Selective lipid recruitment by an archaeal DPANN symbiont from its host

Abstract

The symbiont Ca. Nanohaloarchaeum antarcticus is obligately dependent on its host Halorubrum lacusprofundi for lipids and other metabolites due to its lack of certain biosynthetic genes. However, it remains unclear which specific lipids or metabolites are acquired from its host, and how the host responds to infection. Here, we explored the lipidome dynamics of the Ca. Nha. antarcticus – Hrr. lacusprofundi symbiotic relationship during co-cultivation. By using a comprehensive untargeted lipidomic methodology, our study reveals that Ca. Nha. antarcticus selectively recruits 110 lipid species from its host, i.e. nearly two-thirds of the total number of host lipids. Lipid profiles of co-cultures displayed shifts in abundances of bacterioruberins and menaquinones and changes in the degree of bilayer-forming glycerolipid unsaturation. This likely results in increased membrane fluidity and improved resistance to membrane disruptions, consistent with compensation for higher metabolic load and mechanical stress on host membranes when in contact with Ca. Nha. antarcticus cells. Notably, our findings differ from previous observations of other DPANN symbiont-host systems, where no differences in lipidome composition were reported. Altogether, our work emphasizes the strength of employing untargeted lipidomics approaches to provide details into the dynamics underlying a DPANN symbiont-host system.

Repository Contents

1_DPANN_lipidome.zip: includes all source data and code scripts used for figures in this study. Files are organized as follows and are associated with the corresponding parts of the manuscript: Figure 1b, Figure 1c, Figure 1d, Figure 1e, Figure 1f, Figure 1g, Figure 3a, Figure 3b, Supplementary Figures 7-8.

Figure 1. Overview of the experimental design and the general lipidome composition in the Hrr. lacusprofundi-Ca. Nha. antarcticus system. (b) qPCR based growth measurements of pure Hrr. lacusprofundi cultures and co-cultures of Hrr. lacusprofundi with Ca. Nha. antarcticus. Error bars show the standard deviation of calculated 16S rRNA copy number. (c) Optical density at 600 nm (OD600) growth measurements of pure Hrr. lacusprofundi cultures and co-cultures of Hrr. lacusprofundi with Ca. Nha. antarcticus. Error bars show the standard deviation of measured OD600 values. (d) The number of individual lipid species in major lipid classes among all the samples. (e) Principal Component Analysis (PCA) based on the abundance of intact polar lipid species, showcasing the variance in general lipidomic features among distinct cultures or over varying culture durations. (f) Information theory analysis showing lipidome diversity and specialization based on the Shannon entropy of the lipidomic frequency distribution. Error bars in the data represent variability across replicates. (g) Hierarchical clustering heatmap depicting the distribution of major lipid classes among distinct cultures or over varying culture durations. The colour bar on the right side represents Z-score normalization scale (ranges from -3 to +3 standard deviation). Sample abbreviations: Ca. Nha. antarcticus (Nha), Hrr. lacusprofundi (HP), co-cultures (Cc). Lipid abbreviations: archaeol core lipids (AR), phosphatidylglycerol (PG), phosphatidylglycerosulfate (PGS), phosphatidic acid (PA), phosphatidylglycerophosphate methyl ester (PGP-Me), biphosphatidylglycerol (PGPG), cardiolipin (CL), sulphated diglycosyl (SDG), monoglycosyl (1G), diglycosyl (2G), archeaol lipids containing a sulfur-containing head group except for PGS (S), menaquinone (MK), an "extended " archaeological chain", i.e. with a C25 isoprenoid carbon chain (EXT-AR), unsaturation in the archaeol chain (uns). The two "n" in MK (n:n) stand for numbers of the isoprenoid unit in the side chain and unsaturation in the isoprenoid chain, respectively. MK(n:n-1) signifies one less double bond in the nth isoprenoid chain.

Figure 3. The presence, absence, and changes in lipid composition in the Hrr. lacusprofundi-Ca. Nha. antarcticus system (a) The relative abundance of representative lipid species within the most dominant lipid classes. Statistical differences in lipid species among the samples were assessed using the Tukey's Honest Significance Difference test (TukeyHSD), and results were visualized with the Compact Letter Display (CLD) (P < 0.05). (b) The intersection of lipid species across samples is illustrated through an UpSet plot. A threshold of 0.01% relative abundance of total lipids was applied to determine the presence of a lipid in a specific sample; lipids with less than 0.01% of total lipid abundance were considered absent in that sample. The dark connected dots denote lipid species shared among these samples. Abbreviations: demethylmenaquinone (DMK), methylmenaquinone (MMK), dimethylmenaquinone (DMMK). The representative lipid species are 1G-AR (m/z 832.760, C49H102O8N+), 2G-AR (m/z 994.813, C55H112O13N+), AR (m/z 653.681, C43H89O3+), Bacterioruberin (m/z 741.581, C50H77O4+), CL-AR-AR (m/z 1522.313, C89H183O13P2+), MK(8:8) (m/z 717.560, C51H73O2+), PG-AR (m/z 807.684, C46H96O8P+), PGP-Me-AR (m/z 901.666, C47H99O11P2+), PG-PG-AR (m/z 961.687, C49H103O13P2+).

Supplementary Fig. 7 The distribution of different unsaturation degrees of the summed bilayer-forming glycerolipids and menaquinones among the samples. Statistical differences in lipid species among the samples were assessed using the Tukey's Honest Significance Difference test (TukeyHSD), and results were visualized with the Compact Letter Display (CLD) (P < 0.05).

Supplementary Fig. 8 The intersection of lipid species across samples including the enrichment is illustrated through an UpSet plot. A threshold of 0.01% relative abundance of total lipids was applied to determine the presence of a lipid in a specific sample; lipids with less than 0.01% of total lipid abundance were considered absent in that sample. The dark connected dots denote lipid species shared among these samples.

2_Lipidome source data.xlsx: includes an original table regarding lipidome identification, abundance, precursor mass, retention time, classification as well as ID (name) in the molecular network.

3_Microscopy_Genomics_Scripts.zip: Includes all scripts used for analyses of microscopy data and assembly and annotation of genomes. Scripts are:

Image_Analysis.ijm - ImageJ macro for automated analyses of 16S rRNA FISH images. Analyses were run with a default install of Fiji.

Microscopy_Analyses.R - R script for production of Supplementary Fig. 5. Colours and letter coding were added to plots manually after export from the R environment.

Assembly_Annotation.rmd - Markdown file with workflow used for genome assembly and annotation described in Figure 4, and Supplementary Tables 2 - 5. Assembly and annotations were processed on an in house HPC consisting of 4x Xeon Gold 6140 2.3 GHz processors using bash, python and perl. The system runs a Linux operating system, Red Hat Enterprise 7.5.