Limnospira (Cyanobacteria) chemical fingerprint reveals local molecular adaptation

Abstract

Limnospira can colonize a wide variety of environments (e.g., freshwater, brackish, alkaline, or alkaline-saline water) and develop dominant and even permanent blooms that overshadow and limit the diversity of adjacent phototrophs, especially in alkaline and saline environments. Previous phylogenomic analysis of Limnospira allowed us to distinguish two major phylogenetic clades (I and II) but failed to clearly segregate strains according to their respective habitats in terms of salinity or biogeography. In the present work, we attempted to determine whether Limnospira displays metabolic signatures specific to its different habitats, particularly brackish and alkaline-saline ecosystems. The impact of accessory gene repertoires on respective chemical adaptations was also determined. In complement of our previous phylogenomic investigation of Limnospira (Roussel et al., 2023), we develop a specific analysis of the metabolomic diversity of 93 strains of Limnospira, grown under standardized lab culture conditions. Overall, this original work showed distinct chemical fingerprints that were correlated with the respective biogeographic origins of the strains. The molecules that most distinguished the different Limnospira geographic groups were sugars, lipids, peptides, photosynthetic pigments, and antioxidants. Interestingly, these molecular enrichments might represent consequent adaptations to conditions of salinity, light, and oxidative stress in their respective sampling environments. Although the genes specifically involved in the production of these components remain unknown, we hypothesized that within extreme environments, such as those colonized by Limnospira, a large set of flexible genes could support the production of peculiar metabolite sets providing remarkable adaptations to specific local environmental conditions.

Importance: Limnospira are ubiquitous cyanobacteria with remarkable adaptive strategies allowing them to colonize and dominate a wide range of alkaline-saline environments worldwide. Phylogenomic analysis of Limnospira revealed two distinct major phylogenetic clades but failed to clearly segregate strains according to their habitats in terms of salinity or biogeography. We hypothesized that the genes found within this variable portion of the genome of these clades could be involved in the adaptation of Limnospira to local environmental conditions. In the present paper, we attempted to determine whether Limnospira displayed metabolic signatures specific to its different habitats. We also sought to understand the impact of the accessory gene repertoire on respective chemical adaptations.

Keywords: adaptation; cyanobacteria; genomics; metabolomics; micro-diversity.

Fig 1

Heatmap analyses of pigment, lipid, and metabolite matrices. Heatmap analysis of the 40 pigment concentrations of the 93 analyzed L. platensis strains (normalization according to each feature) (A). Heatmap analysis of the normalized 14 fatty acid concentrations of the 91 analyzed L. platensis strains (B). Total fatty acid concentration (mg∙g⁻¹ dw) discriminates between two groups of lower (1) and higher (2) lipid contents (Welch test, P < 0.001). (C) Heatmap analysis of the 161 HPLC-MS/MS annotated ions of the 88 analyzed L. platensis strains, expressed in normalized relative abundances. The blue-red scale indicates the semi-quantification value of each pigment, fatty acid, and metabolite normalized according to Pareto’s formula.

Fig 2

PCA (A). Representation of the total matrix metabolite composition of the strains for the groups given by HCPC (B). Characterization of each group regarding the tested factors (C) and the multiple correspondence analysis (D).

Fig 3

MFA (A), BCA with a cos² threshold of 0.4 (B), and cos² plot (C), based on the chemical composition of the strains. The tested factor is the sampling site.

Fig 4

Most discriminated fatty acids and pigments (cos² > 0.4) were compared between sampling sites by Kruskal-Wallis test and pairwise Wilcoxon tests. Pigment concentrations are given in mg∙g⁻¹ dw for phycobiliproteins and chlorophylls a and µg∙g⁻¹ dw for other pigments (A). Fatty acid concentrations are given in mg∙g⁻¹ dw, and proportions in % of total fatty acids (B).

Fig 5

Relative abundances of most discriminated annotated ions (cos² > 0.4) were compared between sampling sites (Kruskal-Wallis and pairwise Wilcoxon tests).