Publikationen - Molekulare Signalverarbeitung

BackgroundThe formation of flowers is one of the main model systems to elucidate the molecular mechanisms that control developmental processes in plants. Although several studies have explored gene expression during flower development in the model plant Arabidopsis thaliana on a genome-wide scale, a continuous series of expression data from the earliest floral stages until maturation has been lacking. Here, we used a floral induction system to close this information gap and to generate a reference dataset for stage-specific gene expression during flower formation.ResultsUsing a floral induction system, we collected floral buds at 14 different stages from the time of initiation until maturation. Using whole-genome microarray analysis, we identified 7,405 genes that exhibit rapid expression changes during flower development. These genes comprise many known floral regulators and we found that the expression profiles for these regulators match their known expression patterns, thus validating the dataset. We analyzed groups of co-expressed genes for over-represented cellular and developmental functions through Gene Ontology analysis and found that they could be assigned specific patterns of activities, which are in agreement with the progression of flower development. Furthermore, by mapping binding sites of floral organ identity factors onto our dataset, we were able to identify gene groups that are likely predominantly under control of these transcriptional regulators. We further found that the distribution of paralogs among groups of co-expressed genes varies considerably, with genes expressed predominantly at early and intermediate stages of flower development showing the highest proportion of such genes.ConclusionsOur results highlight and describe the dynamic expression changes undergone by a large number of genes during flower development. They further provide a comprehensive reference dataset for temporal gene expression during flower formation and we demonstrate that it can be used to integrate data from other genomics approaches such as genome-wide localization studies of transcription factor binding sites.

The INDOLE-3-BUTYRIC ACID RESPONSE5 (IBR5) gene encodes a dual specificity phosphatase that regulates plant auxin responses. IBR5 has been predicted to generate two transcripts through alternative splicing, but alternative splicing of IBR5 has not been confirmed experimentally. The previously characterized ibr5-1 null mutant exhibits many auxin related defects such as auxin insensitive primary root growth, defective vascular development, short stature and reduced lateral root development. However, whether all these defects are caused by the lack of phosphatase activity is not clear. Here we describe two new auxin insensitive IBR5 alleles, ibr5-4, a catalytic site mutant, and ibr5-5, a splice site mutant. Characterization of these new mutants indicates that IBR5 is post-transcriptionally regulated to generate two transcripts, AT2G04550.1 and AT2G04550.3, and consequently two IBR5 isoforms, IBR5.1 and IBR5.3. The IBR5.1 isoform exhibits phosphatase catalytic activity that is required for both proper degradation of Aux/IAA proteins and auxin-induced gene expression. These two processes are independently regulated by IBR5.1. Comparison of new mutant alleles with ibr5-1 indicates that all three mutant alleles share many phenotypes. However, each allele also confers distinct defects implicating IBR5 isoform specific functions. Some of these functions are independent of IBR5.1 catalytic activity. Additionally, analysis of these new mutant alleles suggests that IBR5 may link ABP1 and SCFTIR1/AFBs auxin signaling pathways.

Microarrays are commonly applied to study the transcriptome of specific species. However, many available microarrays are restricted to model organisms, and the design of custom microarrays for other species is often not feasible. Hence, transcriptomics approaches of non-model organisms as well as comparative transcriptomics studies among two or more species often make use of cost-intensive RNAseq studies or, alternatively, by hybridizing transcripts of a query species to a microarray of a closely related species. When analyzing these cross-species microarray expression data, differences in the transcriptome of the query species can cause problems, such as the following: (i) lower hybridization accuracy of probes due to mismatches or deletions, (ii) probes binding multiple transcripts of different genes, and (iii) probes binding transcripts of non-orthologous genes. So far, methods for (i) exist, but these neglect (ii) and (iii). Here, we propose an approach for comparative transcriptomics addressing problems (i) to (iii), which retains only transcript-specific probes binding transcripts of orthologous genes. We apply this approach to an Arabidopsis lyrata expression data set measured on a microarray designed for Arabidopsis thaliana, and compare it to two alternative approaches, a sequence-based approach and a genomic DNA hybridization-based approach. We investigate the number of retained probe sets, and we validate the resulting expression responses by qRT-PCR. We find that the proposed approach combines the benefit of sequence-based stringency and accuracy while allowing the expression analysis of much more genes than the alternative sequence-based approach. As an added benefit, the proposed approach requires probes to detect transcripts of orthologous genes only, which provides a superior base for biological interpretation of the measured expression responses.

