Location: Plant Gene Expression Center
2020 Annual Report
Objectives
The long-term goal of our research is to identify genes that regulate plant architecture in maize. We recently positionally cloned four genes that were defined by mutant phenotype. The phenotypes affect multiple aspects of architecture including leaf shape, internode length, tassel branching and sex determination. The phenotypes vary depending on inbred background. Of the four genes, one encodes a plasma membrane bound protein, one encodes a kinase, a third encodes an enzyme and the fourth is a conserved gene of unknown function. In order to connect their interesting phenotypes to mechanism we are identifying interacting partners, carrying out RNAseq and conducting metabolomic analysis. The work will increase our dataset from four genes to entire pathways. We will then combine knowledge of these pathways to transcription factor targets that are being developed in collaboration with others. This combined information will provide a network of connectivity that could be useful for breeding. For example, if we hope to change leaf angle, we can ask which genes appear to function solely in leaf angle and not also in leaf width or tassel branching. If we are selecting for improved abiotic stress, we can examine our network and see what genes are likely to have large or small effects.
Objective 1: Dissect gene networks that regulate leaf architecture and internode elongation in maize to provide targets for breeding more productive maize.
Subobjective 1A. Identify proteins that interact with NOD (NARROW ODD DWARF) and confirm the interaction, in vivo and in vitro.
Subobjective 1B. Identify proteins that are phosphorylated by LGN and carry out transcriptome analysis.
Subobjective 1C. Map modifiers of nod that are responsible for the inbred differences.
Objective 2: Characterize genes that regulate tassel branching and sex determination in maize for higher yields.
Subobjective 2A. Prove identity of Tasselseed5 (Ts5) gene by obtaining a revertant allele and by overexpressing the gene.
Subobjective 2B. Map the modifiers that differentiate Ts5 in Mo17 compared to B73.
Subobjective 2C. Obtain additional feminized upright narrow (fun) alleles and carry out RNAseq analysis.
Objective 3: Determine coordinated and independent pathways that regulate leaf, inflorescence, and internode development in maize for enhanced productivity.
Approach
For Objective 1, we hypothesize that NARROW ODD DWARF (NOD), a plasma membrane localized protein known to function in calcium signaling, is an essential protein in plants with a role in development and immunity. We are using proteomics to identify interacting partners and testing these interactions with biochemical and genetic experiments. We also hypothesize that LIGULELESS NARROW (LGN) is critical, given the severe mutant phenotype when it is not able to phosphorylate other proteins. We will determine the targets of this kinase and determine how and when it interacts with NOD. We have antibodies to both of these proteins that function in westerns and in Co-immunoprecipitation. Both NOD and LGN mutants are distinct in different inbreds. We mapped a modifier to LGN and plan to identify the modifiers for NOD. We hypothesize that there are distinct loci responsible for the ligule defects and other loci responsible for the auto-immunity defects. The modifiers will be identified using genotyping by sequencing (GBS) methods. We also have the possibility of mapping the modifiers by crossing to recombinant inbred lines in the GBS method doesn’t work.
For objective 2, we hypothesize that Ts5 encodes an enzyme in the jasmonic acid (JA) pathway. We will obtain a revertant of Ts5 using ethyl methyl sulfonate (EMS). If this doesn’t work, we will verify its function by following JA metabolites during wounding. We also plan to overexpress the gene in Brachypodium and determine the effect on plant development. Ts5 is completely feminized in the Mo17 inbred and it is mild in B73. We crossed Ts5 to the recombinant B73 Mo17 inbred lines (IBM) and identified 10 major quantitative trait locus (QTL). We combined this data with an RNA sequencin (RNAseq) experiment that identified the differentially expressed genes between Ts5 and normal tassels. Four genes were identified and we will obtain mutants in these genes to examine their function. The fun mutant is also feminized, but may not be in the JA pathway. From analysis of double mutants, we hypothesize it is in the brassinosteroid (BR) pathway. We are determining the BR levels and will analyze an RNAseq dataset to explore this hypothesis. Because FUN is a gene of unknown function, additional alleles will be useful for understanding the domains. These will be obtained by EMS screens.
For Objective 3, we will combine our different datasets into a network analysis. We hypothesize that few genes function in only one tissue and will determine the overlap in the tassel network and leaf network. This analysis may lead to genes that are not yet identified by a mutant phenotype and would be worth study in the future.
