Invasion dynamics of antimicrobial-resistant E. coli in river biofilms: impacts on the resistome, microbiomes, and horizontal gene transfer

Invasion dynamics of antimicrobial-resistant E. coli in river biofilms: impacts on the resistome, microbiomes, and horizontal gene transfer | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Invasion dynamics of antimicrobial-resistant E. coli in river biofilms: impacts on the resistome, microbiomes, and horizontal gene transfer Giulia Gionchetta, Jangwoo Lee, Oliver Hansen, Karin Beck, Helmut Bürgmann This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8270852/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract River biofilms are exposed to invasion by antibiotic resistant bacteria (ARB) due to episodic or chronic exposure to wastewater, yet the ecological processes determining the fate of invaders and their resistance plasmids remain poorly understood. We experimentally exposed river-grown biofilms, originating from sites with contrasting microbial diversity and wastewater influence, to invasion by a genetically tagged ARB- E. coli carrying a transferable IncPα plasmid with the nptII resistance gene. Over two weeks, we quantified the dynamics of the invader and its plasmid, using qPCR and the plasmid-to-strain genome ratio in the biofilm as an indicator of horizontal gene transfer (HGT). We further characterized microbiomes and resistomes via 16S rRNA gene sequencing and metagenomics. Independent quantification methods provided highly consistent estimates of invasion dynamics: the invader established transiently across all biofilms, with ARB- E. coli abundance peaking within 48h and subsequently declining to near-background levels within 14 days. Plasmid-to-strain genome ratios decreased, indicating limited HGT and progressive plasmid loss. Wastewater-impacted biofilms showed slower declines, suggesting higher plasmid persistence potential in disturbed environments. The total community resistome exhibited pronounced but short-lived shifts whereas indigenous resistomes and microbiome composition remained stable. Yet, in one replicate from the wastewater-impacted site, specific indigenous ARGs of public health relevance increased, indicating that disturbance can promote localized ARG proliferation even without sustained invader establishment. Our results show that interactions between invaders and indigenous biofilm are dynamic and strongly shaped by community composition. This supports the One Health concept, highlighting how environmental context modulates AMR propagation risks in freshwater ecosystems. Biological sciences/Ecology Earth and environmental sciences/Ecology Earth and environmental sciences/Environmental sciences Biological sciences/Microbiology Antibiotic resistance antibiotic resistance genes disturbance mobile genetic elements diversity pollution Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction The role of the environment in the spread and evolution of antimicrobial resistance (AMR) is one focus of the One Health approach. From this perspective, the environment is seen as a large reservoir of active or latent antibiotic resistance genes (ARGs), which can transfer to actual or potential pathogens, especially under selective pressures. Given that the spread of resistant bacteria and their antibiotic resistance determinants among environmental microbiomes can possibly influence downstream transmission risks and evolutionary trajectories by altering bacterial population genetics [ 1 ], it is important to understand the dynamics and environmental controls of these processes. Horizontal gene transfer (HGT) is one of the most significant pathways of resistance evolution [ 2 ]. Specifically, HGT between bacteria can lead to the spread of antibiotic resistant traits into previously susceptible pathogens or into abundant environmental species [ 3 ]. The conjugative transfer of plasmids carrying one or multiple ARGs is a particularly important mechanism of HGT and resistance dissemination, because it can lead to the transfer of multiple resistances in a single transfer event. Microbiomes of river biofilms exposed to wastewater (WW) discharge could be especially prone to the conjugative transfer of antibiotic resistant traits due to the chronic pollution of WW with antibiotic residues and AMR vectors (i.e. ARB, resistance plasmids and genes) [ 4 ], with the extent of impact varying according to physio-chemical factors, such as temperature and turbidity [ 5 ] as well as the impact of environmental stressors or selective agents (i.e. heavy metals, biocides or antibiotic residues) often co-released in WW effluents [ 6 ]. Furthermore, the composition and abundance of environmental ARGs (i.e., the resistome of a microbiome) is frequently observed to be related to changes in microbial community composition, that may, but must not necessarily, be related to specific AMR-selective pressures [ 7 – 9 ]. For instance, recent studies revealed that in response to disturbance, resistomes were structurally correlated with microbiomes in water compartments, such as freshwater [ 8 ] or WW [ 7 ]. However, there is presently a lack of such analyses for river biofilms. There have been a few studies where certain groups of microbial taxa such as pathogenic or WW-related groups were correlated with selected ARGs [ 9 ]. However, the magnitude and importance (i.e., relative to HGT) of this effect remains unexplored, highlighting the need for a study that simultaneously assesses both factors. Studies exploring resistome rsponse to invasive microbial species are likewise currently not available. However, several studies explored the effects of invasion on the invaded microbial communities [ 10 , 11 ]. Many of these studies found a significant impact, for example transient or permanent increase in overall taxonomic diversity, with varied results depending upon the biotic or abiotic conditions experienced by the indigenous communities [ 12 ]. Given that changes in resistomes can be attributed to presence of invading species, HGT, changes in microbiomes, or all of these, this question necessitates further exploration. In an earlier study, we observed that invasions of river-grown biofilms with E. coli MG1655 (strain CM2372, derived from E. coli EDCM367 [ 13 ], ARB- E. coli ) were transient, regardless of temperature and flow conditions [ 14 ]. However, in the previous study we did not monitor the persistence of the conjugative plasmid of this organism, which could potentially transfer to native bacteria persisting in the biofilm. Likewise, the ability of invasion and HGT to alter river biofilm microbiomes and thus indirectly their resistomes has not been assessed. Here we performed experiments with a non-pathogenic ARB- E. coli , possessing a plasmid ( pG527) harbouring the nptII resistance gene invading indigenous biofilms obtained from four locations of the river Glatt, Switzerland. Biofilms from these locations differed in microbial community composition and diversity and their pre-exposure to WW. The goals of this study were to (1) quantify the survival of the ARB- E. coli in the different biofilms and, (2) assess persistence of its resistance-conferring plasmid pG527 to evaluate HGT. Further to (3) evaluate whether a transient invasion would drive changes in the overall community structure of the biofilm microbiomes and their resistomes. We hypothesized that previous exposure to WW and indigenous biofilm diversity are influencing factors. Specifically, we hypothesized that (1) the most upstream (not previously exposed to evident source of pollution) and least diverse biofilms would be the most vulnerable to the invasion and to (2) HGT by conjugative plasmid transfer. We further expected that (3) the ARB- E. coli invasion would induce correlated changes of resistomes and microbiomes over time, indicating resistomes are linked to shifts in community structure. The fate and the invasion dynamics of ARB- E. coli and its plasmid, as well as the biofilms total and native ( E. coli effect eliminated) resistomes and microbiomes were monitored through qPCR, shotgun metagenomics and amplicon sequencing, tracking temporal changes in resistomes and microbiomes, and their potential association, employing multivariate analysis. 