Arabidopsis 2010: Deriving the Gene Circuitry and Network Motifs of the Arabidopsis Defense Response

为了确定基因网络的基因功能，将从拟南芥白粉病侵染的时间和空间大规模的mRNA谱中推导出植物对病原菌的防御回路。激光捕获显微切割将被用来以一定的采样率和足够的时间点从不同种群的植物细胞(包括受感染的表皮细胞和邻近的表皮细胞和叶肉细胞)中分离出RNA，以捕捉潜在的生物过程。为了分析这些信使核糖核酸的表达数据，将开发新的统计方法。例如，时间对齐算法将用于在实验之间同步数据，并评估改变的响应时间模式，包括时间延迟和周期压缩/扩展。通过对植物防御回路进行建模，可以对影响感染进展的调控因素和下游产物进行严格评估。连续表示将被用来派生响应的电路，允许人们发现与不同的功能响应相关联的网络节点和子节点。预测基因功能的实验测试将集中在水杨酸(SA)影响的网络的亚节点上的基因，这些基因是使用SA生物合成突变体阐明的。此外，定义这些网络和节点允许发现和评估网络主题，例如用于解决不同环境和物种中的常见生物“问题”的前馈环路。这些在空间和时间上分辨的表达数据、统计和计算方法和工具以及识别的防御电路和网络基序都将对拟南芥社区具有重要价值。以前从整个叶片样本和很少的时间点收集的数据集无法解决防御反应的大部分复杂性。这些努力将产生大量关于基因功能和调控的预测。模型驱动的实验将集中在受SA影响的节点和子节点中的基因。这些基因并不是先验的，但将通过建模工作来确定。可能的基因包括：参与诱导ICS1簇和PR1簇的转录因子(S)，推测的SA葡萄糖转移酶，参与异氨基甲酸合成SA的第二种酶的候选基因，以及可能的转录抑制因子，负责依赖SA对PDF1.2簇的抑制。表达数据将存放在公共数据库中，例如作为国家科学基金会2010年拟南芥奖授予X.Dong和共同PIs的一部分而开发的综合微阵列数据库系统，并将免费提供使用。此外，开发的算法和计算工具将可供下载，并将作为工具包括在免费的微阵列分析平台BioConductor中。有关该项目的信息可在http://plantbio.berkeley.edu:16080/~wildermuth/.Broader Impact上获得。这项研究将通过为分子、遗传和生化数据提供正式的数学框架，更全面地了解植物与病原体之间的相互作用。所阐明的植物-病原菌相互作用的策略可能也适用于其他寄主-病原菌的相互作用。特别是，已识别的功能控制模块很可能在不同的病理系统之间共享。将向工厂和一般科学界广泛提供为分析和建模工作开发的信息和统计方法和工具。此外，年轻的数学家、工程师和实验生物学家的密切合作和交叉培训产生了真正的跨学科科学家，他们在使用定量和基于系统的方法解决生物学问题方面具有独特的地位。

In order to determine the gene function for a network of genes, the circuitry of plant defense against pathogen will be derived from temporal and spatial large-scale mRNA profiling of powdery mildew infection of Arabidopsis. Laser capture microdissection will be utilized to isolate RNA from distinct populations of plant cells (including infected epidermal cells and neighboring epidermal and mesophyll cells) at a sampling rate and with sufficient time points to capture the underlying biological processes. To analyze this mRNA expression data, novel statistical approaches will be developed. For example, time-alignment algorithms will be used to synchronize data among experiments and evaluate altered temporal patterns of response including time delays and period compression/expansion. Modeling the circuitry of plant defense allows for the rigorous assessment of the regulatory factors and downstream products impacting the progression of infection. Continuous representations will be used to derive the circuitry of response, allowing one to uncover network nodes and subnodes associated with distinct functional responses. Experimental testing of predicted gene function will focus on genes in subnodes of salicylic acid (SA)-impacted networks elucidated using an SA biosynthetic mutant. In addition, defining these networks and nodes allows for discovery and assessment of network motifs such as feed-forward loops used to solve a common biological "problem" in diverse circumstances and species. This spatially and temporally resolved expression data, the statistical and computational approaches and tools, and the identified defense circuitry and network motifs will all be of significant value to the Arabidopsis community. Previous datasets collected from whole leaf samples and few time points were unable to resolve much of the complexity of the defense response. Numerous predictions of gene function and regulation will emerge from these efforts. Model-driven experiments will focus on genes in SA-impacted nodes and subnodes. These genes are not known a priori, but will be determined from the modeling efforts. Likely genes include: transcription factor(s) involved in the induction of ICS1 and PR1 clusters, a putative SA glucosyltransferase, candidate genes for the second enzyme involved in the synthesis of SA from isochorismate, and a putative transcriptional repressor responsible for the SA-dependent repression of the PDF1.2 cluster. The expression data will be deposited in public databases such as the Integrated Microarray Database System being developed as part of the NSF Arabidopsis 2010 award to X. Dong and co-PIs and will be freely available for use. In addition, developed algorithms and computational tools will be available for download and will be included as tools in Bioconductor, a free microarray analysis platform. Information about this project is available at http://plantbio.berkeley.edu:16080/~wildermuth/.Broader Impacts This research will result in a more comprehensive understanding of plant-pathogen interactions by providing a formal mathematical framework for molecular genetic and biochemical data. The strategies elucidated for this plant-pathogen interaction are likely applicable to other host-pathogen interactions. In particular, identified functional control modules are likely to be shared across pathosystems. The informatic and statistical methodologies and tools developed for the analysis and modeling efforts will be made widely available to both the plant and general scientific communities. In addition, the intimate collaboration and cross-training of young mathematicians, engineers, and experimental biologists yields truly interdisciplinary scientists uniquely positioned to address biological questions using quantitative and systems-based approaches.