Marine viral dynamics and incorporation into microbial association networks

海洋微生物极其丰富，是全球碳、氮、磷、铁生物地球化学循环的主要参与者和驱动力。在过去的二十年里，我们了解到病毒是海洋系统中普遍存在的元素，对细胞海洋生物具有重要的生态、生物地球化学、遗传和进化影响，但我们对海洋病毒群落结构的动态以及它与宿主（主要是细菌和浮游植物）的群落结构的关系知之甚少。这些信息对于建立关于病毒的作用以及这些作用如何随时间和空间变化的适当概念和实际模型至关重要。这个项目的目标是：(1)主要是表征自然病毒群落的一个重要子集及其动态，以及细菌宿主群落，因为它们在南加州大学经过充分研究的海洋微生物观测站（洛杉矶和圣卡塔利娜岛之间的中途）每天到每月的时间尺度上发生变化，测试关于重复模式、宿主范围效应和分类时间关系的假设；(2)其次，通过统计将特定类型的病毒与特定的潜在宿主联系起来，将这些病毒纳入微生物关联网络。本研究的方法包括：(a)每天、每周和每月巢式收集细菌和病毒的核酸样本，(b)从以前容易在海水中发现的几个特征较好的广泛病毒群（即t4样肌病毒、t7样足病毒）中扩增保守基因作为代系统发育标记，以及细菌rRNA基因(c)通过群落指纹图谱筛选后，从混合扩增产物中进行广泛测序；(d)以不同的分辨率将序列或指纹片段划分为操作分类单元（otu）； (e)用统计方法评估结果，以检查时间模式、与其他病毒otu、细菌、原生生物（每月一次）和环境参数的关系（包括时差关系）；(f)将病毒otu以数学方式纳入微生物关联网络。现有的微生物观测站已经每月收集环境参数、细菌和原生生物的数据，因此病毒工作是对该项目的补充，提供了一个主要的增值部分。同样，该项目将向微生物观测站添加精选的每日和每周微生物数据。从文献数据和PI的初步结果显示，他们有技术和能力来实现第一个目标，据我们所知，这将是第一个这样的范围和类型的数据集。研究人员已经在2006年发表文章称，该地点5m深度的细菌群落在细菌群落组成上显示出可预测的重复年循环，因此对可预测的重复病毒群落的期望并非不合理。他们也有一些初步的数据显示一些重复的病毒事件。第二个目标要求病毒和其他测量参数之间确实存在显著的统计关系，PI预计会是这种情况，但当然无法预测；如果它们不能被证明，这一结果本身将提供信息，并将限制微生物/病毒相互作用的可能模式。更广泛的影响：该项目将有一个重要的培训组成部分，与研究生做相当数量的工作（现场，实验室，生物信息学和写作部分）作为她的论文研究的一部分。调查人员还将把该项目与当地一所中学的科学教师培训计划联系起来，将结果纳入培训模块，并在我们的海洋实验室进行社区外展，中学生经常访问该实验室。与更广泛的全球社区的联系将包括一个项目网站，其中包含协议、工作的非技术和技术描述、结果和结论。

Marine microbes are tremendously abundant and are major players and driving forces in global biogeochemical cycles of carbon, nitrogen, phosphorus, and iron. We learned over the past two decades that viruses are pervasive elements in marine systems, with significant ecological, biogeochemical, genetic, and evolutionary effects on cellular marine organisms, but we have remarkably little information about the dynamics of marine viral community structure and how it relates to the community structure of their hosts (largely bacteria and phytoplankton). Such information is critical for developing proper conceptual and practical models of the roles of viruses and how these change over time and space. The goals of this project are (1) primarily, to characterize a significant subset of the natural virus community and its dynamics, along with bacterial host communities, as they change over daily to monthly time scales at the USC well-studied marine Microbial Observatory site (midway between Los Angeles and Santa Catalina Island), testing hypotheses regarding repeating patterns, host range effects, and taxa-time relationships, and (2) secondarily, to incorporate these viruses into microbial association networks by statistically connecting particular types of viruses to specific potential hosts. Approaches for this study include: (a) nested daily, weekly, and monthly collection of bacteria and viruses for nucleic acid samples, (b) amplification of conserved genes, as proxy phylogenetic markers, from a few moderately-well-characterized broad viral groups previously readily found in seawater (i.e. the T4-like myoviruses, T7-like podoviruses), as well as bacterial rRNA genes (c) extensive sequencing, after screening by community fingerprinting, from the mixed amplified products, (d) binning of the sequences or fingerprint fragments into operational taxonomic units (OTUs) at different levels of resolution, (e) evaluation of the results with statistical approaches to examine temporal patterns, relationships (including time-lagged ones) with other viral OTUs, bacteria, protists (monthly only), and environmental parameters, (f) incorporating the viral OTUs mathematically into microbial association networks. Data on environmental parameters, bacteria, and protists are already being collected monthly for an existing Microbial Observatory, so the viral work is complementary to this project, providing a major value-added component. Similarly, this project will add selected daily and weekly microbial data to the Microbial Observatory. Data from the literature and from the PI's preliminary results show they have the technology and capability to meet the first goal, and to our knowledge this would be the first such data set of its scope and kind. The investigators have already published in 2006 that the bacterial communities at the 5m depth of this site show a predictable repeating annual cycle in bacterial community composition, so the expectation of a predictable repeating viral community is not unreasonable. They also have some preliminary data showing some repeated viral occurrences. The second goal requires that there are indeed significant statistical relationships between the viruses and other measured parameters, which the PI anticipates to be the case, but of course cannot predict; if they cannot be demonstrated, this result itself would be informative and would constrain the possible modes of microbial/viral interactions.BROADER IMPACTS: The project will have a significant training component, with a graduate student doing a considerable amount of the work (field, lab, bioinformatic, and writing components) as part of her dissertation research. The investigators also will link the project to a local secondary school science teacher training program, incorporating results into a training module, and also to community outreach at our marine lab, often visited by secondary school students. Outreach to the broader global community will include a project website with protocols, nontechnical and technical descriptions of the work, results, and conclusions.