Different lanthanide elements induce strong gene expression changes in a lanthanide-accumulating methylotroph

Abstract

Since its discovery, Ln-dependent metabolism in bacteria attracted a lot of attention due to its bio-metallurgical application potential regarding Ln recycling and circular economy. The physiological role of Ln is mostly studied dependent on presence and absence. Comparisons of how different (utilizable) Ln affect metabolism have rarely been done. We noticed unexpectedly pronounced changes in gene expression caused by different Ln supplementation. Our research suggests that strain RH AL1 distinguishes different Ln elements and that the effect of Ln reaches into many aspects of metabolism, for instance, chemotaxis, motility, and polyhydroxyalkanoate metabolism. Our findings regarding Ln accumulation suggest a distinction between individual Ln elements and provide insights relating to intracellular Ln homeostasis. Understanding comprehensively how microbes distinguish and handle different Ln elements is key for turning knowledge into application regarding Ln-centered biometallurgy.

Keywords: EDX; FFTEM; RNAseq; TEM; lanthanides; lanthanome; methylotrophy.

Fig 1

Overview of the cultivation setup and workflow for the carried out RNAseq experiments. Pre-cultures of Beijerinckiaceae bacterium RH AL1 were grown with pyruvate (0.2%, wt/vol, 18.175 mM) and washed with basal MM2 before being used as inoculum for two sets of incubations. Methanol (0.5%, vol/vol, 123 mM) was used as the carbon source for both sets, one investigating the effect of different (i) La concentrations (50 nM vs 1 µM), and one (ii), the effect of different Ln elements [La vs Nd vs Ln cocktail (Ce, Nd, Dy, Ho, Er, Yb)]. Cultivations were performed in triplicates (n = 3). Medium MM2 supplemented with methanol but without Ln source served as negative control (NC). Samples for DNA extraction and downstream quantitative PCR (qPCR) to determine cell numbers based on lanM gene copies were taken at the beginning and end of the incubation and during mid- to late-exponential phase. Biomass samples for RNA extraction were taken during mid- to late-exponential growth as well. Total RNA from each biological triplicate was enriched for mRNA by means of subtractive hybridization, before being subjected to library preparation, and Illumina sequencing. Pre-processed sequencing data were the starting point for differential gene expression analysis.

Fig 2

Methylotrophic growth of Beijerinckiaceae bacterium RH AL1 with methanol as the carbon source (0.5%, vol/vol, 123 mM) and either different concentrations of La or different Ln elements. (A) Growth curves of strain RH AL1 grown with either two different concentrations (50 nM and 1 µM) of La or different Ln elements [1 µM of either La or Nd or 0.9 µM of a Ln cocktail (Ce, Nd, Dy, Ho, Er, Yb)]. Cultures without added Ln served as negative control. Cultivations were performed in triplicates (n = 3). Growth was monitored spectrophotometrically (OD_600nm). The gray shade indicates the time interval considered for calculating doubling time (t_d) and growth rate (µ). Samples for molecular work were taken at time points t₀–t₂ including samples for RNA extraction (T₁ ). (B) Growth monitoring by lanM gene-targeting quantitative PCR. Samples for DNA extraction and qPCR were taken at the beginning (T₀ ), the mid- to late-exponential (T₁ ), and stationary phase (T₂ ).

Fig 3

Differential gene expression in response to La concentration and Ln elements. (A) Volcano plots indicate differentially expressed genes (DEGs) for the four different comparisons 1–4. Genes with changes in gene expression above |0.58| log₂FC, gene expression values higher than 4 (log₂CPM), and P-values smaller than 0.05 were considered for downstream analysis. The gray dashed lines indicate the log₂FC threshold of |0.58|. (B) The proportions of total DEGs, upregulated genes, and downregulated genes in relation to the number of genes encoded in the genome are shown as a bar chart. (C) UpSet plots were used to highlight sets of genes (>100 genes) that were either shared between different comparisons or unique. The lower half of (C) indicates potential overrepresentation of KEGG modules and KEGG pathways. The categories listed under KEGG modules include partially multiple KEGG modules: glycolysis/gluconeogenesis (M00001, M00003), pentose phosphate cycle (M00004), citrate cycle (M00009, M00011, M00173), serine pathway (M00346), and oxidative phosphorylation (M00144). FDR, false discovery rate.

Fig 4

Differential gene expression relating to selected metabolic aspects in response to different La concentrations and Ln elements. The responsive genes are categorized and listed in Table S11. For each comparison, the up- and downregulation of genes are illustrated through ridge plots based on log₂FC values. Marker sizes correspond to the log₂CPM values. Selected genes associated with motility, carbohydrate metabolism, PHA metabolism, serine cycle, sulfur metabolism, iron transport, and metal efflux were highlighted. PHA, polyhydroxyalkanoate; log₂CPM, log₂ counts per million.

Fig 5

Gene expression of lanthanome(-related) genes. We defined four groups of genes linked to the lanthanome: lanthanome core, PQQ biosynthesis, LC- and lut-clusters. The core lanthanome comprises genes for key proteins involved in lanthanide-dependent methanol oxidation. Lut- and LC-clusters refer to identified gene homologs of the previously described lut-cluster (lanthanide utilization and transport) and lanthanide chelator cluster (25, 31 ). The row dendrogram is based on euclidean distances. The colored sidebar indicates the affiliation of individual genes regarding the mentioned grouping. Gene expression is given in log₂CPM (counts per million), indicated by a blue-red color scale, and shown for the different biological replicates per condition (n = 3).

Fig 6

Electron microscopic examination of periplasmic Ln deposits. (A) Deposits were identified and localized by TEM (left panels). The middle panels are close-ups of the areas indicated by black arrows. Periplasmic deposits were also localized and identified by FFTEM (right panels). (B) The size of cells grown with La (30), Nd, or the Ln cocktail was compared by measuring cell areas. We also compared the area occupied by polyhydroxybutyrate (PHB) vacuoles and the area of the cell occupied by them. The analysis was done for three images (magnification, 4,000×; image area, 540.5 µm²), and between 27 and 30 cells were analyzed per condition (La, Nd, cocktail). Statistical significance was tested by means of one-way ANOVA, combined with a post-hoc Tukey-Kramer test. Asterisks indicate significant differences (P ≤ 0.05).

Fig 7

Elemental analysis of periplasmic Ln deposits from incubations with Nd and the Ln cocktail. TEM and FFTEM specimens were subjected to elemental analysis based on EDX. Red (deposit) and blue (reference) boxes indicate measured areas. ROI, region of interest; cps, counts per second. Scale bar, 250 nm.

Fig 8

Summary of metabolic aspects responding to differences in La concentration and added Ln elements in Beijerinckiaceae bacterium RH AL1 when grown with methanol as the carbon source. The color code of the text boxes refers to individual metabolic aspects. Colored proteins in the cell scheme highlight the differential expression of the genes encoding the corresponding protein. DGE, differential gene expression; GDH, glucose dehydrogenase.