A SYSTEMS BIOLOGY APPROACH TO UNCOVERING THE HIDDEN REDOX REACTIONS OF PLANT METABOLISM

• Sponsoring Institution: National Institute of Food and Agriculture
• Project Status: COMPLETE
• Funding Source: HATCH
• Reporting Frequency: Annual
• Accession No.: 1019934
• Project Start Date: Jul 1, 2019
• Project End Date: Jun 30, 2024
• Program Code: [(N/A)]- (N/A)

Recipient Organization
UNIVERSITY OF FLORIDA
G022 MCCARTY HALL
GAINESVILLE,FL 32611

Performing Department
Horticultural Science

Non Technical Summary
This project aims to understand how crops make certain small molecules that are vital for human and animal nutrition (vitamins, cofactors) or that serve for plant growth and defense against pathogens such as insects, bacteria, and fungi. The origin of these compounds is currently not well understood, because they are difficult to isolate and detect. Furthermore, many of the enzymes that plants use to make these compounds are not known. We propose to combine cutting-edge methods in computational biology, biochemistry and genetics to identify these 'missing' enzymes and find out what they do and where they are localized inside the plant cells. This proposed research will help design and breed crops that are better sources of certain vitamins, and inform growers about what the best practices are to improve the shelf-life of plant vitamins.

Animal Health Component

5%

Research Effort Categories

Basic

75%

Applied

5%

Developmental

20%

Classification

Knowledge Area (KA)	Subject of Investigation (SOI)	Field of Science (FOS)	Percent
201	7010	1000	100%

Knowledge Area
201 - Plant Genome, Genetics, and Genetic Mechanisms;

Subject Of Investigation
7010 - Biological Cell Systems;

Field Of Science
1000 - Biochemistry and biophysics;

Keywords

quinones

vitamins

cofactors

redox reactions

photosynthesis

respiration

Goals / Objectives
The main goals of this project are to uncover, characterize, and then engineer some hidden redox reactions required for the biosynthesis aromatic compounds that are central to crop productivity and nutritional quality. The cognate metabolic branches include the biosynthesis of photosynthetic and respiratory cofactors (ubiquinone [also called Coenzyme Q], phylloquinone, and plastoquinone), plant defense compounds (coumarins, indole alkaloids, methyl salicylate), vitamins for vertebrates (vit. E, vit. K1), and plant cell wall (ferulate, sinapate).To reach these goals, our objectives are to:1- Identify oxidoreductases involved in the methylation of aromatic compounds.1.1 We will model the regulatory networks of type II NAD(P)H dehydrogenases across the eukaryotic and prokaryotic kingdoms.1.2 We will search for physical associations in bacterial genomes (operons, gene fusions)2- Functionally characterize the reactions of oxidoreduction that are linked to the activity of methyl transferases2.1 We will use reverse genetics, targeted metabolic profiling and in vivo isotopic labeling to characterize plants and cyanobacterial candidates identified in Objective 1.2.2 We will perform enzymatic assays on the recombinant versions of these candidates2.3 We will determine the subcellular localizaion of these candidates3- Identify metabolons and engineer biological synthomes that protect redox active aromatic intermediates from spontaneous re-oxidation.3.1 We will identify 'hidden' metabolons ('metabolic channels') in plant and bacterial genomes searching in particular for gene fusion events3.2 We will design dehydrogenase/methyltransferase fusions in order to create metabolons, where biosynthetic intermediates are protected from spontaneous reoxidation

