Steranes and hopanes are organic compounds found in ancient rocks that have been used to date the rise of oxygenic photosynthesis. Because of their unique carbon skeletons, these molecules can unambiguously be recognized as molecular fossils of steroids and hopanoids (steroid analogs in bacteria), important constituents of cell membranes. While key steps in the biosynthesis of steroids require O2, hopanoid biosynthesis does not. Modern steroids and hopanoids are structurally diverse, yet only their carbon skeletons are preserved after diagenesis. Remarkably, the total amount of hopanes trapped within ancient rocks is thought to be roughly equivalent to the amount of organic carbon present on Earth today. One of the most important geostable hopanoid modifications is methylation at C-2, and molecular fossils of this type are called 2-methylhopanes (deriving from 2-methylbacteriohopanepolyols, 2-MeBHPs, in modern cells). Cyanobacteria—bacteria that engage in oxygenic photosynthesis—used to be considered the only quantitatively important source of 2-MeBHPs; accordingly, the occurrence of 2-methylhopanes in sediments that are 2.7 billion years old was taken as evidence that photosynthetically derived O2 first appeared on Earth at least that long ago. But because several independent geochemical proxies indicate that a major global redox transition did not occur until several hundred million years later, we decided, in collaboration with organic geochemists, to examine key assumptions underpinning the use of hopanes and steranes as O2 biomarkers.

When we began, although a considerable amount was known about steroid cell biology, what the O2 threshold necessary for steroid biosynthesis is—and the impact this value has on models of atmospheric oxygenation—was unclear. By carefully controlling the O2 available to our cultures, we found that steroid biosynthesis can occur with dissolved O2 concentrations in the nanomolar range. This low requirement helps explain the temporal decoupling between the sterane biomarker record of O2 utilization and the dating of a global redox transition: models of atmospheric oxygenation are consistent with the hypothesis that O2 could have cycled as a trace gas in the marine environment for millions of years prior to its atmospheric accumulation. Key to this discovery was our investment in the ability to culture diverse bacteria in hypoxic and anoxic environments where O2 could be precisely measured. This ability also enabled the isolation of Rhodopseudomonas palustris TIE-1, an anoxygenic phototroph that we serendipitously discovered could produce 2-MeBHPs in as great abundance as cyanobacteria under certain conditions.

Because R. palustris grows quickly and is metabolically versatile, we developed it into a model system in which to study hopanoid cell biology. We elucidated the biosynthetic pathway for diverse hopanoids, the transporter responsible for localizing hopanoids to the outer membrane, and the mechanism and conditions responsible for regulating 2-MeBHP biosynthesis. Our discovery that the C-2 hopanoid methylase (HpnP) is well conserved among all 2-MeBHP–producing bacteria allowed us to circumvent the problem of conditional 2-MeBHP production by using the hpnP gene to identify 2-MeBHP production capacity in other microbial genomes and metagenomes. This survey not only revealed that only a minority of cyanobacteria make 2-MeBHPs but also revealed that a statistically significant correlation exists in modern environments between 2-MeBHP production capacity and an ecological niche defined by low O2, high osmolytes, and sessile microbial communities. In modern environments, this tracks with microenvironments found in microbial mats, stromatolites, and the rhizosphere; relevant to the latter, the occurrence of hpnP is significantly enriched in the genomes of well-characterized plant symbionts.

Motivated by this new correlation, we have expanded our model system set to include Nostoc punctiforme and Bradyrhizobium japonicum, genetically tractable 2-MeBHP–producing bacteria with well-characterized plant partners. In parallel with our work in R. palustris, we are exploring the regulation of hopanoid production by these strains and how hopanoid production affects diverse phenotypes. This has required us to develop novel methods to detect and quantify hopanoids both in single cells and from lipid mixtures extracted from bulk cultures. Using these methods, we are systematically characterizing the membrane composition of diverse hopanoid-producing wild-type and mutant strains grown in vitro and in planta. These results are informing biophysical studies to test the effects of hopanoids on membrane fluidity, permeability, and curvature. Finally, in collaboration with chemical biologists, we are building a molecular toolkit to identify proteins and other biomolecules that interact with hopanoids.

It is now clear that while the O2 requirement for sterane biosynthesis is compatible with other proxies for dating the rise of O2, 2-methylhopanes cannot be used as biomarkers of O2 photosynthesis. Our new goal is to provide a better interpretation of sedimentary hopanes by gaining a deeper understanding of their modern counterparts. Do hopanoids facilitate plant-microbe symbioses in specific ways? With which other membrane components do they interact? What explains their phylogenetic distribution? Unlike steroids in eukaryotes, hopanoid production by bacteria is only essential under certain conditions, offering the possibility of using bacterial systems to explore fundamental questions of membrane homeostasis that are not as readily addressed in eukaryotes.

Using the present to inform the past: interpreting ancient molecular fossils by understanding their modern counterparts