01 Delong 2-8 6/9/05 8:58 AM Page 2 INSIGHT REVIEW NATURE|Vol 437|15 September 2005|doi:10.1038/nature04157 Genomic perspectives in microbial oceanography Edward F. DeLong1 and David M. Karl2 The global ocean is an integrated living system where energy and matter transformations are governed by interdependent physical, chemical and biotic processes. Although the fundamentals of ocean physics and chemistry are well established, comprehensive approaches to describing and interpreting oceanic microbial diversity and processes are only now emerging. In particular, the application of genomics to problems in microbial oceanography is significantly expanding our understanding of marine microbial evolution, metabolism and ecology. Integration of these new genome-enabled insights into the broader framework of ocean science represents one of the great contemporary challenges for microbial oceanographers. Marine ecosystems are complex and dynamic. A mechanistic under- greatly aid in these efforts. The correlation between organism- and habi- standing of the susceptibility of marine ecosystems to global environ- tat-specific genomic features and other physical, chemical and biotic mental variability and climate change driven by greenhouse gases will variables has the potential to refine our understanding of microbial and require a comprehensive description of several factors. These include biogeochemical process in ocean systems. marine physical, chemical and biological interactions including All these advances — improved cultivation, environmental genomic thresholds, negative and positive feedback mechanisms and other approaches and in situ microbial observatories — promise to enhance nonlinear interactions. The fluxes of matter and energy, and the our understanding of the living ocean system. Below, we provide a brief microbes that mediate them, are of central importance in the ocean, recent history of marine microbiology and outline some of the recent yet remain poorly understood. Detailed field studies over the past results from genomic technology and the challenges to be faced for inte- three decades have established the current ‘microbial loop’ hypothesis grating these new data into the larger fabric of ocean science. wherein microbes have a central position in the conversion of dis- solved organic matter into higher trophic levels (Fig. 1). An explicit A brief historical perspective and comprehensive test of the microbial loop hypothesis, however, has In the past 30 years, there has been a remarkable growth in under- not yet been achieved. In addition, the central role of microbial activ- standing of marine microbiota (Fig. 2). During this time, ities in maintaining the oxidative state of our planet, and biogeochem- researchers have recognized the crucial role that microbes play in ical cycles other than the carbon cycle, are not well captured in the ocean ecosystems. This stems in part from technical advances, such current microbial loop hypothesis. Significant obstacles remain to be as improved epifluorescence microscope techniques and ATP-based overcome in the measurement and modelling of ocean microbial biomass metrics7, that have revealed bacterioplankton standing processes. stocks that are several orders of magnitude greater than had been For microbial oceanographers, a major challenge is to more accu- estimated by viable counting techniques8. The use of radiotracers1,9 rately incorporate the details of diversity, physiology and ecology into to estimate planktonic bacterial growth rates and turnover has also oceanographic and biogeochemical models. For example, bacterio- led to revised qualitative and quantitative models of microbial con- plankton community processes have often been modelled as being solely tributions to marine food webs10,11. Research using quantitative aut- heterotrophic1, understating the diverse physiologies and metabolic ofluorescent cell counts and flow cytometry eventually led to the processes that operate in situ. Indeed, the importance of oxygenic pho- discovery of abundant photoautotrophic picoplankton, including toautotrophic picoplankton, currently acknowledged to be dominant Synechococcus12,13 and Prochlorococcus14 species, that dominate pho- components of planktonic communities, was overlooked in early tosynthetic activities in open ocean gyres. Around the same time, oceanographic models. In addition, planktonic microbial metabolism hydrothermal vents and their rich macrofauna and microflora were has sometimes been presumed to resemble that found in common lab- discovered15, as well as the first bacterial isolates with an obligate oratory strains, some of which have questionable ecological relevance. growth requirement for elevated hydrostatic pressure16. The devel- Recent