The rise of atmospheric O(2) was a milestone in the history of life. Although O(2) itself is not a climate-active gas, its appearance would have removed a methane greenhouse present on the early Earth and potentially led to dramatic cooling. Moreover, by fundamentally altering the biogeochemical cycles of C, N, S and Fe, its rise first in the atmosphere and later in the oceans would also have had important indirect effects on Earth's climate. Here, we summarize major lines of evidence from the geological literature that pertain to when and how O(2) first appeared in significant amounts in the atmosphere. On the early Earth, atmospheric O(2) would initially have been very low, probably <10(-5) of the present atmospheric level. Around 2.45 billion years ago, atmospheric O(2) rose suddenly in what is now termed the Great Oxidation Event. While the rise of oxygen has been the subject of considerable attention by Earth scientists, several important aspects of this problem remain unresolved. Our goal in this review is to provide a short summary of the current state of the field, and make the case that future progress towards solving the riddle of oxygen will benefit greatly from the involvement of molecular biologists.

Hopanes preserved in both modern and ancient sediments are recognized as the molecular fossils of bacteriohopanepolyols, pentacyclic hopanoid lipids. Based on the phylogenetic distribution of hopanoid production by extant bacteria, hopanes have been used as indicators of specific bacterial groups and/or their metabolisms. However, our ability to interpret them ultimately depends on understanding the physiological roles of hopanoids in modern bacteria. Toward this end, we set out to identify genes required for hopanoid biosynthesis in the anoxygenic phototroph Rhodopseudomonas palustris TIE-1 to enable selective control of hopanoid production. We attempted to delete 17 genes within a putative hopanoid biosynthetic gene cluster to determine their role, if any, in hopanoid biosynthesis. Two genes, hpnH and hpnG, are required to produce both bacteriohopanetetrol and aminobacteriohopanetriol, whereas a third gene, hpnO, is required only for aminobacteriohopanetriol production. None of the genes in this cluster are required to exclusively synthesize bacteriohopanetetrol, indicating that at least one other hopanoid biosynthesis gene is located elsewhere on the chromosome. Physiological studies with the different deletion mutants demonstrated that unmethylated and C(30) hopanoids are sufficient to maintain cytoplasmic but not outer membrane integrity. These results imply that hopanoid modifications, including methylation of the A-ring and the addition of a polar head group, may have biologic functions beyond playing a role in membrane permeability.

Abstract2-Methylhopanes, molecular fossils of 2-methylbacteriohopanepolyol (2-MeBHP) lipids, have been proposed as biomarkers for cyanobacteria, and by extension, oxygenic photosynthesis. However, the robustness of this interpretation is unclear, as 2-methylhopanoids occur in organisms besides cyanobacteria and their physiological functions are unknown. As a first step towards understanding the role of 2-MeBHP in cyanobacteria, we examined the expression and intercellular localization of hopanoids in the three cell types of Nostoc punctiforme: vegetative cells, akinetes, and heterocysts. Cultures in which N. punctiforme had differentiated into akinetes contained approximately 10-fold higher concentrations of 2-methylhopanoids than did cultures that contained only vegetative cells. In contrast, 2-methylhopanoids were only present at very low concentrations in heterocysts. Hopanoid production initially increased 3-fold in cells starved of nitrogen but returned to levels consistent with vegetative cells within two weeks. Vegetative and akinete cell types were separated into cytoplasmic, thylakoid, and outer membrane fractions; the increase in hopanoid expression observed in akinetes was due to a 34-fold enrichment of hopanoid content in their outer membrane relative to vegetative cells. Akinetes formed in response either to low light or phosphorus limitation, exhibited the same 2-methylhopanoid localization and concentration, demonstrating that 2-methylhopanoids are associated with the akinete cell type per se. Because akinetes are resting cells that are not photosynthetically active, 2-methylhopanoids cannot be functionally linked to oxygenic photosyntheis in N. punctiforme.