP proteins. Overall, we provide the first direct evidence for PP-1c binding via its C-terminus to an SH3 protein domain.

Strigolactone insensitivity affects differential shoot and root transcriptome in barley.

Strigolactones (SLs) are plant hormones that play a crucial role in regulating various aspects of plant architecture, such as shoot and root branching. However, the knowledge of SL-responsive genes and transcription factors (TFs) that control the shaping of plant architecture remains elusive. Here, transcriptomic analysis was conducted using the SL-insensitive barley mutant hvd14.d (carried mutation in SL receptor DWARF14, HvD14) and its wild-type (WT) to unravel the differences in gene expression separately in root and shoot tissues. This approach enabled us to select more than six thousand SL-dependent genes that were exclusive to each studied organ or not tissue-specific. The data obtained, along with in silico analyses, found several TFs that exhibited changed expression between the analyzed genotypes and that recognized binding sites in promoters of other identified differentially expressed genes (DEGs). In total, 28 TFs that recognize motifs over-represented in DEG promoters were identified. Moreover, nearly half of the identified TFs were connected in a single network of known and predicted interactions, highlighting the complexity and multidimensionality of SL-related signalling in barley. Finally, the SL control on the expression of one of the identified TFs in HvD14- and dose-dependent manners was proved. Obtained results bring us closer to understanding the signalling pathways regulating SL-dependent plant development.

Twilight length alters growth and flowering time in Arabidopsis via LHY/CCA1.

Decades of research have uncovered how plants respond to two environmental variables that change across latitudes and over seasons: photoperiod and temperature. However, a third such variable, twilight length, has so far gone unstudied. Here, using controlled growth setups, we show that the duration of twilight affects growth and flowering time via the LHY/CCA1 clock genes in the model plant Arabidopsis. Using a series of progressively truncated no-twilight photoperiods, we also found that plants are more sensitive to twilight length compared to equivalent changes in solely photoperiods. Transcriptome and proteome analyses showed that twilight length affects reactive oxygen species metabolism, photosynthesis, and carbon metabolism. Genetic analyses suggested a twilight sensing pathway from the photoreceptors PHY E, PHY B, PHY D, and CRY2 through LHY/CCA1 to flowering modulation through the GI-FT pathway. Overall, our findings call for more nuanced models of day-length perception in plants and posit that twilight is an important determinant of plant growth and development.

Proteome- and metabolome-level changes during early stages of clubroot infection in Brassica napus canola.

Clubroot is a destructive root disease of canola (Brassica napus L.) caused by Plasmodiophora brassicae Woronin. Despite extensive research into the molecular responses of B. napus to P. brassicae, there is limited information on proteome- and metabolome-level changes in response to the pathogen, especially during the initial stages of infection. In this study, we have investigated the proteome- and metabolome- level changes in the roots of clubroot-resistant (CR) and -susceptible (CS) doubled-haploid (DH) B. napus lines, in response to P. brassicae pathotype 3H at 1-, 4-, and 7-days post-inoculation (DPI). Root proteomes were analyzed using nanoflow liquid chromatography coupled with tandem mass spectrometry (nano LC-MS/MS). Comparisons of pathogen-inoculated and uninoculated root proteomes revealed 2515 and 1556 differentially abundant proteins at one or more time points (1-, 4-, and 7-DPI) in the CR and CS genotypes, respectively. Several proteins related to primary metabolites (e.g., amino acids, fatty acids, and lipids), secondary metabolites (e.g., glucosinolates), and cell wall reinforcement-related proteins [e.g., laccase, peroxidases, and plant invertase/pectin methylesterase inhibitors (PInv/PMEI)] were identified. Eleven nucleotides and nucleoside-related metabolites, and eight fatty acids and sphingolipid-related metabolites were identified in the metabolomics study. To our knowledge, this is the first report of root proteome-level changes and associated alterations in metabolites during the early stages of P. brassicae infection in B. napus.

Arabidopsis thaliana histone deacetylase 14 (HDA14) is an α-tubulin deacetylase that associates with PP2A and enriches in the microtubule fraction with the putative histone acetyltransferase ELP3.