2. Material and methods 2.1 Study locations and biofilm colonization The study sites are located along the river Glatt, a 25 km-long tributary of the Thur in northeastern Switzerland (Fig. S1 ). The two “upstream” sites ( upL and upH ) are located about three kilometres upstream of the Oberglatt WWTP, one in the Glatt and one in a tributary - both sites and their mostly forested watersheds experience minimal human impact. The two “downstream” sites are located a few hundred meters up- (“before”, dwB ) and downstream (“after”, dwA ) of the WWTP discharge point. Both sites are downstream of a tributary (Dorfbach) draining the town of Gossau, and agricultural fields, indicating increased human impact. Sampling locations were chosen based on preliminary microbiome diversity screenings of native epilithic biofilms [ 15 ]: upL and upH had low and high native biofilm microbial diversity, respectively, while dwB and dwA had high and highest biofilm biodiversity. In May 2022, the river biofilms for the experiments were grown in situ for 5 weeks on glass slides in environmental exposure units (SI1 and Fig. S2 ; [ 14 ]). 2.2 Flumes experiment and biofilm collection Twelve flumes were set up in a 20°C climate chamber, with three flumes per sampling site: one control and two receiving the E. coli invader (Fig. S3A). Each flume contained 6.5 L of filtered river water in a continuous recirculation system (8 L total water volume, details in SI2). To ensure sufficient biomass recovery and to minimize the impact of slide-to-slide variability, every biomass sample consisted of pooled biofilm of three randomly chosen microscope glass slides, gently scraped with a spatula and washed with sterile water from both sides (Fig. S3B). Samples were taken at time 0 and 1, 2, 4, 7, 10, and 14 days post-invasion. Biofilms were centrifuged, supernatant removed, weighed (wet biomass, mg), and 0.25 g stored at -20°C for DNA extraction. 2.3 Inoculum of antibiotic resistant E. coli MG1655 preparation A non-pathogenic E.coli MG1655 ΔLacZY strain possessing a native conjugative IncPα plasmid carrying the nptII resistance gene, was grown overnight at 37°C on a shaker table operating at 160 rpm in three falcon tubes each with 10 mL of Luria-Bertani (LB) broth containing 50 µg/mL of kanamycin ([ 13 , 16 ], SI3). To start the invasion event (time 0), MG1655 cell suspension was mixed into the 10% water volume (800 mL) added during the weekly water renewal, in each flume apart from those of controls, to achieve a final concentration of 10 7 cells/mL in the recirculation system. 2.4 DNA extraction and quantitative PCR (qPCR) DNA extraction was conducted using the DNeasy PowerMax Soil Kit (Qiagen, Germany) following the manufacturer’s instructions. Extraction blanks confirmed the absence of DNA contamination. The quality and quantity of the extracted DNA were assessed with a NanoDrop One spectrophotometer (Thermo Fisher Scientific, USA). DNA was stored at -20°C for future analysis. E. coli MG1655 and its conjugative IncPα plasmid were quantified via qPCR for samples collected on days 0, 1, 2, 4, 7, 10, and 14 (Fig. S3B). The plasmid pG527, containing the target sequences, served as the qPCR standard for E. coli MG1655 ΔlacZY strain and its IncPα plasmid [ 13 , 16 ]. It was linearized with XbaI (Thermo Fisher Scientific, USA), purified with the QIAquick PCR purification kit (Qiagen, Germany), and verified by gel electrophoresis. qPCR assays were performed using Taqman chemistry on a LightCycler 480 II (Roche, Switzerland) in a 10 µL reaction volume. Details on primers, probes, and conditions are provided in Table S1 , with further information on the qPCR procedure and standard curve efficiency in Table S2 [ 8 , 17 , 18 ]. E. coli invader MG1655 was quantified using qPCR targeting the ΔlacZY chromosomal gene, while pG527 was used for the plasmid [ 16 ]. Bacterial abundance was assessed through qPCR of the 16S rRNA gene using universal primers [ 19 ]. The qPCR results were expressed in units of gene copies per glass slide surface area (cm 2 ), and relative abundances of ΔlacZY and pG527 were obtained by normalizing to 16S rRNA gene copy numbers (gene copies/16S). 2.5 Shotgun metagenomics and amplicon sequencing workflows Shotgun metagenomics and 16S rRNA gene amplicon sequencing analysis were performed using Illumina platforms, NovaSeq6000 with a paired-end (2×150bp) strategy and MiSeq (2×300bp), respectively. For shotgun metagenomics, DNA samples from each invasion treatment replicate and one control replicate corresponding to sampling times day 0, 1, 2, 4, 10, 14 (7 was omitted) were sequenced for each sampling site (n = 72). Library construction and sequencing was performed by IKBM facility (Kiel University, Germany), yielding a median output of 21.9 giga-bases (interquartile range of 20.8–23.6). Shotgun metagenomic reads were cleaned by IKBM by removing adapter sequences and low-quality reads (N > 10% and quality score ≤ 5). The read qualities were double checked by the authors using FastQC v0.11.4 [ 20 ]. For ‘total’ (i.e., including the indigenous community resistome and that of the E. coli invader) resistome analysis, DeepARG short-sequence pipeline v1.0.2 was used to annotate ARGs using the consolidated DeepARG DB v2.0 database [ 21 ]. This pipeline finds and quantifies ARG-like reads from short read data and normalizes to 16S rRNA gene abundance [ 21 ]. For all downstream analysis, we used this normalization (i.e., ARG copies per 16S rRNA gene) and refer to it as relative abundance [ 21 , 22 ]. We further evaluated ARG-subtypes at more granular level (e.g., from OXA to OXA-1, 2 …) using a customized workflow written in Python v3.5.3, which was developed previously [ 23 ]. To separate ARGs associated with the E. coli invader from those associated with the invaded community, we followed a bioinformatics workflow modified from relevant references [ 24 , 25 ]. We first mapped the trimmed reads against reference sequences for E. coli invader chromosome (Strain MG1655∆lacZY) and for the plasmid (GenBank: JX469830.1) using Bowtie2 v2.2.8, with ‘--very-sensitive’ option. Using Samtools v2.6.3., the unmapped reads were separated from the mapped reads, then subjected to further analysis using DeepARG to identify ‘indigenous’ resistomes, as outlined above (Fig. 1 ). Using the mapped reads, we further calculated coverage for each of chromosome and plasmid, followed by normalization using 16S rRNA gene, as outlined above for total resistome analysis (Fig. 1 ). For the comparison between biofilms grown on natural and artificial substrates, 16S rRNA gene amplicon sequencing was performed on duplicates of DNA from rocks and glass slides biofilms for each site (n = 16). To monitor changes in the microbiome, duplicates of DNA samples from both treatment flumes obtained on day 0, 1, 2, 4, 7, 10, and 14 of the experiment were sequenced (n = 56). Amplicon sequencing of bacterial 16S rRNA genes was performed using primers targeting the V3-V4 region (V3F: 5′-CCTACGGGAGGCAGCAG-3′; V4R: 5′-GGACTACHVGGGTWTCTAAT-3) [ 26 ] to generate paired-end reads, that were collectively analysed using the DADA2 algorithm [ 27 ] in QIIME2 [ 28 ]. On average, 9288 reads per sample were obtained, distributed across 56 samples with a minimum of 1553 and a maximum of 21127 reads per sample. Forward and reverse reads were merged into amplicon-sequence variants (ASVs), classified taxonomically, and used to build an ASV table (4,090 ASVs). Indigenous microbiomes were analysed after screening out reads not mapped against 16S rRNA gene sequences for E. coli invader, following the same protocol for indigenous resistome. This resulted in 9230 reads per sample on average and 4,039 ASVs. 2.6 Statistical analysis and visualization Statistical analyses and all plots were generated in R (v4.2.0, R Core Team). To evaluate microbiome diversity (Vegan v2.6-4, vegan R package) at sampling sites and composition differences by substrate (natural rock vs. glass slide), we calculated alpha and beta diversity from 16S rRNA gene sequences. Shannon-index differences were tested with one-way ANOVA (sampling sites as fixed factors) followed by Tukey test. Community composition differences among microbial communities grown on natural rocks or glass slides were analysed with NMDS based on Bray-Curtis dissimilarity [ 29 ] and group significance was assessed with a permutation test (n = 999). E. coli invader and plasmid abundance differences over time were tested with one-way ANOVA (time as a fixed factor; package: stats package v4.1.2 in R) followed by Tukey test. Resistome Shannon-index was calculated from ARG relative abundances, following the same method as microbiome analysis, using Vegan v2.6-4 ( vegan package). Metagenome beta diversity was analysed by Bray-Curtis dissimilarity and visualized with NMDS. Group comparison by treatment used ANOSIM ( vegan ), and ARG subtype correlation was visualized with biplot analysis in vegan (p ≤ 0.05). The top 20 significantly correlated ARGs associated with the sample of interest were further identified, based upon the residual sum of squares, representing the deviation between regression lines and the sample (i.e., visualized as vectors on the ordination plot) [ 30 ]. The top 20 genes were further visualized in heatmaps using gplots v3.1.3. Procrustes analysis was conducted to correlate resistome and microbiome ordination spaces using vegan , visualizing deviations as blue vector arrows, with significance assessed by a permutation test (n = 999). Spearman correlations between qPCR and metagenomics data (in copies per 16S rRNA gene) for E. coli and the pG527 plasmid were tested using stats package v4.1.2 in R, excluding an outlier (the 2nd replicate of D14- dwA ). Trendlines with confidence interval were calculated using the scales package (method = linear model) and plotted in R. 3. Results 3.1 Characterization of biofilm microbiome diversity Shannon diversity varied by sampling location, regardless of samples originating from natural rocks [ 15 ]. or grown on glass slides (n = 16; Fig. S4). Both substrate types exhibited similar diversity patterns, with upstream locations being less diverse than downstream ones. The difference between the two downstream sites was not statistically significant (Fig. S4). Beta diversity analysis revealed significant differences between microbial communities growing on natural rocks compared to glass slides (PERMANOVA, p = 0.004; Fig. S5A). Cyanobacteria and Bacillota were notably more abundant in natural rock biofilms (Fig. S5B). The number of taxa (ASVs) on glass slides was not significantly different from that on natural rocks (Kruskal-Wallis pairwise, p = 0.174), with both types predominantly composed of Pseudomonadota , Bacteroidota , and Actinomycetota (Fig. S5B). 3.2 Tracking ARB- E. coli invader and potential HGT transfer The abundance of the ARB- E. coli invader and its plasmid within indigenous biofilm communities were tracked independently with specific qPCR assays. The abundance of the invader and its plasmid varied among replicates and decreased over time regardless of biofilm origin (Fig. 2 A, B). The chromosomal ΔlacZY target indicating E. coli abundance peaked 24 hours post-invasion, then declined but remained detectable until the end of sampling (Fig. 2 A). The resistance plasmid abundance was consistently ~ 2-log higher than that of the invader (Fig. 2 B), but abundance patterns were overall consistent with E. coli . The plasmid and strain abundances showed considerable variability among replicates of the same origin (Fig. 2 A, B). We therefore analysed replicates individually. However, all sites and replicates showed the same transitory invasion pattern and the negative slope of the trendline of the plasmid-to-strain suggests that plasmid abundance was decreasing faster than E. coli chromosomal DNA, indicating a low potential conjugation rate and/or loss of the plasmid from the donor strain (Fig. 2 C). The ratio decreased significantly from the beginning to the end of the invasion in most of the invaded biofilms (ANOVA, p = 0.003 upL ; p = 0.038 upH ; p-value = 1.88e-05 dwB ; Fig. S6). However, the rate of the decrease varied. For instance, it was more stable in upH than in upL from day 7 to the end of the experiment. In the downstream dwA location, situated after the WW effluent, the ratio did not decrease significantly (ANOVA p-value = 0.0891 dwA ) indicating a higher stability of the plasmid and/or possible horizontal gene transfer (Fig. S6). At both downstream locations, ratios were initially lower than upstream but then decreased only slightly and were higher than in the upstream by the end of the experiment. Metagenomics supported the qPCR results: we obtained significant correlations between the E. coli and plasmid relative abundance measured by qPCR and calculated from metagenome data (p ≤ 0.008, Fig. S7). In addition, we quantified E. coli using both metagenomic read mapping and 16S rRNA amplicon sequencing. Estimates obtained from metagenomic data were highly consistent with those derived from 16S-based E. coli recounts, confirming that both approaches captured similar temporal patterns of invasion across treatments and biofilm origins (Fig. S8). The 16S- and metagenomic derived E. coli recounts reflected the same dynamics observed in the relative abundances (Fig. 2 A) with a sharp increase in E. coli on days 1–2 followed by a rapid decline to background levels by day 4–7. This agreement across independent molecular methods strengthens the robustness of our invasion assessments and supports the reliability of the observed differences between biofilm origins. 3.3 Responses of indigenous resistomes to ARB- E. coli We applied a custom bioinformatics workflow that allowed to determine the contribution of the invading E. coli to the biofilm metagenomes and to extract information about the indigenous metagenome and resistome. We found that the indigenous resistomes were highly diverse, and contained some ARGs of potential public health concern [ 31 – 34 ], such as genes for class-A (NMC), class-B (GIM, and IMP), and class-D (OXA) carbapenemases, and the ermB gene for macrolide-lincosamide-streptogramin (MLS) resistance (Fig. S9). Among them, some OXA genes (e.g, OXA-3, 5, 53, -119, -129, -198, and − 205) and ermB occurred abundantly across samples but their relative abundances did not exhibit substantial temporal dynamics (Fig. S9). Alpha-diversity of the indigenous resistomes (i.e., ARG subtypes identified in the metagenome), and total resistome (i.e. including both ARB- E. coli and indigenous ARGs) was quantified using the Shannon-index. The alpha diversity of the total resistomes increased markedly shortly after invader addition (Day 0 to 2) in the invaded biofilms (Fig. 3 A). Both replicates (Rep1 and 2) exhibited generally similar temporal dynamics. However, only three of four downstream biofilms ( dwA Rep 1, and dwB Rep 1 and 2) showed still elevated resistome diversity on day 4 compared to upstream sites where it had already dropped to values only slightly above day 0 values. In contrast, only minor changes were observed in the control biofilms. The dynamics thus generally mirrored the transient