Rice is the most important staple crop in China and an ideal model plant. How to increase yield and improve quality has been the major challenge in rice research and breeding. However, to meet the security of food supply, the crop yield per unit area has to be increased by 50% before 2030. My lab has been focused on systematical dissection of key genes regulating rice plant architecture, elucidation of their molecular mechanisms and the genetic regulatory network, and application of this fundamental knowledge to breed new high-yield elite varieties.

Plant architecture is a primary determinant of crop yield and has profound adaptability in response to internal and external factors. Research on rice plant architecture and understanding of the molecular basis and regulatory mechanism are of great importance to both basic science and practical application. My lab has molecularly characterized a series of rice mutants that show various altered shoot architectures, including monoclum 1 (moc1), tillering and dwarf 1 (tad1), ideal plant architecture 1 (ipa1), lazy1 (la1), dwarf 27 (d27), dwarf 53 (d53), and ipa1 interacting protein1 (ipi1). Through in-depth characterization of their functions, we demonstrated that the tillering phenotypes mainly result from either the deficiency in the branching plant hormone strigolactone biosynthetic and signaling pathways or the transcription factors that play key roles in shoot architecture formation. The understanding of these gene functions enable us to elucidate a comprehensive regulatory network underlying the formation of shoot architecture and greatly expanded our knowledge of plant development.

Based on the knowledge from the study on shoot architecture of rice, our lab and collaborators have made enormous efforts in breeding new elite varieties through developing a molecular breeding system. These results demonstrate that this system could truly speed up the rice breeding course and generate new super rice with multiple advantages. However, to fully meet the challenge of food supply, more efforts need to be made by providing more genetic resources and designing different allele combinations to meet the unique demands in different regions in China.

Plant development is characterized by spectacular levels of plasticity, such that a single genotype can give rise to a wide range of phenotypes. The degree of shoot branching is an excellent example of this phenomenon with branching levels being influenced by diverse environmental and developmental factors throughout the life cycle. Fundamentally this is played out at the level of the behavior of axillary buds, established at the base of each leaf on the plant. Axillary buds can remain dormant or they can activate to produce a shoot. The commitment of axillary buds to switch between dormancy and activity involves the integration of local and systemic factors, and is mediated at least in part by the self-organising properties of the transport network for the plant hormone auxin. A central feature of this system is the competition between axillary buds for access to a common auxin transport path down the main stem to the root, along with positive feedback between auxin and its transport, which reinforces active transport routes. This system supports robust switching behavior in buds, dependent on their relative rather than absolute properties, which can be both globally and locally modulated.

Plants have developed smart strategies for detoxifying toxic metals in the environment. At this symposium, I will show two examples on how plants cope with toxic metals including aluminum (Al) and manganese (Mn), which are two major factors liming crop production on acid soils.

Al tolerance in barley is controlled by a single major gene on chromosome 4H, namely HvAACT1 (Al-activated citrate transporter 1), belonging to the multidrug and toxic compound extrusion (MATE) family. It functions to mediate citrate release from the roots to detoxify Al externally in the rhizosphere. We found that the differential expression of HvAACT1 in different barley accessions is due to the insertion of a 1023-bp CACTA-like transposon in the upstream genomic sequence of HvAACT1 in some Al-tolerant accessions in East Asia. This 1-kb insertion promotes enhanced expression level of HvAACT1 and also spatially alters the HvAACT1 expression pattern from the mature root region to the root tips, where it functions to release citrate to the xylem for iron (Fe) translocation from the roots to the shoots. Recently, we found a novel mechanism regulating HvAACT1 expression in European barley accessions. A multi-retrotransposon-like (MLR) sequence insertion of at least 15.3 kb was detected in the upstream genomic sequence of HvAACT1 in several European accessions. Furthermore, DNA demethylation of this MLR insertion was found to be involved in increasing the expression of HvAACT1 uniquely in Al-tolerant European accessions. These results indicate that barley accessions in East Asia and Europe have developed independent but equivalent strategies to withstand Al toxicity in acid soils.

On the other hand, Mn is an essential element for plants, but only a small amount is required for their healthy growth. However, Mn concentration in soil solution varies from submicromolar to hundreds of micromolars. Rice is very tolerant of high Mn and is able to accumulate high Mn in the shoots. We found that Mn uptake in rice is mediated by two different transporters; OsNramp5 and OsMTP9 polarly localized at the exodermis and endodermis of the roots, but they don’t respond to environmental Mn changes at both transcriptional and protein level. In contrast, OsNramp3 localized at the nodes responds to different Mn concentration at the protein level, but not at the transcriptional level. It functions as a switch in response to Mn changes; preferentially delivering an adequate amount of Mn to the developing tissues at Mn-limited condition, but switching Mn to the older leaves through rapid degradation of this protein at high Mn. Detoxification of high Mn in the leaves is mediated by at least four transporters localized at the plasma membrane, tonoplast and trans-Golgi, respectively.