Project Methods
Computational Biology: We have developed a systems biology approach to detect the footprints that 'missing' metabolic genes have left in genomes and transcriptomes. Our strategy rests on the concept that on an evolutionary time scale and driven by the need for the co-regulation of gene expression, genes that are involved in the same cellular process tend to physically associate in genomes (e.g. operons, fusion events), to co-occur in the same regulatory network, and to be taxonomically co-distributed. By combining comparative genomics and gene network reconstruction, we will take advantage of these genomics and transcriptomics imprints to establish functional linkages between known and unknown genes. In the case of metabolic pathways, it becomes possible to identify 'pathway holes' i.e. missing enzymes for reactions known to occur but for which no candidate gene could previously be found, or -as is more specific to our proposed work- to detect unsuspected reactions and their corresponding enzymes. Crucially, because the method mines across phylogenetic lineages, it is able to detect functional associations that otherwise would have remained invisible in a single organism.Reverse genetics, targeted metabolic profiling and in vivo isotopic labeling: Guided by our in silico modeling and comparative genomics data mining, we will analyze plant and bacterial mutants for changes in the levels of prenylated quinones and related intermediates. We will conduct isotopic tracer experiments using 13C to identify the biosynthetic precursors and associated metabolic branches.Enzymatic assays: Candidate genes will be expressed in E. coli in order to determine their substrate specificity, cofactor requirement and kinetic features. In the case of proteins that are refractory to recombinant expression, we will implement solubilization/renaturation protocols that take advantage of the anti-aggregation properties of non-detergent sulphobetaines during protein dialysis and re-folding.Subcellular localization experiments: We will use fusions to fluorescent reporter proteins coupled to imaging by confocal microscopy to determine the subcellular localization of the identified enzymes. In the case of dual targeted enzymes (e.g. peroxisomes/chloroplasts; chloroplasts/mitochondria), we will perform hemi-complementation experiments of cognate Arabidopsis mutants using engineered versions of these proteins targeted to a single organelle.Protein engineering: We will design codon-optimized versions and gene fusions of the oxidoreductase/methyltransferase pairs that we identify to generate transgenics in which unstable metabolic intermediates will be be-routed towards desirable end products (e.g. vitamin K).Effort: Our research will provide unprecedented opportunities for graduate students and postdoctoral researchers involved in the project to learn cutting-edge methods in gene network modeling, comparative genomics, and biochemical genetics.Evaluation: Students and postdoctoral researchers will work under the direct supervision of the project director. Findings will be discussed during weekly lab meetings. Annotated candidate genes will be deposited in the SEED database (http://pubseed.theseed.org).

Progress 10/01/19 to 09/30/20

Outputs
Target Audience:International and national experts in bioenergetics (photosynthesis and respiration) and plant specialized metabolism via publications, conferences, and invited seminers). Local: i) Training of undergraduate and graduate students; ii) Training of postdoctoral researchers; iii) Collaboration with the USDA-CMAVE laboratory, Gainesville, FL Changes/Problems:Our on-site research was paused from March to June 2020 due COVID-19 lockdown at the University of Florida. Furthermore, because we were barred from accessing our laboratory and greenhouse facilities during this time, we lost many of the plant transgenics (monooxygenase RNAi lines, RquA overexpressor lines, candidate methyltransferase CRISPR knockout lines) that we had generated for this project. We estimate the incurred loss of time to 5-6 months. Furthermore, although on-site research was paused, our data mining, in silico modeling and comparative genomics work has progressed much faster than originally anticipated. We have therefore come up with several new candidate genes (3 methyltransferases, 2 FAD oxidoreductases and 3 ABC binding proteins) to test. We anticipate that this coming year our effort will be focused on performing almost exclusively wet bench work in order to catch up on the delays.I am confident that we can still deliver this coming year what was originally proposed in our original plan of work . Due to the closure of local schools last March, we have not been able to perform our outreach activities in 2020. What opportunities for training and professional development has the project provided?The project contributed to the training and career development of: - 2graduate students:Dr. Ann Bernert, whograduated in March 2020,and Dr. Scott Latmer, who graduated in November 2019 - 1 graduate rotating student (Megan Kelly)[Please note that thisstudentisa coauthoron one of our published studythis reporting year] - 2 graduate students (Shea Keene and Timothy Johnson) via collaborations [Please note that these students are coauthors on two of our published studiesthis reporting year] - 2 postdoctoral researchers (Dr. Eric Soubeyrand and Dr. Antoine Berger) - 3undergraduate students (Samantha McDonal, Taylor Nolff and David Berryman) [Samantha and David havejoined graduate schools, while Taylor joined Medical School in the US Army] How have the results been disseminated to communities of interest?Two manuscripts (Biochem. J., Molecules) and 2 PhD dissertations have been published. - Samantha McDonald and David Berryman presented the results of their findings to the 20th Annual Undergraduate Research Symposium at the University of Florida Genetics Institute. -Taylor Nolff presented the results of her bionformatics study in our group as part of a 3-credit Bioinformatics class in Microbiology (BSC4913) - Dr. Soubeyrand presented his findings at the Annual Meeting of the Canadian Society of Plant Biology last July (oralcommunication + poster) - Dr. Ann Bernert and Dr. Scott Latimer defended their PhD publically (March 2020 and October 2019, respectively). What do you plan to do during the next reporting period to accomplish the goals?i) We will complete the molecular and biochemical characterization of a mitochondrial monooxygenase and a cytosolic transporter that are critical for the biosynthesis of plant coenzyme Q. We have generated newCRISPR/Cas9 (knockout) and RNAi (silenced) mutant lines in Arabidopsis (we havelost all our previous transgenic linesin Spring 2020 during the university shutdown). We will use these mutants in feeding assays with heavy isotopes ([13C]-Phenylalanine and [13C]-Tyrosine) to quantify the decrease in biosynthetic fluxes as compared to wild-type controls. We will produce the recombinant versions of the monooxygenase and transporter subunits from Arabidopsis intoEscherichia coliandSaccharomyces cerevisae.The monooxygnase will be assayedin vitro, while the transporter subunits will be used for protein-protein interaction studies using size exclusion chromatography and native polyacrylamide gel separation. Theseprojectsl will becarried out by Dr. Scott Latimer and Dr. Antoine Berger. ii) We will initiate the functional characterization of two plant methyltransferases involved in the biosynthesis of prenylated quinones. In particular, we will attempt to determine, if the plant enzymes have retained the broad substrate preference that allow their eubacterial ancestors to moonlight in multiple biosynthetic pathways (e.g. vitamin K, vitamin E, plastoquinone, Coenzyme Q). To do that, the genes corresponding to these two plant enzymes will be introduced in anE. coliknockout mutant (menG/ubiE) corresponding to the methyl transfer steps in the biosynthesis of menaquinone (vitamin K2) and coenzyme Q. Beyond the gain in basic knowledge, the significance of this project is to develop new enzymatic tools for manipulating vitamin and cofactor metabolism via synthetic biology approachesin prokaryotes and eukaryotes. This work will be carried out by Dr. Lauren Stutts, who was recently hired in our group. ii) Expanding on our recent discovery that plants are able to branch the biosynthesis Coenzyme Q from the catabolism of kaempferol,we will examine how plants control the metabolic branch-point between these two pathways. Specifically, we will quantify to what extent the 3-O-glycosylation of kaempferol protects this flavonol from peroxidative cleavage. We have initiated a collaboration with Pr. AntonSchäffner (Institute of Biochemical Plant Pathology, Munich, Germany) to accomplish this goal. iii) We will generate new Arabidopsis transgenics engineered to produce a rare prenylated quinone, called rhodhoquinone, which confers the extraordinary capability to certain organisms to operate their respiratory chain in absence of oxygen (here again we have lost previous constructs during the university shutdown last year).