Progress Report
In support of Objective 1, research continued in Albany, California, on LIGULELESS NARROW (LGN) and NARROW ODD DWARF (NOD). LGN encodes a receptor-like kinase and NOD encodes a membrane-bound protein that may function in Calcium signaling. To identify NOD interacting proteins, we carried out coimmunoprecipitation (CoIP) using a polyclonal antibody that specifically interacts with NOD. We collected tissue from three-week old plants, dissecting off older leaves to enrich for the meristem and leaf primordia. The nod-1 null mutant was used as a negative control. Each immunoprecipitation included 30 shoot apices. The third round of experiments enriched for plasma membrane proteins and included a phosphatase inhibitor to promote phosphorylation-dependent affinity purification. Mass spectrometry analysis of this third round of coimmunoprecipitation revealed LGN as an interactor.
Given the similarity between nod and Lgn-R mutants and the fact that LGN also localizes in the plasma membrane, we explored this interaction in more detail. To test the in planta NOD-LGN interaction, we developed a specific polyclonal antibody against LGN to do reciprocal CoIP and test in vivo interaction with NOD in maize tissue. NOD CoIP was carried out using the inbred B73, Lgn-R heterozygous, and nod-1 homozygous shoot apex tissue. Both mutants were in the B73 inbred background. LGN was detected by western blot in the B73 (wild-type) CoIP but also in the background of the Lgn-R mutant, showing that the single amino acid change in Lgn-R does not affect the interaction with NOD. No signal was detected in nod-1, demonstrating the absence of background. The size of LGN protein interacting with NOD migrates at a higher molecular weight than the LGN in the input sample. Slower migration of these proteins suggests posttranslational modifications, most likely phosphorylation.
We developed an antibody to LGN to carry out the reverse experiment. LGN CoIP was carried out using wild type and Lgn-R shoot apex tissue and NOD was detected by western blot in the CoIP complex. The immunoprecipitated LGN either by Anti-NOD or by Anti-LGN shows a strong signal, but immunoprecipitated NOD by Anti-LGN shows a weak signal that is slightly shifted in size compared to NOD detected in the input. This difference suggests that the NOD version interacting with LGN is a modified one, possibly phosphorylated which is less abundant in the LGN complex or has less affinity to the Anti-NOD antibody.
We evaluated a double mutant between Lgn-R heterozygotes and nod-1 in two different inbred backgrounds. In the A619 inbred, the single Lgn-R heterozygote is nearly the same as normal siblings in terms of plant height and leaf width. In A619, the nod-1 mutant is a dwarf, but with normal leaves and architecture. The double mutant (nod-1/nod-1; Lgn-R/+) has a synergistic phenotype, being a fraction of the size of either single mutant, highly tillered, and failing to make reproductive parts. The double mutant appears similar to Lgn-R homozygotes. In the B73 inbred, the single Lgn-R heterozygote has a visible phenotype with narrow, liguleless leaves and few tassel branches. nod-1 in B73 is bushy with small leaves and no reproductive parts. The double mutant in B73 is an enhancement of nod-1, but not changing the overall mutant architecture. Thus, we see different outcomes of the genetic interactions depending on the inbred background.
In order to understand some of these phenotypic differences, we also carried out a metabolic analysis in collaboration with an ARS group in Gainesville, Florida. The analysis was of the second leaf at maturity. Using the same plants, we also carried out RNA-sequencing on the entire shoots after removing mature leaves. The results mirrored the phenotypes in that the double mutant in A619 was similar to Lgn-R in its transcriptome and metabolome. The analysis in B73 showed that the transcriptome and metabolome of the double mutant was similar to nod-1. In both cases, the double mutants have initiated a stress response as if the plants were attacked by pathogens. This auto-immune response is seen in both the metabolomic response and the RNA-sequencing.
For Objective 2, we published a manuscript in Communications Biology describing the cloning of Tasselseed5. A University of California Berkeley graduate student carrying out research in the ARS lab finished his thesis on the other mutant under study, feminized upright narrow (fun). The thesis focuses on double mutant analysis, which suggests that fun operates in the brassinosteroid pathway and not the jasmonic acid pathway. We sent immature tassels to our ARS collaborator in Gainesville, Florida, for analysis of hormone levels. To understand in more detail the role of FUN protein, we developed an antibody. With the help of a former postdoctoral scholar we were able to purify the antibody such that a single band was detected in wild type and not in the mutant. The band is of the correct size. We used the antibody for immunolocalizations in tassels and found a signal in wild-type but not the mutant. The localization is interesting with signal along the pedicel of the spikelet, which contains the flowers. This pattern is similar to that of tasselseed1, a mutant in the jasmonic acid pathway with a similar phenotype.