F-box proteins (FBPs) represent one of the largest and fastest evolving gene/protein families in the plant kingdom. The FBP superfamily can be divided in several subfamilies characterized by different C-terminal protein-protein interaction domains that recruit targets for proteasomal degradation. Hence, a clear picture of their phylogeny and molecular evolution is of special interest for the general understanding of evolutionary histories of multi-domain and/or large protein families in plants. In an effort to further understand the molecular evolution of F-box family proteins, we asked whether the largest subfamily in Arabidopsis thaliana, which carries a C-terminal F-box associated domain (FBA proteins) shares evolutionary patterns and signatures of selection with other FBPs. To address this question, we applied phylogenetic and molecular evolution analyses in combination with the evaluation of transcriptional profiles. Based on the 2219 FBA proteins we de novo identified in 34 completely sequenced plant genomes, we compared their evolutionary patterns to a previously analyzed large subfamily carrying C-terminal kelch repeats. We found that these two large FBP subfamilies generally tend to evolve by massive waves of duplication, followed by sequence conservation of the F-box domain and sequence diversification of the target recruiting domain. We conclude that the earlier in evolutionary time a major wave of expansion occurred, the more pronounced these selection signatures are. As a consequence, when performing cross species comparisons among FBP subfamilies, significant differences will be observed in the selective signatures of protein-protein interaction domains. Depending on the species, the investigated subfamilies comprise up to 45% of the complete superfamily, indicating that other subfamilies possibly follow similar modes of evolution.

The COP9 signalosome (CSN) is an eight subunit protein complex conserved in all higher eukaryotes. In Arabidopsis thaliana, the CSN regulates auxin response by removing the ubiquitin-like protein NEDD8/RUB1 from the CUL1 subunit of the SCFTIR1/AFB ubiquitin-ligase (deneddylation). Previously described null mutations in any CSN subunit result in the pleiotropic cop/det/fus phenotype and cause seedling lethality, hampering the study of CSN functions in plant development. In a genetic screen to identify enhancers of the auxin response defects conferred by the tir1-1 mutation, we identified a viable csn mutant of subunit 3 (CSN3), designated eta7/csn3-3. In addition to enhancing tir1-1 mutant phenotypes, the csn3-3 mutation alone confers several phenotypes indicative of impaired auxin signaling including auxin resistant root growth and diminished auxin responsive gene expression. Unexpectedly however, csn3-3 plants are not defective in either the CSN-mediated deneddylation of CUL1 or in SCFTIR1-mediated degradation of Aux/IAA proteins. These findings suggest that csn3-3 is an atypical csn mutant that defines a novel CSN or CSN3-specific function. Consistent with this possibility, we observe dramatic differences in double mutant interactions between csn3-3 and other auxin signaling mutants compared to another weak csn mutant, csn1-10. Lastly, unlike other csn mutants, assembly of the CSN holocomplex is unaffected in csn3-3 plants. However, we detected a small CSN3-containing protein complex that is altered in csn3-3 plants. We hypothesize that in addition to its role in the CSN as a cullin deneddylase, CSN3 functions in a distinct protein complex that is required for proper auxin signaling.