ARB- E. coli invasion (Fig. 2 A). In the indigenous resistome, alpha diversity was less dynamic, only replicate 1 of dwA and replicate 2 of dwB appear to show a small transient response (Fig. 3 B). Some changes and a divergence of samples was observed at later timepoints (day 1 and 14) in both invaded and control biofilms. Similar to alpha-diversity, beta-diversity analysis also showed very different patterns for the total resistomes of invaded biofilms in comparison to indigenous resistomes (Fig. 4 A and B). For the total ARGs, clear temporal changes connected to the transitory invasion are apparent while differences between biofilm origins were not significant (i.e., sampling sites; ANOSIM r = 0.032, p = 0.141; Fig. 4 A). Specifically, the total resistomes shifted a few days after the invasion (Day1 ~ 2) along the NMDS1 axis, indicating the ARB- E. coli impact on biofilm resistomes during the invasion events is captured on this axis. Total resistomes then returned to the pre-disturbance configuration (i.e., upH, dwB , and dwA in Fig. 4 A; upL, upH , and dwA in Fig. 4 B). Controls did not exhibit pronounced temporal changes along the ordination axis (Fig. S10). The indigenous resistomes significantly differed by biofilm origin (ANOSIM r = 0.79, p = 0.001, Fig. 4 B) but not over time. Some outliers, such as dwA -Rep2 D1, dwB -Rep1 D2, and upL -Rep2 D14 were noted (Fig. 4 B and S10). Similarly, controls mostly did not change much over the course of monitoring, forming tight clusters by biofilm origin, with one exception, upH -D14 (Fig. 4 B and S10). 3.4 Correlation between resistome and microbiome Between-group comparison using ANOSIM revealed that dissimilarity of total microbiomes significantly differed by sampling locations (r = 0.79, p = 0.001), not exhibiting pronounced temporal variation (Fig. 5 A). This indicates that biofilm origin rather than transient ARB- E. coli invasion mainly determined microbiome composition throughout the experiment. Results for indigenous microbiomes mirrored those for total microbiome, indicating that the presence of the single invader species had only a minor impact on the overall community structure (Fig. 5 B). Procrustes analysis revealed significant association between indigenous resistome and indigenous microbiome at a community level at 5% significance level (p = 0.025) (Fig. 5 C), indicating that resistomes and microbiomes were structurally correlated, suggesting that differences by origin (i.e., sites) in microbial community composition might explain those observed in the indigenous resistome. 4. Discussion 4.1 Contrasting microbiome diversity characterized the study sites Essential for the intended experiments was to obtain riverine biofilms on substrates suitable for experimentation yet representing natural contrasts in diversity and community composition. Our experimental procedures were in line with the standardization proposed by Freixa and colleagues [ 36 ] for the use of artificial substrates in freshwater biofilms studies, and the experimental approach using exposure units to grow biofilm in-situ (Fig. S2 ) proved viable and achieved the desired result. The comparison between biofilms grown on natural substrates (i.e., river rocks) and artificial substrates (i.e., glass-slides deployed in the river) ensured that the natural contrasts of microbial diversity and composition of the fresh waters biofilms were retained when using the artificial substrates (glass slides) for experimental purposes [ 35 ]. Differences between rocks and glass-slides biofilm diversity indices and community compositions were likely due to differences in light exposition, nutrient status and water flow they might be exposed to during the colonization period [ 36 , 37 ], and the comparatively young age of glass-slide biofilms. 4.2 The presence of the invading ARB- E. coli was short-lived, and conjugative transfer of its plasmid was limited The fate of the invading ARB- E. coli was overall independent of the origin and composition of the invaded biofilm. Similarly to what was observed in our previous study using the same E. coli strain [ 14 ], invasion was only transitory, referring to only low levels of the invader DNA remaining after 14 days. However, the degree of invasion during the first 2 days, indicated by the abundances of ARB- E. coli in the biofilm communities, was considerable and differed by the origin of biofilm. Although invaders did not establish long-term dominance, this could still be relevant under chronic exposure scenarios, where a continuous external source may sustain their presence in the biofilm despite their limited ability to naturally establish. For successful invasion, the invader must persist and be able to exploit the resources of the given habitat, overcoming the indigenous community [ 38 ]. Theoretically, healthy environments with high microbial diversity make the invasion process more difficult as a community with great diversity enhances its capacity to exploit the habitat resources [ 39 , 40 ]. Consistent with this hypothesis, the upstream low diversity biofilm ( upL ) allowed larger abundances of ARB- E. coli invader to establish than those of the upstream high diversity biofilm, potentially indicating a greater availability of ecological niches and less intraspecific competition among the low diversity community. Considering that biofilms provide conditions facilitating plasmid transfer through conjugation (i.e., close contact between cells as well as minimal shear forces), it has been assumed that conjugative gene transfer in microbial biofilms occurs with high efficiency [ 41 , 42 ]. As we observed overall decreasing plasmid-to-host ratios, the evidence suggests absence or a limited impact of horizontal gene transfer (HGT) (Fig. 2 C and Fig. S6), while indicating that the remaining ARB- E. coli populations carries increasingly lower numbers of plasmids over time. Plasmid loss is a phenomenon known to occur in non-selective environments, but generally as an evolutionary process during population growth [ 43 , 44 ] and often persistence is higher than expected [ 45 ]. Thus, the mechanism of the observed excess plasmid loss remains unclear. Biofilm communities located upstream, far from known pollution sources showed a more rapid loss of plasmid in comparison to the downstream biofilms (Fig. S6). This slower plasmid loss downstream may reflect selection pressures favoring plasmid retention in polluted environments, although contributions from conjugative transfer cannot be excluded. However, these results do not provide sufficient evidence that biofilm communities grown in the upstream “non-impacted” locations differ from the more impacted downstream locations in their natural resilience against the spread of antibiotic resistance, since an increase in the plasmid-to-host ratio was never observed. A limitation of this study is that only one type of Inc plasmid carrying resistance genes has been tested. The fate of AMR plasmids is the result of the complex interactions between the plasmid, its hosts and the environment [ 46 ]. For instance, abiotic factors that directly select for plasmid-encoded traits, such as antibiotics or heavy metals, can inhibit the growth of plasmid-free cells. Other abiotic factors such as pH, temperature or nutrients, may directly affect some of the mechanisms that determine whether a plasmid will persist in a population: plasmid fitness cost, replication and segregation rate, and horizontal transfer rate [ 15 , 46 , 47 ]. Therefore, the ‘invasibility’ of the biofilm by plasmids, or the proportion of biofilm cells that acquired plasmids within a few hours, could depend critically on the type of plasmid, on the time of biofilm exposure to the donor, on the biofilm age and on the ability of the plasmid donor to attach to the biofilm [ 42 ]. 