Continuing research relating to Objective 2 yielded great results in the last year. We positionally cloned the Fascicled1 (Fas1) mutant, originally identified in the 1940s. The mutant has highly branched ears and tassels. We have two alleles thanks to a collaboration with researchers in Wuhan, China. Both our group and theirs mapped the mutation to two genes, one encoding ZMM8, a MADS-box gene, and the other encoding Drooping Leaf2 (DRL2), a YABBY gene. When we examined RNA levels of these genes in mutant tassels, we found both genes were expressed at high levels compared to the normal siblings. We also saw a similar result using in situ hybridization. Normally ZMM8 is expressed in floral meristems. In Fas1 mutants, the RNA is found in the inflorescence meristem. DRL2 is normally expressed in the lateral organs of floral meristems. In Fas1 mutants, it is also expressed in the inflorescence meristem.
Previous analysis of Fas1 over the years suggested instability. We found two revertant per 3000, which suggests a transposon or some sort of event that was not stable. We carried out DNA blot analysis to pursue this hypothesis. We found two to three copies of both genes in the Fas1 mutant alleles. We also discovered that one of our two revertants has lost one of the copies of ZMM8. These results suggest a tandem duplication or triplication is the cause of the dominant Fas1 phenotype.
Given the fact we do not know the ancestry of the Fas1 mutant, we explored the recent release of the maize diversity panel of inbreds to see if any lines looked like Fas1 in sequence. One of the maize inbreds, B97, also has a duplication at this locus and the sequences are nearly identical; only the intron sizes differ. We are fairly confident that B97 has no mutant inflorescence phenotype but will confirm that later this year. We are also confirming that ZMM8 and DRL2 are not misexpressed in B97. If these assumptions are born out, then it is not just the extra copies of ZMM8 and DRL2 that give rise to the mutant phenotype, but a novel promoter that leads to their misexpression.
To explore why misexpression leads to the fascicled phenotype, we focused on the YABBY gene. In most species, YABBY genes mark the abaxial (bottom) of leaves and the periphery of radial organs. Work in Arabidopsis has shown that if YABBY genes are misexpressed in the center of the meristem, the meristem dies. We hypothesized that the misexpression of DRL2 in the center of the inflorescence meristem “killed” the central zone, promoting meristematic activity in the peripheral zone, thus leading to the fascicled phenotype. To follow up on this hypothesis, we carried out RNA sequencing (RNAseq) analysis. Indeed, we found that adaxial (top) markers were expressed at low levels and the levels of other abaxial markers was high.
For Objective 3, we continue to carry out many RNAseq experiments with collaborators. RNAseq experiments include two stages of tassels, one at one millimeter (mm) and one at two to three mm when branches were just initiating or had initiated but not started making spikelets. These two stages were captured for 8 different genotypes, some mutants with increased branching and some with decreased branching. We took four of these genotypes and did laser capture analysis of the tissue in the axil of the branches at a two to four mm stage. Network analysis is being carried out by collaborators using all these datasets.
A previous drought experiment funded by the National Science Foundation involved 80 RNAseq libraries of ear, tassel and leaf tissue. These are also being analyzed by collaborators. In addition to these large experiments, we have used RNAseq to identify genes in our positional cloning experiments and these datasets will also be useful for building networks.
Accomplishments
1. Gene duplications lead to novel phenotypes. The Fascicled1 (Fas1) mutant of maize produces cobs that branch from their base with kernels missing on the inner surface, and a tassel that branches instead of growing a single growing tip. ARS researchers in Albany, California, and collaborators in Wuhan, China, mapped the mutation to a two gene interval in two separate alleles. The interval contains two transcription factors normally expressed in maize flowers. In the Fas1 mutant, the two genes have been duplicated leading to new promoters that initiate expression in the inflorescence meristem, thereby leading to the loss of apical dominance. Identifying the genes that regulate maize ear development is useful for improving maize yield.