Impacts
What was accomplished under these goals? i) Using our RNA seq data and metabolic reconstructions, we identified an Arabidopsisgene of unknown function as the top co-expressor of the methyltransferases COQ3 and COQ5 that are involved in the biosynthesis of ubiquinone in plants. We showed that the corresponding protein displays some homologies with bacterial monooxygenases UbiH, UbiF and UbiI. Using functional complementation assays of the ubih, ubif and ubii knockout mutants in E. coli, we showed that the Arabidopsis gene restored ubiquinone production in the ubif mutant but not in the ubih and ubii ones. The plant enzyme thus appears to be monofunctional.Using GFP-fusion assays, we demonstrated that the plant enzyme is targeted to the mitochondrion. We also showed that the corresponding plant knockout mutant is embryo lethal. We therefore constructed some silenced lines (RNAi), which were found to be viable albeit with some significant decrease in intracellular ubiquinone content (~30% of wild-type level). A manuscript reporting these findings is in preparation. ii) Using [13C]-Phenylalanine feeding assays, gene co-expression analysis and reverse genetics, we showed that an Arabidopsis acyl-activating enzyme, product of gene 4-CL8, catalyzes the committed step in the beta-oxidative shortening of p-coumarate into 4-hydroxybenzoate. We showed that a cognate knockout mutant displayed a 20%decrease in ubiquinone content compared with wild-type plants, while 4-CL8 overexpression boosted ubiquinone content up to 150% of the control level. Furthermore, [13C]-Phenylalanine feeding assays confirmed that the enrichment of ubiquinone' s ring was decreased in the knockout as compared with wild-type controls. We also showed that this metabolic blockage could be bypassed by the exogenous supply of 4-hydroxybenzoate, the product of p-coumarate beta-oxidation. Using confocal microscopy experiments, we demonstrated that this Arabidopsis p-coumarate-CoA ligase is imported into peroxisomes. A paper reporting the identification and characterization of this new ubiquinone biosynthetic enzyme in plantswas published. iii) In collaboration with the group of Pr. Catherine Clarke at UCLA, wedemonstrated that human kidney cells use the plant favonol kaempferol as a precursor for the biosynthesis of coenzyme Q. A paper reporting this finding was published. iv) We discovered that 3 Arabidopsis glycosyltransferases, called UGT78D1, UGT78D2 and UGT78D3 control the flux of ubiquinone biosynthesis via stabilization of kaempferol.These data are congruent with our previous finding that unprotected (i.e. deglycosylated) kaempferol is cleaved by peroxidases in vivo, releasing 4-hydroxybenzoate, which in turn is incorporated into ubiquinone.