4.3 Does biofilm diversity and composition matter for the fate of AMR? The presence of the invader profoundly impacted the resistomes diversity and structure, but this was primarily traced to the direct presence of ARB- E. coli itself using our method of observing the indigenous resistome. Nevertheless, this finding is remarkable in showing that an invasion that did not affect the overall microbiome composition resulted in pronounced changes to the resistome. We further reported on changes in the indigenous community resistomes that were not associated with the invasive ARB- E. coli . As introduced earlier, both HGT and shifts of community are known drivers for changes in resistome [ 2 , 3 , 8 , 9 ]. The extent of HGT does not appear to have been extensive enough to affect overall indigenous resistome structure. This is in line with the results of qPCR-based analysis (3.2, Fig. 2 , 4.2). The shifts in microbiome composition over the course of monitoring were likewise limited, as evidenced by the microbial community structure clustering mainly by biofilm origin rather than time (Fig. 5 A, B). The significant structural correlation between microbiomes and resistomes therefore mostly confirms that differences in community composition between biofilm origins influenced the resistomes. The early fluctuations observed in the indigenous resistome (Fig. 3 B and 4 B), although minor compared to the temporal trends observed in the total resistome, raise the question of whether they reflect true biological responses or technical noise introduced during indigenous metagenome filtering and normalization. By manually crosschecking our data, we did not observe evidence that the bioinformatic filtering workflow introduced artifacts that would be problematic, i.e. incomplete filtering of resistance genes of E. coli MG1655. The consistency across replicates also suggests that part of this variation may represent genuine transient community adjustments following the initial disturbance from exposure to the invading organism. From day 7 onward, the plasmid-to-strain ratio remained markedly more stable in upH than in upL, indicating a greater persistence of the plasmid in the upH community (Fig. S6). Likewise, in the downstream dwA biofilms, located immediately after the WWTP effluent, the ratio did not significantly decrease (Fig. S6), supporting a higher persistence of the introduced plasmid and/or increased HGT potential in this environment. This stability, despite the decline in E. coli abundance, aligns with the notion that downstream environments receiving continuous anthropogenic inputs may sustain persistent alterations of their resistomes even when invaders fail to establish long-term. In both locations, the higher diversity site appeared to support a slightly higher persistence of the plasmid over the duration of the experiment. This was not expected according to our initial hypothesis, and we can currently not speculate regarding the cause. The magnitude of community shift and HGT, has been known to depend strongly upon biotic interactions (i.e., composition of introduced and indigenous communities) [ 12 , 15 ]. Our study in this respect confirms that the composition of the indigenous community is an important factor - anthropogenically impacted indigenous biofilm communities (i.e., dwA ) exhibited the lowest ecological resilience in response to ARB-invasion, at least in one of the repeated cases. WW-impacted biofilms evolve in highly specialized environments, where a chronic perturbation (i.e., wastewater effluents) exists. The effluents contain high biomass with a stable microbial and resistome composition that is distinct from undisturbed waters [ 8 , 48 ]. Thus, the locations receiving high volumes of wastewater effluents can be expected to be more favourable environments for colonization by organisms with a wastewater origin [ 4 , 49 ]. Exposed to chronic perturbation or constant invasion pressure, communities may reach alternative stable states adapted to function in the new environment [ 50 , 51 ]. Removal of chronic perturbation does not necessarily lead to the reversal of community assembly because resilience can be reduced. In this sense, dwA might have exhibited reduced resilience under the absence of a chronic perturbation in the case of our experiment setting (i.e., using filtered river waters, where biomass particles were removed), allowing certain organisms to colonize successfully in the later stage of ecological succession. This mainly shows that perturbation can have rather unpredictable effects on the resistome of a microbial community. The response of the indigenous resistome was also stochastic - only one of the replicates exhibited significant alteration in the later stage of ecological succession (Day14). As discussed earlier, the composition of microbial community itself and the perturbation, can be important factors; the composition of the microbiome in the two replicates for dwA may not have been identical, leading to varying responses to the invader between the replicates. These observations warrant further study, including on biofilms affected by multiple sources of contamination and multiple ARB-invasions. Conclusions From a One-Health perspective, biofilms are considered potential hotspots where transferrable AMR could evolve and thrive. In this study, we reported that the indigenous freshwater biofilm microbiomes were largely resilient and did not respond to the invasion of an ARB- E. coli whereas resistomes were temporally affected. The limited colonization success of the invader and limited extent of indirect impacts of the invasion (i.e., HGT and changes in indigenous resistomes) within freshwater biofilms, indicates that the potential health risks posed by freshwater biofilms, might not be as high as we initially hypothesized. However, specific groups of indigenous ARGs, including those of public health concern ‘can’ still increase in biofilms from wastewater-impacted sites. This suggests that the potential public health risks can persist locally, particularly in highly wastewater-impacted freshwater environments. Therefore, future studies involving additional wastewater-impacted sites are required. Once confirmed, these results could motivate measures to maintain highly resilient microbial biofilms in rivers as one way to reduce the spread of resistances in natural environments. Declarations Declaration of competing interest All authors declare no financial or non-financial competing interests. Funding This study was supported by the ANTIVERSA project (BiodivERsA, European Union) and the Swiss National Science Foundation (Switzerland) grant 186531. GG was supported by a Junior Leader Incoming Fellowship contract (LCF/BQ/PI23/11970040) funded by La Caixa Foundation (ID 100010434). Author Contribution GG : Conceptualization, Investigation, Formal analysis, Visualization, Writing - Original Draft, Review and Editing; JL : Investigation, Formal analysis, Visualization, Writing - Original Draft, Review and Editing; OH : Investigation, Review and Editing; KB : Investigation, Review and Editing; HB : Conceptualization, Funding acquisition, Supervision, Writing - Review and Editing. Acknowledgement We acknowledge our ANTIVERSA colleagues for the joint development of the experimental design used in this study and specifically X. Bellanger and C. Merlin from LCPME (Nancy, France) for providing E. coli MG1655 (CM2372) ∆lacZY and pG527 IncPα plasmid, used to study microbial invasion and for providing the standards for their quantification through qPCR. We thank Eawag workshop team for their help building the exposure units and the flumes, as well as with the experimental setup. Finally, we thank the Oberglatt Fishery Department for their support and permission to access to the Glatt River for colonizing