Very few phylogenetic clades of Bacteria claim traits of their own that others don’t share. The phylum Bacteroidetes stands out from other bacteria with a rare characteristic: the ability to produce sphingolipids. Although eukaryotes make this special class of lipids with roles in cell membrane structure and signaling, the Bacteria and Archaea generally do not. The exceptions: Bacteroidetes and some Proteobacteria (e.g., Sphingomonas). Intriguingly, these bacterial sphingolipid-producers are also often host-associated. In the human gut, the Bacteroidetes makes up a large fraction of the microbiome. We asked: could the sphingolipids produced by gut bacteria affect the host’s own sphingolipid levels? Do we carry a source of endogenous sphingolipid with us?

To get at this question we enlisted the help of a sphingolipid specialist, Dr. Tilla Worgall of Colombia University. She enthusiastically trained the lead author, Dr. Elizabeth Johnson (now Professor Johnson), to measure a panel of sphingolipids on the mass spec (with a lot of back and forth on the bus between Ithaca and NYC). Then a wonderfully brave and talented new graduate student, (now Dr.) Stacey Heaver, jumped on board. Together they launched into new territory for the lab: lipid analysis, cell culture, Bacteroides genetics, click chemistry (Figure 1), inoculation of germfree animals with a wildtype strain of Bacteroides thetaiotaomicron and a mutant deficient in sphingolipid production. All while the lab was in the process of moving from Cornell University to the Max Planck Institute for Developmental Biology in Tübingen, Germany.

Figure 1: Bacteroides thetaiotaomicron lipids are labelled with click chemistry. Left: Cell pellet after growth in medium supplemented with click-taggable palmitic acid (PA). Right: the cells are visualized by fluorescent microscopy: PA is tagged with green fluoresence.

We reached out to great colleagues for additional help: Dr. Andy Goodman (Yale) provided the initial strain of B. theta deficient in sphingolipid synthesis to use until Stacey knocked the gene (SPT) out. Dr. Andrew Gewirtz and Alexis Bretin (GSU) performed the experiment with high-fat diet fed mice administered B. theta: a great help given that our lab was between countries. 

The final product: together we showed that sphingolipids produced in the gut do indeed affect the host’s liver ceramide levels and de-novo production rate. Deletion of a single gene (SPT) in the bacterial genome can impact the host lipodome. We previously linked this effect to altered insulin sensitivity and noted a modest effect of B. theta supplementation (figures 5c,d here). We look forward to better understanding how this endogenous source of sphingolipid affects host development and metabolism.