advances, such as new cultivation approaches2, cultivation-inde- opment of cultivation-independent phylogenetic surveys using pendent identification and enumeration strategies3,4, and environmen- ribosomal RNA (rRNA) sequencing3,17 and fluorescence in situ tal genomics5,6, are improving this situation. For example, we can now hybridization4 set the stage for contemporary environmental identify the genes and biochemical pathways that differentiate microbial genomic studies. Soon after their development, rRNA-based phylo- species living in different habitats. It will soon be possible to map the genetic survey techniques using the then-novel technique poly- metabolic and functional gene distributions of indigenous microbial merase chain reaction (PCR) revealed the widespread distribution species in space and time and within different oceanic provinces. The and abundance of several previously unrecognized marine micro- recent establishment of microbial observatories at selected sites will also bial groups, including Pelagibacter (also known as SAR11)18, abun- 1Department of Civil and Environmental Engineering & Division of Biological Engineering, 48-427 Massachusetts Institute of Technology, Cambridge, Massachusetts 02139, USA; 2School of Ocean and Earth Science and Technology, University of Hawaii, Honolulu, Hawaii 96822, USA. 336 © 2005 Nature Publishing Group © 2005 Nature Publishing Group 01 Delong 2-8 6/9/05 8:58 AM Page 3 NATURE|Vol 437|15 September 2005 INSIGHT REVIEW Classic food chain Microbial food web Figure 1 | Marine microbial interactions in the upper ocean. Schematic representation of the Sunlight ocean food web showing on the left the classic pathway of carbon and energy flow through photosynthetic Eukarya, to herbivores and on to Air higher trophic levels. Depicted on the right is the microbial food web, which uses energy stored in Seawater the non-living, detrital carbon pool to produce microbial biomass that can re-enter the classic CO2 CO2 pathway of carbon and energy flow. Cell- associated ectoenzymes (Ecto) enable bacteria to Viruses Viruses Archaea use high-molecular-weight (HMW) DOC in Algae addition to the more traditional low-molecular- weight (LMW) and gaseous carbon substances. Gases LMW Also shown in the microbial food web are viral (?) particles and Archaea. At the present time, there is only rudimentary knowledge of the role of Herbivores Auto- DOC E Archaea in the oceanic food web. Shown at the c hetero t bacteria bottom of this diagram is the downward flux of o Mucus particulate carbon (and energy), which is now Colloids HMW net feeders thought to fuel most subeuphotic zone processes. The classic algae-herbivore grazer pathway (left Protozoans Higher side) is most important in this regard. Adapted trophic from ref. 51. levels (?) Particulate carbon flux to mesopelagic zone dant new groups of planktonic marine Archaea19–21 and novel Genomics of cultivated marine microbial species eukaryotic picoplankton22, to list a few examples. Adaptation to the unique characteristics of their environment defines More recently, the prevalence of bacteriochlorophyll-containing23 the essence of marine microbial species. High salinity, low and variable and rhodopsin-containing24 bacterioplankton was recognized, pro- nutrient concentrations and steep gradients of temperature, light, pres- viding new perspectives on the nature of light-use strategies in ocean sure and nutrients with increasing depth are some of the environmen- surface waters. Surprisingly, it has only very recently been realized that tal variables that influence the evolution of form and function in viral particles can exceed total microbial cell numbers by an order of microbial plankton. The adaptive instructions for coping with the magnitude in marine plankton25 and that they represent potentially marine environment are encoded in the genomic blueprints of ocean important vectors of bacterioplankton mortality and lateral gene microbiota. Genome sequences are therefore useful tools for explor- transfer26. ing the relationship between genotype, phenotype and environment in Many new marine microbial species continue to be described, and native marine microbial species (Table 1). A few examples provided unexpected physiological and biochemical properties await further below illustrate how recent sequence information has contributed to discovery and description. There remains much to learn about the our understanding of the adaptations, physiology and ecology of distributions, variability and biogeochemical influence of naturally marine microbes. occurring microbes in the sea. To this end, new technologies are now providing life science and ocean science with remarkable new Genomic perspective on the forests of the sea research opportunities. In