biofilm on the glass slides and sampling. Data Availability All the sequences were deposited in the NCBI (National Center for Biotechnology Information) under SRA accession (sequence read archive) number: PRJNA1070575 (Microbiome data) and PRJNA1129054 (Metagenomics data). All data are presented in the manuscript and supplementary material. Raw data that support the findings of this study are available from the corresponding author upon request. References Hernando-Amado S, Coque TM, Baquero F, Martínez JL. Defining and combating antibiotic resistance from One Health and Global Health perspectives. Nat Microbiol 2019; 4: 1432–1442. Smalla K, Cook K, Djordjevic SP, Klümper U, Gillings M. Environmental dimensions of antibiotic resistance: assessment of basic science gaps. FEMS Microbiology Ecology 2018; 94. Woods LC, Gorrell RJ, Taylor F, Connallon T, Kwok T, McDonald MJ. Horizontal gene transfer potentiates adaptation by reducing selective constraints on the spread of genetic variation. Proc Natl Acad Sci USA 2020; 117: 26868–26875. Matviichuk O, Mondamert L, Geffroy C, Gaschet M, Dagot C, Labanowski J. River Biofilms Microbiome and Resistome Responses to Wastewater Treatment Plant Effluents Containing Antibiotics. Front Microbiol 2022; 13: 795206. Reichert G, Hilgert S, Alexander J, Rodrigues De Azevedo JC, Morck T, Fuchs S, et al. Determination of antibiotic resistance genes in a WWTP-impacted river in surface water, sediment, and biofilm: Influence of seasonality and water quality. Science of The Total Environment 2021; 768: 144526. Bagra K, Bellanger X, Merlin C, Singh G, Berendonk TU, Klümper U. Environmental stress increases the invasion success of antimicrobial resistant bacteria in river microbial communities. Science of The Total Environment 2023; 904: 166661. Ju F, Beck K, Yin X, Maccagnan A, McArdell CS, Singer HP, et al. Wastewater treatment plant resistomes are shaped by bacterial composition, genetic exchange, and upregulated expression in the effluent microbiomes. The ISME Journal 2019; 13: 346–360. Lee J, Ju F, Maile-Moskowitz A, Beck K, Maccagnan A, McArdell CS, et al. Unraveling the riverine antibiotic resistome: The downstream fate of anthropogenic inputs. Water Research 2021; 197: 117050. Luo G, Li B, Li L-G, Zhang T, Angelidaki I. Antibiotic Resistance Genes and Correlations with Microbial Community and Metal Resistance Genes in Full-Scale Biogas Reactors As Revealed by Metagenomic Analysis. Environ Sci Technol 2017; 51: 4069–4080. Jones ML, Rivett DW, Pascual-García A, Bell T. Relationships between community composition, productivity and invasion resistance in semi-natural bacterial microcosms. eLife 2021; 10: e71811. Mawarda PC, Lakke SL, Dirk Van Elsas J, Salles JF. Temporal dynamics of the soil bacterial community following Bacillus invasion. iScience 2022; 25: 104185. Albright MBN, Sevanto S, Gallegos-Graves LV, Dunbar J. Biotic Interactions Are More Important than Propagule Pressure in Microbial Community Invasions. mBio 2020; 11: e02089-20. Merlin C, McAteer S, Masters M. Tools for Characterization of Escherichia coli Genes of Unknown Function. J Bacteriol 2002; 184: 4573–4581. Gionchetta G, Snead D, Semerad S, Beck K, Pruden A, Bürgmann H. Dynamics of antibiotic resistance markers and Escherichia coli invasion in riverine heterotrophic biofilms facing increasing heat and flow stagnation. Science of The Total Environment 2023; 893: 164658. Klümper U, Gionchetta G, Catão E, Bellanger X, Dielacher I, Elena AX, et al. Environmental microbiome diversity and stability is a barrier to antimicrobial resistance gene accumulation. Commun Biol 2024; 7: 706. Sagrillo C, Changey F, Bellanger X. Bacteriophages vehiculate a high amount of antibiotic resistance determinants of bacterial origin in the Orne River ecosystem. Environmental Microbiology 2022; 24: 4317–4328. Heuer H, Smalla K. Manure and sulfadiazine synergistically increased bacterial antibiotic resistance in soil over at least two months. Environmental Microbiology 2007; 9: 657–666. Bustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. Clinical Chemistry 2009; 55: 611–622. Takai K, Horikoshi K. Rapid Detection and Quantification of Members of the Archaeal Community by Quantitative PCR Using Fluorogenic Probes. Appl Environ Microbiol 2000; 66: 5066–5072. Andrews, Simon. FastQC: a quality control tool for high throughput sequence data. Babraham Bioinformatics, Babraham Institute, Cambridge, United Kingdom . Arango-Argoty G, Garner E, Pruden A, Heath LS, Vikesland P, Zhang L. DeepARG: a deep learning approach for predicting antibiotic resistance genes from metagenomic data. Microbiome 2018; 6: 23. Yang Y, Jiang X, Chai B, Ma L, Li B, Zhang A, et al. ARGs-OAP: online analysis pipeline for antibiotic resistance genes detection from metagenomic data using an integrated structured ARG-database. Bioinformatics 2016; 32: 2346–2351. Lee, J., Xiang, K., Au, E. et al. Longitudinal monitoring of sewershed resistomes in socioeconomically diverse urban neighborhoods. Commun Med 5, 7 (2025). Tarracchini C, Alessandri G, Fontana F, Rizzo SM, Lugli GA, Bianchi MG, et al. Genetic strategies for sex-biased persistence of gut microbes across human life. Nat Commun 2023; 14: 4220. Liu Y, Elworth RAL, Jochum MD, Aagaard KM, Treangen TJ. De novo identification of microbial contaminants in low microbial biomass microbiomes with Squeegee. Nat Commun 2022; 13: 6799. Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Lozupone CA, Turnbaugh PJ, et al. Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. Proc Natl Acad Sci USA 2011; 108: 4516–4522. Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. Nat Methods 2016; 13: 581–583. Bolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 2019; 37: 852–857. Legendre L, Legendre P. Numerical ecology. 2012. Elsevier , Amsterdam. Parkins M, Lee J, Xiang K, Au E, Sarabi S, Acosta N, et al. Longitudinal monitoring of sewershed resistomes in socioeconomically diverse urban neighbourhoods. 2024. Queenan AM, Bush K. Carbapenemases: the Versatile β-Lactamases. Clin Microbiol Rev 2007; 20: 440–458. Rossolini GM, D’Andrea MM, Mugnaioli C. The spread of CTX-M-type extended-spectrum β-lactamases. Clinical Microbiology and Infection 2008; 14: 33–41. Evans BA, Amyes SGB. OXA β-Lactamases. Clin Microbiol Rev 2014; 27: 241–263. Mugnier P, Dubrous P, Casin I, Arlet G, Collatz E. A TEM-derived extended-spectrum beta-lactamase in Pseudomonas aeruginosa . Antimicrob Agents Chemother 1996; 40: 2488–2493. Sabater S, Guasch H, Ricart M, Romaní A, Vidal G, Klünder C, et al. Monitoring the effect of chemicals on biological communities. The biofilm as an interface. Anal Bioanal Chem 2007; 387: 1425–1434. Freixa A, Ortiz-Rivero J, Sabater S. Artificial substrata to assess ecological and ecotoxicological responses in river biofilms: Use and recommendations. MethodsX 2023; 10: 102089. Biggs BJF. Artificial substrate exposure times for periphyton biomass estimates in rivers. New Zealand Journal of Marine and Freshwater Research 1988; 22: 507–515. Kinnunen M, Dechesne A, Proctor C, Hammes F, Johnson D, Quintela-Baluja M, et al. A conceptual framework for invasion in microbial communities. The ISME Journal 2016; 10: 2773–2779. Hector A, Dobson K, Minns A, Bazeley-White E, Hartley Lawton J. Community diversity and invasion resistance: An experimental test in a grassland ecosystem and a review of comparable studies. Ecological Research 2001; 16: 819–831. Mallon CA, Le Roux X, Van Doorn GS, Dini-Andreote F, Poly F, Salles JF. The impact of failure: unsuccessful bacterial invasions steer the soil microbial community away from the invader’s niche. The ISME Journal 2018; 12: 728–741. Molin S, Tolker-Nielsen T. Gene transfer occurs with enhanced efficiency in biofilms and induces enhanced stabilisation of the biofilm structure. Current Opinion in Biotechnology 2003; 14: 255–261. Król JE, Wojtowicz AJ, Rogers LM, Heuer H, Smalla K, Krone SM, et al. Invasion of E. coli biofilms by antibiotic resistance plasmids. Plasmid 2013; 70: 110–119. Smith MA, Bidochka MJ. Bacterial fitness and plasmid loss: the importance of culture conditions and plasmid size. Can J Microbiol 1998; 44: 351–355. Bellanger, X., Guilloteau, H., Bonot, S. & Merlin, C. Demonstrating plasmid-based horizontal gene transfer in complex environmental matrices: A practical approach for a critical review. Sci. Total Environ. 