Publications

• Type: Journal Articles Status: Published Year Published: 2019 Citation: Soubeyrand E, Kelly M, Keene SA, Bernert AC, Latimer S, Johnson TS, Elowsky C, Colquhoun TA, Block AK, Basset GJ. Arabidopsis 4-COUMAROYL-COA LIGASE 8 contributes to the biosynthesis of the benzenoid ring of coenzyme Q in peroxisomes. Biochem J. 2019 Nov 29;476(22):3521-3532. doi: 10.1042/BCJ20190688. PMID: 31688904.
• Type: Journal Articles Status: Published Year Published: 2020 Citation: Fern�ndez-Del-R�o L, Soubeyrand E, Basset GJ, Clarke CF. Metabolism of the Flavonol Kaempferol in Kidney Cells Liberates the B-ring to Enter Coenzyme Q Biosynthesis. Molecules. 2020 Jun 27;25(13):2955. doi: 10.3390/molecules25132955. PMID: 32605010; PMCID: PMC7412559.
• Type: Theses/Dissertations Status: Published Year Published: 2020 Citation: DISCOVERY OF RQUA FUNCTION AND PROSPECTS FOR OPTIMIZING QUINONE PROFILES IN BACTERIA, YEAST, AND PLANTS. Bernert AC. (PhD Dissertation)
• Type: Theses/Dissertations Status: Published Year Published: 2019 Citation: DISCOVERY OF BIOSYNTHETIC GENES AND BIOLOGICAL FUNCTIONS OF PLANT TERPENOID QUINONES. Latimer S. (PhD Dissertation)

Progress 07/01/19 to 09/30/19

Outputs
Target Audience:National and international experts in plant metabolism (publications, conferences, seminars). Local 1) Training of undergraduateand graduate students Local 2) Collaboration with USDA-CMAVE unit, Gainesville, FL. Local 3): Outreach toAlachua County 4-H andFort Clarke Middle School, Gainesville, FL. Changes/Problems: Nothing Reported What opportunities for training and professional development has the project provided?The project contributed to the training and career development of: - 2 graduate students (Ann Bernert and Scott Latimer) - 2 postdoctoral researchers (Dr. Eric Soubeyrand and Dr. Antoine Berger) - 3 undergraduate researchers (Samantha McDonald, David Berryman, Taylor Nolff) How have the results been disseminated to communities of interest?- Three manuscripts reporting our findings have been published (PNAS 116, 2374-2383); BBA 1864, 1226-1234; Plant Cell Environ. 43, 223-234) What do you plan to do during the next reporting period to accomplish the goals?1) We will set up in vitro methyltransferase assays using s-adenosylmethionine as a methyl donor and demethylated precursors of ubiquinone as acceptors to demonstrate that Ndb1 and succinate dehydrogenase catalyze the pre-requisite reduction of the unmethylated intermediates for their subsequent transmethylation. 2) Our collaborator, Dr. Anna Block (USDA-CMAVE) will set up similar assays with the purified recombinant versions of the corn methyltransferase candidates using various phenolics as substrates (p-coumarate, anthranilate and salicylate in particular). 3) Having shown that RquA alone is sufficient for the formation of rhodoquinone from ubiquinone in E.coli, we will engineer a codon-optimized/mitochondria-targeted version of this enzyme for expression in Arabidopsis. Rhodoquinone being used in its native organism to create an electron shunt at the level of complex II of the respiratory chain (anaerobic respiration), transgenics will be tested in vitro (extracts) and in vivo (hypoxia) for their capability to perform malate dismutation in mitochondria.

