A Comprehensive Study of Biohopanoid Production in Alphaproteobacteria: Biosynthetic, Chemotaxonomical, and Geobiological Implications.

Bacteriohopanepolyol derivatives (BHPDs) and dia- and catagenetic products formed from these bacterial membrane modifiers are extensively used as biomarkers in molecular ecological and geoscience studies. Some BHPDs can be assigned to specific phylogenetic bacterial groups. With the vastly increasing availability of complete bacterial genomes, hopanoid production can be readily predicted by the presence of specific genes encoding biosynthetic enzymes involved in their production, but the time-consuming physiological confirmation remains a critical element in the interpretation of such data in a biosynthetic and paleontological context. Alphaproteobacteria (APB) have often been proposed as important BHPD producers in a wide variety of environments and produce indicative BHPDs containing an additional methyl group in the A-ring, complicating the assignment of 2-methyl hopanes to N2-fixing cyanobacteria in paleontological studies. Here we provide the first comprehensive study of the relationship between genotype and phenotype with respect to the production of C30 hopanoids and BHPDs by APB. Genome analysis of > 6000 reference genomes of APB revealed that ca. 23% possess the genetic capacity to produce BHPDs, which is substantially higher than for all bacteria. However, BHPD biosynthesis genes were unevenly distributed between taxonomic and phylogenetic groups and not consistently found in mono-phylogenetic clusters. To study the relationship of genotype and phenotype with respect to the production of BHPDs, we cultivated 52 strains (50 terrestrial and 2 marine species) of the three major orders of the APB: Hyphomicrobiales, Rhodospirillales, and Sphingomonadales. These include species of 29 genera that have not previously been examined for BHPDs. Intact BHPDs were analyzed by UHPLC-MSn, resulting in the identification of overall 63 different structures and a wide variety in BHPD distributions. These results were in line with those obtained from Rohmer degradation on intact cells, which were specifically used to accurately assess the degree of methylation at C-2 and C-3 of ring A of the BHPDs. This revealed a 1-2 orders of magnitude lower degree of methylation at C-2 of BHPDs than for tetrahymanol (which was detected in three species all belonging to the Nitrobacteraceae) and C30 hopanoids, which has important implications for the interpretation of the molecular fossil record. Our results also showed that the presence of BHPD biosynthetic genes, often organized in a biosynthetic gene cluster, in all cases results in actual production of BHPDs. Thus, the presence of BHPD genes is a good predictor for the actual production of BHPDs. However, the presence of genes encoding proteins that result in methylation at C-2 and C-3 of BHPDs does not always lead to the production of methylated BHPDs, complicating the interpretation of the presence of the hpnP and hpnR genes in their genomes. Rohmer degradation-derived BHPD concentrations in APB species that do produce hopanoids can vary by two orders of magnitude and are not directly related to a specific phylogenetic group, indicating that the origin of sedimentary BHPDs may be biased towards specific species that produce relatively high amounts of BHPDs. These findings constrain the use of BHPDs as biomarkers for specific groups of bacteria in environmental and palaeontological studies.