Application of a newly developed widely-targeted metabolomics strategy simultaneously detected hundreds of both primary and secondary metabolites in rice and disclosed a number of subspecies-specific metabolites that may reflect, as well as affect the subspecies differentiation of rice. Distinct and overlapped accumulation was observed and complex genetic regulation of metabolism was revealed in two different tissues by subsequent metabolic QTL (mQTL) mapping and metabolic genome-wide association study (mGWAS). Hundreds of loci with high resolution and large effects were uncovered. Interactive gene/metabolite identification/annotation was facilitated for both functional genomics and metabolomics. Data mining revealed a large number of candidate genes underlying metabolites that are of physiological and agronomical importance, and also be applied to the bulk identification of tailing enzymes contributing most to metabolic diversity. Similar approaches were also applied to the understanding of maize kernel metabolome. Furthermore, comparative mGWAS between rice and maize resulted in greatly increased power and resolution in both species. In addition, parallel metabolic and phenotypic GWAS identified new candidate genes responsible for both metabolic and morphologic traits such as grain width and leaf senescence, revealing direct linkage between the metabotype and the phenotype. Our study not only reveals novel biochemical and genetic insights of important aspects of plant and human such as development, stress resistance, and nutrition/health-promoting, but also provides vast amount of high-quality data for further understanding plant metabolome which may help bridge the gap between the genome and phenome. The strategy is a powerful tool for large-scale gene identification, pathway elucidation, and for knowledge-based crop genetic improvement.

Plant surveillance of microbes and microbial counter-surveillance provide a fascinating framework for deciphering host-pathogen coevolution and disease resistance signaling/execution mechanisms. Understanding the control of immunity/cell death outputs, also by observing what happens when things go wrong (as in autoimmunity), has important implications for enhancing plant growth, fitness and resilience to pathogens. I’ll describe our analysis of host recognition of biotrophic pathogens and the processes by which intracellular (NLR) immune receptors, sensing pathogen interference with host cells, transmit recognition to downstream resistance pathways. Using molecular genetic, genomic, transcriptomic and protein structural approaches in Arabidopsis we’ve identified distinct NLR-induced defense sectors that contribute to immunity and we’re assessing how these pathways operate within the broader environmental stress network. Arabidopsis also serves as a springboard to interrogate stress network wiring in other plant species with different NLR receptor repertoires. I’ll further describe our analysis of natural genetic variation in Arabidopsis defense/growth trade-offs and from this the identification of potentially useful underlying variants.

Plants deploy two major classes of immune receptors to sense pathogens and activate defenses. Cell surface-localized receptor kinases and receptor-like proteins, collectively called pattern recognition receptors (PRRs), perceive microbe- or host-derived molecular patterns associated with infection and constitute the first layer of plant immunity. Pathogenic microbes, however, are capable of secreting diverse effector proteins into the plant cell to inhibit pattern-triggered immunity (PTI). The second class immune receptors, which are cytoplasmic nucleotide-binding, leucine-rich repeat receptors (NLRs), are specialized to recognize pathogen effectors and trigger immunity when PTI fails. Although a quarter century has passed since the isolation of the first NLR-coding gene in plants, we know little concerning biochemical basis of NLR regulation.

We have been using Arabidopsis-bacterial pathogen interactions as a model to understand how pathogen effectors subvert PTI and how plants recognize pathogens through the two classes of immune receptors. To this end, we previously identified a receptor-like cytoplasmic kinase (RLCK) called BIK1 as a central component of multiple PRR complexes. BIK1 mediate immune signaling by directly phosphorylating major downstream components, and several bacterial pathogen effectors can attack BIK1 to block PTI signaling. One such effector, AvrAC, is an uridylyl transferase encoded by the black rot pathogen Xanthomonas campestris campestris. It uridylylates key serine/threonine residues in the activation loop of BIK1 kinase to promote virulence on plants. ZAR1, an ancient NLR protein with a N terminal coiled coil (CC) domain shared by both Arabidopsis and Nocotianan benthamiana, is capable of recognizing AvrAC. The recognition occurs indirectly through two additional host proteins RKS1, a pseudokinase of the RLCK XII family, and PBL2, a close homolog of BIK1 belonging to the RLCK VII family. RKS1 constitutively associate with ZAR1 and serves as an adaptor. PBL2 serves as a decoy for BIK1 and is uridylylated by AvrAC in the activation loop. PBL2UMP then acts as a ligand and binds RKS1 in the preformed ZAR1-RKS1 complex to trigger disease resistance. By collaborating with structural biologists, we solved cryo-EM structures of the ZAR1-RKS1 binary complex and ZAR1-RKS1-PBL2UMP tertiary complex. Structural comparisons and structure-guided mutagenesis demonstrate that the ZAR1 NLR is maintained in an inactive state stabilized by multiple intramolecular interactions and ADP. The binding to PBL2UMP causes an allosteric change in the activation segment of RKS1 which in turn causes steric hindrance to the nucleotide-binding domain of ZAR1 and impaired binding to ADP. The depletion of ADP primes the ZAR1-RKS1-PBL2UMP complex for further activation by ATP or dATP, which induces oligomerization of ZAR1-RKS1-PBL2UMP, forming a pentameric resistosome that is essential for immune activation. Cryo-EM structure and site directed mutagenesis show that, upon oligomerization, the CC domains of ZAR1 resistosome form a funnel-like structure required for the triggering cell death and disease resistance. Together these results uncover a series of mechanisms through which plant NLR is regulated before and after ligand binding and provides an excellent template for the understanding of other plant NLRs, which constitute the largest family of immune receptors in plants.