The apical plasma membrane of pollen tubes contains different PI4P 5-kinases that all produce phosphatidylinositol-4,5-bisphosphate [PtdIns(4,5)P2] but exert distinct cellular effects. In the present example, overexpression of Arabidopsis AtPIP5K5 or tobacco NtPIP5K6-1 caused growth defects previously attributed to increased pectin secretion. In contrast, overexpression of Arabidopsis AtPIP5K2 caused apical tip swelling implicated in altering actin fine structure in the pollen tube apex. AtPIP5K5, NtPIP5K6-1, and AtPIP5K2 share identical domain structures. Domains required for correct membrane association of the enzymes were identified by systematic deletion of N-terminal domains and subsequent expression of fluorescence-tagged enzyme truncations in tobacco pollen tubes. A variable linker region (Lin) contained in all PI4P 5-kinase isoforms of subfamily B, but not conserved in sequence, was recognized to be necessary for correct subcellular localization of AtPIP5K5, NtPIP5K6-1, and AtPIP5K2. Deletion of N-terminal domains including the Lin domain did not impair catalytic activity of recombinant AtPIP5K5, NtPIP5K6-1, or AtPIP5K2 in vitro; however, the presence of the Lin domain was necessary for in vivo effects on pollen tube growth upon overexpression of truncated enzymes. Overexpression of catalytically inactive variants of AtPIP5K5, NtPIP5K6-1, or AtPIP5K2 did not influence pollen tube growth, indicating that PtdIns(4,5)P2 production rather than structural properties of PI4P 5-kinases was relevant for the manifestation of growth phenotypes. When Lin domains were swapped between NtPIP5K6-1 and AtPIP5K2 and the chimeric enzymes overexpressed in pollen tubes, the chimeras reciprocally gained the capabilities to invoke tip swelling or secretion phenotypes, respectively. The data indicate that the Lin domain directed the enzymes into different regulatory contexts, possibly contributing to channeling of PtdIns(4,5)P2 at the interface of secretion and actin cytoskeleton.

The mitogen-activated protein kinase (MAPK) pathway is a three-tier signaling cascade that transmits cellular information from the plasma membrane to the cytoplasm where it triggers downstream responses. The MAPKs represent the last step in this cascade and are activated when both tyrosine and threonine residues in a conserved TxY motif are phosphorylated by MAPK kinases, which in turn are themselves activated by phosphorylation by MAPK kinase kinases. To understand the molecular evolution of MAPKs in the plant kingdom, we systematically conducted a Hidden-Markov-Model based screen to identify MAPKs in 13 completely sequenced plant genomes. In this analysis, we included green algae, bryophytes, lycophytes, and several mono- and eudicotyledonous species covering >800 million years of evolution. The phylogenetic relationships of the 204 identified MAPKs based on Bayesian inference facilitated the retraction of the sequence of emergence of the four major clades that are characterized by the presence of a TDY or TEY-A/TEY-B/TEY-C type kinase activation loop. We present evidence that after the split of TDY- and TEY-type MAPKs, initially the TEY-C clade emerged. This was followed by the TEY-B clade in early land plants until the TEY-A clade finally emerged in flowering plants. In addition to these well characterized clades, we identified another highly conserved clade of 45 MAPK-likes, members of which were previously described as Mak-homologous kinases. In agreement with their essential functions, molecular population genetic analysis of MAPK genes in Arabidopsis thaliana accessions reveal that purifying selection drove the evolution of the MAPK family, implying strong functional constraints on MAPK genes. Closely related MAPKs most likely subfunctionalized, a process in which differential transcriptional regulation of duplicates may be involved.

SKP1-Cullin1-F-box protein (SCF) ubiquitin-ligases regulate numerous aspects of eukaryotic growth and development. Cullin-Associated and Neddylation-Dissociated (CAND1) modulates SCF function through its interactions with the CUL1 subunit. Although biochemical studies with human CAND1 suggested that CAND1 plays a negative regulatory role by sequestering CUL1 and preventing SCF complex assembly, genetic studies in Arabidopsis have shown that cand1 mutants exhibit reduced SCF activity, demonstrating that CAND1 is required for optimal SCF function in vivo. Together, these genetic and biochemical studies have suggested a model of CAND1-mediated cycles of SCF complex assembly and disassembly. Here, using the SCFTIR1 complex of the Arabidopsis auxin response pathway, we test the SCF cycling model with Arabidopsis mutant derivatives of CAND1 and CUL1 that have opposing effects on the CAND1–CUL1 interaction. We find that the disruption of the CAND1–CUL1 interaction results in an increased abundance of assembled SCFTIR1 complex. In contrast, stabilization of the CAND1–CUL1 interaction diminishes SCFTIR1 complex abundance. The fact that both decreased and increased CAND1–CUL1 interactions result in reduced SCFTIR1 activity in vivo strongly supports the hypothesis that CAND1-mediated cycling is required for optimal SCF function.