2014; 493, 872–882. Carroll AC, Wong A. Plasmid persistence: costs, benefits, and the plasmid paradox. Can J Microbiol 2018; 64: 293–304. Castañeda-Barba S, Top EM, Stalder T. Plasmids, a molecular cornerstone of antimicrobial resistance in the One Health era. Nat Rev Microbiol 2024; 22: 18–32. Smalla K, Jechalke S, Top EM. Plasmid Detection, Characterization, and Ecology. In: Tolmasky ME, Alonso JC (eds). Plasmids . 2015. ASM Press, Washington, DC, USA, pp 445–458. Mansfeldt C, Deiner K, Mächler E, Fenner K, Eggen RIL, Stamm C, et al. Microbial community shifts in streams receiving treated wastewater effluent. Science of The Total Environment 2020; 709: 135727. Zhang Z, Zhang G, Ju F. Using Culture-Enriched Phenotypic Metagenomics for Targeted High-Throughput Monitoring of the Clinically Important Fraction of the β-Lactam Resistome. Environ Sci Technol 2022; 56: 11429–11439. Shade A, Peter H, Allison SD, Baho DL, Berga M, Bürgmann H, et al. Fundamentals of Microbial Community Resistance and Resilience. Front Microbio 2012; 3. Potts LD, Douglas A, Perez Calderon LJ, Anderson JA, Witte U, Prosser JI, et al. Chronic Environmental Perturbation Influences Microbial Community Assembly Patterns. Environ Sci Technol 2022; 56: 2300–2311. Additional Declarations No competing interests reported. Supplementary Files GGnpjamrSItables.xlsx GGnpjamrSI.docx Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 17 Feb, 2026 Reviews received at journal 18 Jan, 2026 Reviews received at journal 04 Jan, 2026 Reviewers agreed at journal 23 Dec, 2025 Reviewers agreed at journal 22 Dec, 2025 Reviewers agreed at journal 20 Dec, 2025 Reviewers agreed at journal 18 Dec, 2025 Reviewers invited by journal 18 Dec, 2025 Editor assigned by journal 05 Dec, 2025 Submission checks completed at journal 04 Dec, 2025 First submitted to journal 03 Dec, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8270852","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":556827948,"identity":"e7ce8774-436a-4823-a38c-e2b2fe38b12b","order_by":0,"name":"Giulia Gionchetta","email":"data:image/png;base64,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","orcid":"","institution":"Institute of Environmental Assessment and Water Research (IDAEA), Spanish Council of Scientific Research (CSIC)","correspondingAuthor":true,"prefix":"","firstName":"Giulia","middleName":"","lastName":"Gionchetta","suffix":""},{"id":556827956,"identity":"60be84dd-6fea-4848-a000-bf77c5303eb4","order_by":1,"name":"Jangwoo Lee","email":"","orcid":"","institution":"University of Calgary","correspondingAuthor":false,"prefix":"","firstName":"Jangwoo","middleName":"","lastName":"Lee","suffix":""},{"id":556827960,"identity":"01b94e83-f806-496f-a45b-31d6eaae11ac","order_by":2,"name":"Oliver Hansen","email":"","orcid":"","institution":"Swiss Federal Institute of Aquatic Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Oliver","middleName":"","lastName":"Hansen","suffix":""},{"id":556827961,"identity":"c83d6ec1-809e-434d-9a59-f89e54952902","order_by":3,"name":"Karin Beck","email":"","orcid":"","institution":"Swiss Federal Institute of Aquatic Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Karin","middleName":"","lastName":"Beck","suffix":""},{"id":556827962,"identity":"f71f28fe-af77-4bf7-9c42-60508904da9d","order_by":4,"name":"Helmut Bürgmann","email":"","orcid":"","institution":"Swiss Federal Institute of Aquatic Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Helmut","middleName":"","lastName":"Bürgmann","suffix":""}],"badges":[],"createdAt":"2025-12-03 13:38:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8270852/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8270852/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97797206,"identity":"4a7af1a8-c2ac-4617-a07d-4f13c95ee5a4","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":699275,"visible":true,"origin":"","legend":"","description":"","filename":"GGnpjamrmanuscriptrev.docx","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/1c89ea85f0efdb28db5282cd.docx"},{"id":97797204,"identity":"6156d090-40d1-4575-aca0-12991e1aec99","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8092,"visible":true,"origin":"","legend":"","description":"","filename":"7d279d029deb40b5bed2f32492e40c70.json","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/464ed6dca77c5f671a4ab7e8.json"},{"id":97797218,"identity":"10033242-30b3-4623-95a9-cfb30603fffd","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":5558547,"visible":true,"origin":"","legend":"","description":"","filename":"GGnpjamrSI.docx","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/799898f725f6113d553300f8.docx"},{"id":97797208,"identity":"8a620533-4d7c-450e-b402-aee381cb2120","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":18631,"visible":true,"origin":"","legend":"","description":"","filename":"GGnpjamrSItables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/4e0155f8d647da52471adc6f.xlsx"},{"id":97797216,"identity":"6d711bc7-6bed-416d-b185-0a9884b8ec7d","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":133430,"visible":true,"origin":"","legend":"","description":"","filename":"7d279d029deb40b5bed2f32492e40c701enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/42b99e9635e0447a0db6b828.xml"},{"id":97797209,"identity":"af0fcdc8-00f8-49f6-8abf-5b0a38c6dafa","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":26482,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/6499dc53a74d99b2903d4edc.png"},{"id":97797212,"identity":"6572e2fa-6566-42e8-8aa3-8243987d4f89","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"png","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":25725,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/fede777b7d8a5c348d190522.png"},{"id":97898425,"identity":"45d1d60f-434a-41e2-8e0f-e5fcdf540364","added_by":"auto","created_at":"2025-12-10 15:39:09","extension":"png","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42121,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/30538c0376d302e8f35f3b10.png"},{"id":97898045,"identity":"50cf5098-af40-4071-a071-fc482cba6002","added_by":"auto","created_at":"2025-12-10 15:38:37","extension":"png","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":49750,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/7b82e7bc05879fa4e3132ffe.png"},{"id":97797210,"identity":"310c6eca-affc-43fc-a4e8-f24b2817d7f6","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"png","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":57787,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/91748389d5ee4fa42ea58269.png"},{"id":97896225,"identity":"aaec4e51-425c-456f-8c8c-650dbeb749d6","added_by":"auto","created_at":"2025-12-10 15:36:11","extension":"xml","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":131671,"visible":true,"origin":"","legend":"","description":"","filename":"7d279d029deb40b5bed2f32492e40c701structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/e85f67aa5fc4340f9f016044.xml"},{"id":97898152,"identity":"eb895472-2f1f-4009-89d1-e66365a4436c","added_by":"auto","created_at":"2025-12-10 15:38:43","extension":"html","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":145110,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/5313d66b32258e79ad958e19.html"},{"id":97797203,"identity":"76e31a00-ad6c-4e6a-95c0-070889ed6752","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":62890,"visible":true,"origin":"","legend":"\u003cp\u003eSchematic overview of the metagenomics workflows applied to separate indigenous metagenomes and invader genome:\u003cstrong\u003e \u003c/strong\u003eA) Indigenous resistome workflow, enabling identification and enumeration of the antimicrobial resistance genes (ARGs) matching reads that were not mapped onto \u003cem\u003eE. coli\u003c/em\u003e MG1655 chromosomal and plasmid genomes; B) \u003cem\u003eE.coli\u003c/em\u003e MG1655 workflow, enabling identification and quantification of the reads associated with \u003cem\u003eE.coli\u003c/em\u003e MG1655 chromosomal and plasmid genomes.