Erwinia amylovora is the causal agent of fire blight of apple and pear trees. Several bacteria have been shown to produce antibiotics that antagonize E. amylovora, including pantocins, herbicolins, dapdiamides, and the vinylglycines, 4-formylaminooxyvinylglycine (FVG) and 4-aminoethoxyvinylglycine (AVG). Pantoea ananatis BRT175 was previously shown to exhibit antibiotic activity against E. amylovora via the production of Pantoea natural product 1 (PNP-1), later shown to be FVG; however, exposure of E. amylovora to FVG results in spontaneously resistant mutants. To identify the mechanism of resistance, we used genome variant analysis on spontaneous FVG-resistant mutants of E. amylovora and identified null mutations in the l-asparagine permease gene ansP. Heterologous expression of ansP in normally resistant Escherichia coli was sufficient to impart FVG susceptibility, suggesting that FVG is imported through this permease. Because FVG and AVG are structurally similar, we hypothesized that resistance to AVG would also be conferred through inactivation of ansP; however, ansP mutants were not resistant to AVG. We found that spontaneously resistant Ea321 mutants also arise in the presence of AVG, with whole-genome variant analysis revealing that resistance was due to inactivation of the arginine ABC transporter permease subunit gene artQ. Heterologous expression of the predicted lysE-like transporter encoded within the Pantoea ananatis BRT175 FVG biosynthetic cluster, which is likely responsible for antibiotic export, was sufficient to confer resistance to both FVG and AVG. This work highlights the important roles of amino acid transporters in antibiotic import into bacteria and the potential utility of antimicrobial amino acid analogs as antibiotics. IMPORTANCE The related antibiotics formylaminooxyvinylglycine (FVG) and aminoethoxyvinylglycine (AVG) have been shown to have activity against the fire blight pathogen Erwinia amylovora; however, E. amylovora can develop spontaneous resistance to these antibiotics. By comparing the genomes of mutants to those of the wild type, we found that inactivation of the l-asparagine transporter conferred resistance to FVG, while inactivation of the l-arginine transporter conferred resistance to AVG. We also show that the transporter encoded by the FVG biosynthetic cluster can confer resistance to both FVG and AVG. Our work indicates the important role that amino acid transporters play in the import of antibiotics and highlights the possible utility in designer antibiotics that enter the bacterial cell through amino acid transporters.

Pollution with nitrogen (N) and phosphorous (P) impairs streams by favoring suspended algae and cyanobacteria over diatom-rich periphyton. Recently, wastewater treatment plants have been upgraded to biological nutrient removal to eliminate both P and N (mainly NH 4 +), although little is known of the effects of this effluent on flowing waters. Here, we used high performance liquid chromatography to quantify how the abundance and composition of phytoplankton and periphyton varied in response to both influx of effluent produced by biological nutrient removal and physico-chemical conditions in small, turbid, P-rich streams of the northern Great Plains. At the catchment scale, analysis with generalized additive models (GAMs) explained 40.5-62.6% of deviance in total phototroph abundance (as Chl a) and 72.5-82.5% of deviance in community composition (as biomarker carotenoids) in both planktonic and benthic habitats when date-and site-specific physico-chemical parameters were used as predictors. In contrast, GAMs using wastewater input (as aqueous δ 15 N) as a predictor explained up to 50% of deviance in Chl a, and~60% of deviance in community composition, in both suspended (51.6% of Chl a, 67.1% of composition) and attached communities (21.5% of Chl a, 58.8% of composition). Phytoplankton was replaced by periphyton within a 60-km wastewater-impacted reach due to dilution of streams by transparent effluent and addition of urban NO 3 À , although predominance of phytoplankton was re-established after confluence with higher-order streams. Overall, influx of effluent shifted turbid, phytoplankton-rich streams to clear ecosystems with abundant epilithon by improving water transparency and providing NO 3 À to favor benthic diatoms and chlorophytes.

Urea can degrade water quality and stimulate toxic phytoplankton in P-rich lakes, yet little is known of its sources, abundance, or transportation in lotic systems, particularly within the Northern Great Plains. We measured physico-chemical parameters biweekly during May–September 2010–2012 at 16 stations along a 250 km lotic continuum to quantify spatial and temporal variation in urea concentrations and discharge, and to identify potential regulatory processes. Urea concentrations were similar to those in regional prairie lakes (range 5.2–792.1, median 78.6 μg N L−1) with variable seasonal mean (± SD) concentrations (96.6  ± 96.1 μg N L−1) and fluxes (4.22 × 105 ± 257.6 μg N s−1). Landscape analysis with generalized additive models explained 68.3% of deviance in urea concentrations, with high temporal variability predicted mainly by positive relationships with nutrient content and chlorophyte abundance, but not temperature, dissolved organic matter, bacterial abundance, or urban effluent. Seasonal analysis revealed that during spring, urea content was correlated negatively with leguminous forage cover (% area) and positively with stream discharge, oilseed and cereal crops, and shrubs or deciduous plants, while during summer, urea concentrations were correlated negatively with discharge and leguminous crop cover, as well as nutrient levels. Mean porewater urea concentrations (528.5 ± 229.8 μg N L−1) were over five-fold greater than stream concentrations, suggesting that hyporheic production may offset declining influx from terrestrial sources during summer. We conclude that urea may be ubiquitous in eutrophic prairie streams and that management of its export from land may reduce detrimental effects on downstream lakes.

Biological nutrient removal (BNR) may be an effective strategy to reduce eutrophication; however, concerns remain about effects on receiving waters of removing both nitrogen (N) and phosphorus (P), rather than P alone. Phytoplankton abundance (as µg chlorophyll a/L) and community composition (as nmol biomarker pigment/L) were quantified over 6 years in two connected eutrophic streams to determine how algae and cyanobacteria varied in response to a shift from tertiary (P removal) to BNR (N and P removal) wastewater treatment. Phytoplankton were sampled biweekly at nine stations May to September and were analysed using generalised additive models (GAMs) to quantify landscape patterns of phototrophs and identify potential causal relationships both before (2010–2012) and after (2017–2019) BNR installation in 2016. Analysis with GAMs showed that 69%–79% of deviance in phytoplankton abundance and composition could be explained by date‐ and site‐specific variance in stream flow, temperature, and solute concentrations (mainly nutrients), whereas similar GAMs using only effluent N content (δ15Nwater) as a predictor explained c. 60% of phototroph deviance. Prior to BNR, phytoplankton levels (mainly chlorophytes) increased with ‐rich effluent, whereas their abundance declined with δ15N after BNR (diatoms, chlorophytes). Overall, declines in total effluent release of N (67%–97%) but not P (c. 0%) due to BNR resulted in a 52 ± 7% decline in phytoplankton abundance relative to upstream values, despite high inter‐annual variation in discharge and baseline chlorophyll a concentration. Nitrogen removal by BNR improved water quality in N‐limited ecosystems.