# Antimicrobial Resistance in Bacterial Strains of Agricultural Interest: Predictions Based on Genomic Data

## TL;DR

This study examines antibiotic resistance in plant growth promoting bacteria and finds that genomic data can help predict resistance, though it doesn't always match experimental results.

## Contribution

The study combines genomic analysis with experimental testing to assess antibiotic resistance in agricultural bacteria.

## Key findings

- Gram-negative PGPB strains showed higher antibiotic resistance than Gram-positive strains.
- Genomic analysis confirmed the presence of resistance genes in resistant strains.
- There was no exact match between resistance gene presence and resistance levels, suggesting regulatory mechanisms are involved.

## Abstract

Background: Plant growth promoting bacteria (PGPB) are non-pathogenic bacteria that enhance plant growth through several mechanisms such as nutrient mobilization, phytohormones production, defense against phytopathogens, and alleviation of plant stress. Hence, these bacteria are used as ecologic biofertilizers to diminish the use of agrochemicals. Nevertheless, some PGPR strains can harbor antibiotic resistance determinants and the possibility of spreading them upon releasing these bacteria is an environmental concern. Objectives: The objectives of this work are as follows: (1) evaluating the antibiotic resistance in a collection of PGPB, and (2) prospecting antibiotic resistance genes in the genomes of PGPB in order to predict the risk for antibiotic resistance dissemination. Methods: The resistance towards 12 antibiotics in a collection of 20 PGPB (10 Gram-positive and 10 Gram-negative strains) has been evaluated using disk diffusion in agar, broth microdilution, and agar dilution tests. In addition, the whole genomes of six strains have been sequenced in order to find the correlation between the resistance levels and AMR genes by using bioinformatic tools. Results: The results indicated a wide range of halo diameters, but in general Gram-negatives showed higher resistance compared to Gram-positives. The four most resistant strains and the two more susceptible strains were selected for further analysis and sequencing the whole genomes. The resistant strains were identified as Achromobacter spanius N6, Leclercia adecarboxylata H17, Priestia aryabhattai strain MHA1, and Bacillus cereus N25. The susceptible strains were identified as Pantoea sp. S3 and Priestia megaterium MS4. Mining antibiotic resistance genes in the genomes confirmed the existence of resistance determinants responsible for the phenotypic behavior, indicating the potential of genomics for predicting antibiotic resistance in PGPB. However, there was not an exact correspondence between the presence of the genes and the level of resistance, suggesting the existence of additional regulatory mechanisms. Conclusions: The information obtained by genomics must be complemented experimentally by tests for antibiotic resistance determination. In this regard, it is necessary to develop a global antibiotic resistance database for PGPB, due to the difficulty of interpretation of the antibiotic susceptibility tests after comparing the experimental results with those tabulated for clinical species.