It is now emerging that many proteins are regulated by a variety of covalent modifications. Using microcystin-affinity chromatography we have purified multiple protein phosphatases and their associated proteins from Arabidopsis thaliana. One major protein purified was the histone deacetylase HDA14. We demonstrate that HDA14 can deacetylate α-tubulin, associates with α/ß-tubulin and is retained on GTP/taxol-stabilized microtubules, at least in part, by direct association with the PP2A-A2 subunit. Like HDA14, the putative histone acetyltransferase ELP3 was purified on microcystin-Sepharose and is also enriched at microtubules, potentially functioning in opposition to HDA14 as the α-tubulin acetylating enzyme. Consistent with the likelihood of it having many substrates throughout the cell, we demonstrate that HDA14, ELP3 and the PP2A A-subunits A1, A2 and A3 all reside in both the nucleus and cytosol of the cell. The association of a histone deacetylase with PP2A suggests a direct link between protein phosphorylation and acetylation.

Phosphorylation mediated regulation of RNA splicing in plants.

For the past two decades, the study of alternative splicing (AS) and its involvement in plant development and stress response has grown in popularity. Only recently however, has the focus shifted to the study of how AS regulation (or lack-thereof) affects downstream mRNA and protein landscapes and how these AS regulatory events impact plant development and stress tolerance. In humans, protein phosphorylation represents one of the predominant mechanisms by which AS is regulated and thus the protein kinases governing these phosphorylation events are of interest for further study. Large-scale phosphoproteomic studies in plants have consistently found that RNA splicing-related proteins are extensively phosphorylated, however, the signaling pathways involved in AS regulation have not been resolved. In this mini-review, we summarize our current knowledge of the three major splicing-related protein kinase families in plants that are suggested to mediate AS phospho-regulation and draw comparisons to their metazoan orthologs. We also summarize and contextualize the phosphorylation events identified as occurring on splicing-related protein families to illustrate the high degree to which splicing-related proteins are modified, placing a new focus on elucidating the impacts of AS at the protein and PTM-level.

Systems-level proteomics and metabolomics reveals the diel molecular landscape of diverse kale cultivars.

Kale is a group of diverse Brassicaceae species that are nutritious leafy greens consumed for their abundance of vitamins and micronutrients. Typified by their curly, serrated and/or wavy leaves, kale varieties have been primarily defined based on their leaf morphology and geographic origin, despite having complex genetic backgrounds. Kale is a very promising crop for vertical farming due to its high nutritional content; however, being a non-model organism, foundational, systems-level analyses of kale are lacking. Previous studies in kale have shown that time-of-day harvesting can affect its nutritional composition. Therefore, to gain a systems-level diel understanding of kale across its wide-ranging and diverse genetic landscape, we selected nine publicly available and commercially grown kale cultivars for growth under near-sunlight LED light conditions ideal for vertical farming. We then analyzed changes in morphology, growth and nutrition using a combination of plant phenotyping, proteomics and metabolomics. As the diel molecular activities of plants drive their daily growth and development, ultimately determining their productivity as a crop, we harvested kale leaf tissue at both end-of-day (ED) and end-of-night (EN) time-points for all molecular analyses. Our results reveal that diel proteome and metabolome signatures divide the selected kale cultivars into two groups defined by their amino acid and sugar content, along with significant proteome differences involving carbon and nitrogen metabolism, mRNA splicing, protein translation and light harvesting. Together, our multi-cultivar, multi-omic analysis provides new insights into the molecular underpinnings of the diel growth and development landscape of kale, advancing our fundamental understanding of this nutritious leafy green super-food for horticulture/vertical farming applications.

Machine learning classification of plant genotypes grown under different light conditions through the integration of multi-scale time-series data.

In order to mitigate the effects of a changing climate, agriculture requires more effective evaluation, selection, and production of crop cultivars in order to accelerate genotype-to-phenotype connections and the selection of beneficial traits. Critically, plant growth and development are highly dependent on sunlight, with light energy providing plants with the energy required to photosynthesize as well as a means to directly intersect with the environment in order to develop. In plant analyses, machine learning and deep learning techniques have a proven ability to learn plant growth patterns, including detection of disease, plant stress, and growth using a variety of image data. To date, however, studies have not assessed machine learning and deep learning algorithms for their ability to differentiate a large cohort of genotypes grown under several growth conditions using time-series data automatically acquired across multiple scales (daily and developmentally). Here, we extensively evaluate a wide range of machine learning and deep learning algorithms for their ability to differentiate 17 well-characterized photoreceptor deficient genotypes differing in their light detection capabilities grown under several different light conditions. Using algorithm performance measurements of precision, recall, F1-Score, and accuracy, we find that Suport Vector Machine (SVM) maintains the greatest classification accuracy, while a combined ConvLSTM2D deep learning model produces the best genotype classification results across the different growth conditions. Our successful integration of time-series growth data across multiple scales, genotypes and growth conditions sets a new foundational baseline from which more complex plant science traits can be assessed for genotype-to-phenotype connections.