Vitamins in the sea.

N early half of all photosynthesis happens in the oceans, in communities of microscopic microbial plankton that can vary greatly in time and space. On the one hand these communities are ephemeral, changing with changing ocean conditions, but on the other they can be counted on to cycle in predictable patterns driven by seasons, weather, and ocean currents (1, 2). In PNAS, Sañudo-Wilhelmy et al. (3) explore vitamin concentrations in the oceans, showing that B vitamins are distributed in complex patterns and offering the idea that they might be important factors controlling microbial plankton community composition. The issue considered is not gross ocean productivity but rather the composition of plankton communities, which can impact geochemical cycles. The search for growth factors that control community composition is an old tradition in oceanography that often appears under the heading “bottom-up control,” to differentiate it from predation by viruses, protists, and zooplankton, which can alter community composition by selective cropping (4). Compounds of phosphorus and nitrogen are widely understood to be the major drivers of ocean productivity and plankton community structure, but, famously, in the 1970s iron was recognized as running a close third. Many other elements contend on the list, including silica, which is required by diatoms as a structural component of their frustules, and a number of trace metals that cells use as enzyme cofactors. The distribution of trace metals in the ocean was explored under the auspices of the National Science Foundation in the “Geotracers” program, which revealed striking patterns in many biologically important metals, including Co, Zn, and Cu (5). Despite recognition of the importance of iron and the success of Geotracers, today only part of the variation in plankton communities can be predicted from chemistry and physics. A thought on many minds is that biological molecules dissolved in the water column might be a missing piece of the puzzle (6). In pursuit of this idea, oceanographers are turning their attention to vitamins. The amount of data directly supporting the hypothesis that vitamins control marine microbial communities remains small. One of the most cited reports demonstrated the stimulation of phytoplankton photosynthesis in the Ross Sea by the addition of both iron and vitamin B12 (7). The field is primed for further work aimed at getting to the bottom of this question. Vitamins are truly “old school,” having been topics of extensive research on nutrition and long recognized as requirements for many phytoplankton strains cultured in laboratories (8). Many microbial plankton species are auxotrophic for one or more vitamins—that is, they lack one or more vitamin biosynthetic pathways and therefore must gather essential vitamins from the environment (9). Possession of a vitamin biosynthetic pathway means independence, at the cost of synthesis. It is not difficult to imagine how this might influence the species distributions in plankton communities: a cell that needs a vitamin can only thrive in a community that provides it. Vitamin traffic may explain some of the network connectedness that has been observed in ocean microbial plankton communities (10). Connectedness has become an important concept in ecology because it can help explain the responses of biological communities to stresses (11). In oceanography connectedness is manifested as correlations in the abundances of species. When two populations are correlated, it could be because they are interacting directly or because they are both controlled by the same environmental factor. Most often connectedness is associated with trophic interactions, for example, in predator–prey relationships. Microbial plankton communities seem to be highly connected, but there is little direct evidence about the nature of the interactions. Vitamins are perhaps the most wellknown and historical examples of biological molecules that are involved in intracellular traffic, but other examples have emerged of metabolic requirements being outsourced to the community (12, 13). In our work with the highly abundant SAR11 clade of marine chemoheterotrophs, we sought to understand how genome streamlining impacted metabolism. Our objectives were both practical—to learn how to grow these enigmatic cells, and theoretical—to understand what metabolic functions were dispensable as selection shrank genome size in these ultra-efficient, small cells. Pelagibacter sp., the main experimental strains of SAR11, have very unusual requirements for reduced sulfur compounds (14) and for glycine or glycine precursors (15). Both requirements were traced to genome reduction. Our interpretation of these observations is that evolution is a good integrator—in the periodically fluctuating ocean milleau, fluxes of reduced sulfur and glycine compounds are sufficient to meet SAR11 requirements most of the time, so the fitness advantages of smaller genomes and lower cell replication costs offset the potential gain in fitness that would come from autonomy when reduced sulfur and glycine compounds are in short supply. Interestingly, Pelagibacter seems to synthesize at least some of its own vitamins (16). This suggests that these compounds, which are just as essential for cell growth as glycine and reduced sulfur, must not be sufficiently available to offset the costs of maintaining biosynthetic pathways for these compounds in the genome. The presence of vitamin biosynthetic pathways in highly streamlined cells suggests that Sañudo-Wilhelmy et al. (3) are on the right track, implicating vitamin distributions in ecology. If vitamins were always available from the environment, why would organisms retain these genes in the B12 B7 B1