Prevalence and dynamics of antimicrobial resistance in pioneer and developing Arctic soils

Abstract

Antimicrobial resistance (AMR) in soil is an ancient phenomenon with widespread spatial presence in terrestrial ecosystems. However, the natural processes shaping the temporal dissemination of AMR in soils are not well understood. We aimed to determine whether, how, and why AMR varies with soil age in recently deglaciated pioneer and developing Arctic soils using a space-for-time approach. Specifically, we assess how the magnitude and spread of AMR changes with soil development stages, including antibiotic resistance genes (ARGs), mobile genetic elements (MGEs), and antibiotic-resistant bacteria (ARB). We showed that ARGs, MGEs, and ARB are present, and exhibit a non-uniform distribution in the developing soils. Their abundance generally increases with soil age but at different rates overall and across different glacier forefields. Our analyses suggest a strong positive relationship between soil age and ARGs and ARB, which we attribute to increased competition between microbes in older soils. We also observed a strong negative relationship between soil age and ARG diversity mediated by soil organic matter - suggesting facilitation due to the alleviation of nutrient limitation. These contrasting results suggest that both competition and facilitation can regulate AMR spread through time in the Arctic, but competition might be the stronger determinant of AMR spread.

Keywords: Antimicrobial resistance; Arctic soils; Metagenomics; Microorganisms; Soil development.

Fig. 1

Conceptual illustration of soil development processes following glacier retreat. The triangles indicate the magnitude of variables indicated in the accompanying text, with widening indicating increasing magnitude and thinning indicating decreasing magnitude. The diagram indicates various expectations that, with increasing distance from the glacier snout: age of soil increases, microbial abundance and diversity increases, nutrient availability increases, competition for nutrients decreases, and antagonistic interactions increases. The complex interplay of processes lead to alternate expectations with regard to the prevalence of ARGs and ARB in different stages of soil development

Fig. 2

Map of the sampling locations across in Svalbard, and photographs of sampling sites. The white/clear whirlpak bags and field notebooks pictured are approximately 15 cm in length

Fig. 3

Antimicrobial resistance in deglaciated Arctic soils. (A) Abundance of 16S rRNA gene, antibiotic resistance genes (ARGs), and mobile genetic elements (MGEs) in soils sampled from different glacier forefields and stages of development. (B) Relationship of ARGs with soil age (time since deglaciation). (C) Abundance of antibiotic resistant cultivable heterotrophic bacteria (ARB) in soils sampled from different glacier forefields and stages of development. (D) Relationship of ARB with soil age (time since deglaciation). Forefield 1 (blue line) corresponds to Austre Brøggerbreen (MHG1, MHG2, MHG3); Forefield 2 (green line) corresponds to Midtre Lovénbreen (MHG4, MHG5, MHG6); Overall (red line) corresponds to all samples evaluated together. The numbers in each cell denote the absolute abundance of different ARGs (log₁₀ copies g^− 1 soil) and ARBs (CFU g^− 1 soil). Grey cells in (A) indicate that ARGs and MGEs were not detected. The points in (B) and (D) represent qPCR replicates for each sample

Fig. 4

Microbial diversity in Arctic soils from different glacier forefields and stages of development. Microbial diversity is evaluated as (A) OTU richness, (B) Shannon diversity, and the relative abundance of dominant (C) microbial phyla and (D) Pseudomonadota. Numbers in each cell of A and B denote OTU richness and Shannon diversity, respectively, for each site

Fig. 5

Path analysis using partial least squares path model (PLS-PM) to evaluate inter-relationships among bacterial diversity, mobile genetic element, antibiotic resistance genes, antibiotic-resistant bacterial abundance, and soil edaphic factors in developing Arctic soils from deglaciated forefields. The numbers next to the arrows indicate the path coefficients representing the relationships. Path coefficients were calculated after 1,000 bootstraps. The values in the box next to each indicator in the latent variable represent the loadings of the measured variable or indicator. To reduce clutter, only significant paths (α = 0.05) are plotted (for the full model, see companion in Fig. S6 and Table S8). The model is assessed using the Goodness of Fit (GoF) statistic, which is 0.692. SOM: soil organic matter; WHC: water holding capacity; BD: bulk density. Blue arrows represent significant positive relationships, while the red arrows represent significant negative relationships. Black arrows represent the relationship of the latent variable with its block of the indicator. R² values indicate the variance explained by the model for each variable. Asterisks next to each path coefficient represent statistical significance (***P ≤ 0.001, **P ≤ 0.01 and *P ≤ 0.05)

Fig. 6

Antibiotic resistome following metagenomics from developing Arctic soils from deglaciated forefields. (A) The number of antibiotic resistance genes (ARGs) detected and (B) their relationship with soil age (time since deglaciation). (C) ARGs identified across categories based on different mechanisms of resistance, (D) different drug classes, and (E) gene families. The number in each cell denotes the number of ARGs for each soil across different categories. Grey cells indicate the absence of detected ARGs. The points in (B) represent individual samples. rbp: RNA-polymerase binding proteins, RND: Resistance-nodulation-cell division, SMR: Small multidrug resistance, RPP: Ribosomal protection proteins, DHFR: Di-hydro folate reductase