The epigenetic mark H3K27me3 controls spatial and temporal patterning throughout the flowering plant life cycle. The histone variant H3.1 is essential for the propagation of H3K27me3 (Jiang and Berger, Science, 2017). Whether this mark is extensively reprogrammed at any point in development has remained unclear. We show that H3K27me3 is selectively erased from sperm chromatin in Arabidopsis through a multi-layered mechanism, which includes the deposition of a sperm-specific histone H3 variant that is immune to methylation at lysine 27. Resetting of H3K27me3 also coincides with selective types of deposition of H3K4me3 in a manner that potentiates expression in mature sperm and at distinct phases of development after fertilization. In stark contrast to the global resetting of epigenetic marks in mammals, plants have evolved mechanism to selectively reprogram H3K27me3. This mechanism acts as an epigenetic timer to control gene expression during plant development.

Next generation sequencing technologies and the associated informatics are transforming biology into big data science. Vegetables are essential for human health and their production constitutes a major section in rural economy, but genomic resource for vegetable breeding was scarce. To overcome this limitation, my lab coordinated and participated de novo genome sequencing several major vegetable crops, including cucumber, tomato, and potato. We also construct genome variation maps for cucumber and tomato, providing millions SNP for molecular breeding. In total, so far we generated 50 Tbps (5 x 10¹³ bps) genomic data. We integrated this big data with molecular biology and biochemistry to discover the mechanisms of several biological and evolutionary processes, including domestication of cucumber and tomato, sex determination and fruit elongation in cucumber, and flavor metabolism in tomato. These studies enable a molecular design of a cucumber genotype with bitter foliage for insect resistance and nonbitter fruits even under stress, as well as a tomato cultivar with higher sugar content and pleasant flavor. Furthermore, we are pursuing to re-domesticate potato by genome design, an initiative to transform the tuber-propagated autopolyploid that is difficult for breeding and logistics, into a breeder- and environment- friendly crop that is a diploid and propagated by true seeds.

My lab is interested in discovering and utilizing standing natural variation from the genus Oryza to breed new varieties of rice that are higher yielding and more nutritious, but have less of an environmental footprint (i.e. rice that requires less water, fertilizer & pesticides, can grow on marginal lands and have reduced greenhouse gas emissions) – a.k.a. “Green Super Rice”. We specialize in the generation of platinum standard reference genomes (PSRefSeqs) that represent the genetic diversity of cultivated rice and its wild relatives. These genomes serve as a pan-Oryza¬genome that can be used as a template to map resequencing data from rice germplasm banks across the globe aimed at capturing virtually all genetic variation that exist across the Oryzasphere. Such information can be used in GWAS studies and genomic selection strategies, coupled with high throughput phenotyping, to breed the crops of the future to help solve the 2050 10-billion people question – i.e. “How do we feed our world without destroying our world”.

Ethylene is a unique phytohormone, which is a simple gas molecule that regulates a wide array of plant processes, including fruit ripening and abscission, leaf and flower senescence, stress tolerance and pathogen defense. We have demonstrated that plant responses to ethylene are mediated by the regulation of EBF1/EBF2-dependent proteasomal degradation of EIN3 and EIL1 transcription factors. EIN2, whose null mutant is completely insensitive to ethylene, acts as an essential ethylene signaling component. Recently, we found that, in the present of ethylene, the C-terminus of EIN2 (CEND) undergoes proteolytic cleavage, and subsequently translocates into the nucleus to promote the accumulation of EIN3/EIL1, as well as to induce the turnover of EBF1/EBF2 proteins. However, EIN2 CEND was found to package into discrete and prominent cytoplasmic foci apart from nuclear localization in response to ethylene. This raises the possibility that there must exist additional mode of EIN2 action other than its nuclear function in ethylene signaling cascade. We report a novel mechanism of EIN2-mediated ethylene signaling, whereby EIN2 imposes the translational repression of EBF1 and EBF2 mRNA. We find that the EBF1/2 3’ untranslated regions (3’UTRs) mediate EIN2-directed translational repression, and identify multiple PolyU motifs as functional cis-elements of 3’UTRs. Furthermore, we demonstrate that ethylene induces EIN2 to associate with 3’UTRs and target EBF1/2 mRNA to cytoplasmic processing-body (P-body) through interacting with multiple P-body factors, including EIN5 and PABs. Our study illustrates the translational regulation as a key step in ethylene signaling, and represents the first demonstration of mRNA 3’UTR functioning as a “signal transducer” to sense and relay cellular signaling in plants. New advances on ethylene signaling will also be discussed.