Impacts
What was accomplished under these goals? 1) Combining homology searches and RNAseq experiments, we have identified 10 maize O-methyltransferases that are induced in response to attack by fall armyworm (Spodoptera frugiperda). Transposon mutants have been isolated for 3 of these methyltransferases; all showed defect in the production of methylanthranilate. We also assessed the response of maize to the combinatorial stress of flooding and infestation with the insect pestS. frugiperda. This combined stress lead to elevated production of the defense hormone salicylic acid, which does not occur in the individual stresses, and the resultant salicylic acid-dependent increase inS. frugiperdaresistance. We also showed that changes in cellular redox status also occured, as indicated by reductions in peroxidase and polyphenol oxidase activity. These data suggest that metabolite changes important for flooding tolerance and anti-insect defence may act both additively and synergistically to provide extra protection to the plant. 2) Using a systems biology approach, we have identified an Arabidopsistype II NAD(P)H dehydrogenase (Ndb1), as well as subunits ofsuccinate dehydrogenase in Arabidopsis and E. coli that are functional interactors of orthologous methyltransferases (UbiE/COQ5 and UbiG/Coq3) involved in the biosynthesis of the respiratory cofactor coenzyme Q (ubiquinone). We have isolated knockouts corresponding to these dehydrogenases in E. coli and Arabidopsis and shown that these mutants display significant loss of coenzyme Q. Notably, the E. coli and Arabidopsis knockouts accumulate the unmethylated substrates of the UbiE/Coq5 and UbiG/Coq3-catalyzed reactions. 3) We have expressed in E. coli, yeast and plant (Arabidopsis) a methyltransferase-like (RquA) known to be crucial for the production of rhodoquinone in the photosynthetic proteobacterium Rhodospirullum rubrum. The engineered cells (E. coli, yeast, Arabidopsis) accumulated rhodoquinone. Repeating the same experiment in a series of ubiquinone biosynthetic mutants in E. coli, we obtained evidence that RquA catalyzes the direct transamination of ubiquinone into rhodoquinone. 4)We analyzed the expression patterns of nuclear, plastid and mitochondrial genomes of the unicellular green algaChlamydomonas reinhardtiigrown under light-dark cycles. We discovered that 85% of transcribed genes show differential expression, with different sets of transcripts being up-regulated over the course of the day to coordinate cellular growth before undergoing cell division. Parallel measurements of select metabolites and pigments, physiological parameters, and a subset of proteins allowed us to infer metabolic events and to evaluate the impact of the transcriptome on the proteome. Among the findings are the observations thatChlamydomonasexhibits lower respiratory activity at night compared with the day; multiple fermentation pathways, some oxygen-sensitive, are expressed at night in aerated cultures. These data lead us to propose that the ferredoxin 9 is the likely electron donor to hydrogenases. Our expression dataset, complemented with coexpression networks and metabolite profiling, should constitute an excellent resource for the algal and plant communities.

Publications

• Type: Journal Articles Status: Published Year Published: 2019 Citation: Strenkert D, Schmollinger S, Gallaher SD, Salom� PA, Purvine SO, Nicora CD, Mettler-Altmann T, Soubeyrand E, Weber APM, Lipton MS, Basset GJ, Merchant SS. Multiomics resolution of molecular events during a day in the life of Chlamydomonas. Proc Natl Acad Sci U S A. 2019 Feb 5;116(6):2374-2383. doi: 10.1073/pnas.1815238116. Epub 2019 Jan 18. PubMed PMID: 30659148; PubMed Central PMCID: PMC6369806.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Bernert AC, Jacobs EJ, Reinl SR, Choi CCY, Roberts Buceta PM, Culver JC, Goodspeed CR, Bradley MC, Clarke CF, Basset GJ, Shepherd JN. Recombinant RquA catalyzes the in vivo conversion of ubiquinone to rhodoquinone in Escherichia coli and Saccharomyces cerevisiae. Biochim Biophys Acta Mol Cell Biol Lipids. 2019 Sep;1864(9):1226-1234. doi: 10.1016/j.bbalip.2019.05.007. Epub 2019 May 21. PubMed PMID: 31121262; PubMed Central PMCID: PMC6874216.
• Type: Journal Articles Status: Published Year Published: 2019 Citation: Block AK, Hunter CT, Sattler SE, Rering C, McDonald S, Basset GJ, Christensen SA. Fighting on two fronts: Elevated insect resistance in flooded maize. Plant Cell Environ. 2019 Aug 14. doi: 10.1111/pce.13642. [Epub ahead of print] PubMed PMID: 31411732.