Coupled metalipidomics-metagenomics reveal structurally diverse sphingolipids produced by a wide variety of marine bacteria

Abstract

Microbial lipids, used as taxonomic markers and physiological indicators, have mainly been studied through cultivation. However, this approach is limited due to the scarcity of cultures of environmental microbes, thereby restricting insights into the diversity of lipids and their ecological roles. Addressing this limitation, here we apply metalipidomics combined with metagenomics in the Black Sea, classifying and tentatively identifying 1623 lipid-like species across 18 lipid classes. We discovered over 200 novel, abundant, and structurally diverse sphingolipids in euxinic waters, including unique 1-deoxysphingolipids with long-chain fatty acids and sulfur-containing groups. Sphingolipids were thought to be rare in bacteria and their molecular and ecological functions in bacterial membranes remain elusive. However, genomic analysis focused on sphingolipid biosynthesis genes revealed that members of 38 bacterial phyla in the Black Sea can synthesize sphingolipids, representing a fourfold increase from previously known capabilities and accounting for up to 25% of the microbial community. These sphingolipids appear to be involved in oxidative stress response, cell wall remodeling and are associated with the metabolism of nitrogen-containing molecules. Our findings underscore the effectiveness of multi-omics approaches in exploring microbial chemical ecology.

Repository content:

1) metalipidome_sphingolipids.zip: includes source data and code scripts used for figures regarding metalipidome and sphingolipids abundance, classification and diversity in this study. Files are organized as follows and are associated with the corresponding parts of the manuscript: Fig. 1a, Fig. 1b,  Fig. 2b,  Fig. 2c, Fig. 2e,  Fig. 2f, Fig. 2g, Fig. 2h, Fig. 4d, Supplementary Fig. 2, Supplementary Fig. 3.

2) Source data_major lipid classification.xlsx: includes original tables regarding metalipidome identification, abundance, precursor mass, retention time, classification as well as ID (name) in the molecular network.

3) Source data_sphingolipids information.xlsx: includes information about sphingolipids identification, precusor mass, retention time, peak intensity, elemental composition and etc.

4) Black_Sea_2013.code.tar.gz: contains the directory structure and code used for the metagenomics part of this project. Each directory contains a 'commands.sh', which contains the code to generate the content in that directory. Other shell and python scripts are always run from within 'commands.sh', with the exception of the files within the 'figures' directory which contains Jupyter labs and a python script that were run individually.

5) MAGs.tar.gz: all the MAGs generated by DAS Tool including CheckM and GTDB-Tk analyses (inside the 'binners' directory). Final taxonomic annotations of MAGs (see 'MAG2info.txt' file) are based on BAT annotatations with GTDB as a reference database, source data in the directory 'CAT_and_BAT_with_GTDB_refdb'.

6) abundance_profile.tar.gz: taxonomic annotation (based on CAT and BAT) and abundance of all scaffolds in the file 'big_table.txt'. Columns that start with 'mappings' are the read counts mapping to the scaffold in the sample from which it is assembled. Since the scaffolds were assembled per sample, only one of the 15 samples contains read mappings per scaffold. The last 15 columns (that start with 'BlackSea') are the depth per 1e8 mapped reads based on the all versus all mappings and were used for co-abundance analyses with sphingolipids. The file 'taxon2counts.txt' summarizes the taxonomic composition of the water column based on summing of the read mappings of the samples from which scaffolds were assembled (the 'mappings' columns in 'big_table.txt'), i.e. they represent all reads that could be associated with a certain taxon in that sample. The 'taxon2counts.txt' file used in Fig. 3c,d.