\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/c4a3f2319cbdacf695279a4a.png"},{"id":97898799,"identity":"4abb5b7a-bd56-4efa-ae83-6a8937ed806d","added_by":"auto","created_at":"2025-12-10 15:39:46","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":98651,"visible":true,"origin":"","legend":"\u003cp\u003eRelative abundance of \u003cstrong\u003eA) \u003c/strong\u003e\u003cem\u003eE. coli\u003c/em\u003e invaders (\u003cem\u003e∆lacZY\u003c/em\u003e gene copies/cm\u003csup\u003e2\u003c/sup\u003e) and \u003cstrong\u003eB)\u003c/strong\u003e plasmid (\u003cem\u003epG527\u003c/em\u003e gene/cm\u003csup\u003e2\u003c/sup\u003e) tracked for 15 days after the invasion (time in days on the x-axis). The time before spiking (Time 0) is not shown being zero for all the cases. Error bars denote the standard deviation associated with the average of technical replicates for each time point. \u003cstrong\u003eC) \u003c/strong\u003ePlasmid-to-strain genome ratio calculated for each replicate as proxy of potential horizontal gene transfer, HGT. Trend lines were calculated over all invaded biofilms, shaded area indicates the 95% confidence interval.\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/7ec5b0488e4cc63a9dac2942.png"},{"id":97898140,"identity":"15ff2e10-66d1-46b0-acab-1ebe2b3e7d7a","added_by":"auto","created_at":"2025-12-10 15:38:42","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":137271,"visible":true,"origin":"","legend":"\u003cp\u003eTemporal dynamics of Shannon index for \u003cstrong\u003e(A)\u003c/strong\u003etotal resistome (ensemble of ARGs); \u003cstrong\u003e(B) \u003c/strong\u003eindigenous resistome: \u003cem\u003eupL\u003c/em\u003e (upstream low diversity), \u003cem\u003eupH\u003c/em\u003e (upstream high diversity), \u003cem\u003edwB\u003c/em\u003e(downstream before wastewater impact), and \u003cem\u003edwA\u003c/em\u003e (downstream after wastewater impact), and corresponding control sets (dashed lines).\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/0458762b72b325fca93374bd.png"},{"id":97898858,"identity":"a1d8459f-48be-4710-8943-e1130b728524","added_by":"auto","created_at":"2025-12-10 15:39:52","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":154126,"visible":true,"origin":"","legend":"\u003cp\u003eThe non-metric multidimensional scaling (NMDS) ordination plot for visualizing dissimilarities in \u003cstrong\u003e(A) \u003c/strong\u003etotal ARGs identified in this study (stress = 0.033); \u003cstrong\u003e(B)\u003c/strong\u003e indigenous ARGs (stress = 0.056); variance ellipses were calculated using standard deviation of point scores (confidence limits = 0.95) for each invaded biofilm. Only data points for invaded biofilms are shown. Empty circles and R1 indicate replicate 1; filled circles and R2 denote replicate 2. Day indicates days after invasion. For controls, see Fig. S10.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/d774e9d32bfb729c062a952f.png"},{"id":97897903,"identity":"baa14000-4ced-4424-8843-dad320549134","added_by":"auto","created_at":"2025-12-10 15:38:26","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":151746,"visible":true,"origin":"","legend":"\u003cp\u003enon-metric multidimensional scaling (NMDS) ordination plot visualizing dissimilarities in \u003cstrong\u003e(A)\u003c/strong\u003e total microbiome composition (stress = 0.13); \u003cstrong\u003e(B) \u003c/strong\u003eindigenous microbiome composition (stress = 0.13) between samples; variance ellipses were calculated using standard deviation of point scores (confidence limits = 0.95) for each invaded biofilm. \u003cstrong\u003e(C)\u003c/strong\u003e Procrustes analysis between indigenous resistomes (symbols) and indigenous microbiome (tips of arrows) (r = 0.31, p = 0.025).\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/43a9b2e5eb2caf844e1d1d69.png"},{"id":98622643,"identity":"cfe58bf2-3243-4d60-a69e-a4a1726565ca","added_by":"auto","created_at":"2025-12-19 16:59:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1460851,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/abaec961-45e7-460d-844f-c8e76ae251c2.pdf"},{"id":97797202,"identity":"1dda963b-e39f-4a3b-8244-e070db01a398","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":18631,"visible":true,"origin":"","legend":"","description":"","filename":"GGnpjamrSItables.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/773bc609c4f068844e814da3.xlsx"},{"id":97797220,"identity":"ff42906a-6ddc-4942-9407-9b72b935963f","added_by":"auto","created_at":"2025-12-09 12:57:32","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":5558547,"visible":true,"origin":"","legend":"","description":"","filename":"GGnpjamrSI.docx","url":"https://assets-eu.researchsquare.com/files/rs-8270852/v1/6faf4e775fefd1859ebb3a38.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Invasion dynamics of antimicrobial-resistant E. coli in river biofilms: impacts on the resistome, microbiomes, and horizontal gene transfer","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eThe role of the environment in the spread and evolution of antimicrobial resistance (AMR) is one focus of the One Health approach. From this perspective, the environment is seen as a large reservoir of active or latent antibiotic resistance genes (ARGs), which can transfer to actual or potential pathogens, especially under selective pressures. Given that the spread of resistant bacteria and their antibiotic resistance determinants among environmental microbiomes can possibly influence downstream transmission risks and evolutionary trajectories by altering bacterial population genetics [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e], it is important to understand the dynamics and environmental controls of these processes.\u003c/p\u003e\u003cp\u003eHorizontal gene transfer (HGT) is one of the most significant pathways of resistance evolution [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Specifically, HGT between bacteria can lead to the spread of antibiotic resistant traits into previously susceptible pathogens or into abundant environmental species [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. The conjugative transfer of plasmids carrying one or multiple ARGs is a particularly important mechanism of HGT and resistance dissemination, because it can lead to the transfer of multiple resistances in a single transfer event. Microbiomes of river biofilms exposed to wastewater (WW) discharge could be especially prone to the conjugative transfer of antibiotic resistant traits due to the chronic pollution of WW with antibiotic residues and AMR vectors (i.e. ARB, resistance plasmids and genes) [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e], with the extent of impact varying according to physio-chemical factors, such as temperature and turbidity [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e] as well as the impact of environmental stressors or selective agents (i.e. heavy metals, biocides or antibiotic residues) often co-released in WW effluents [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eFurthermore, the composition and abundance of environmental ARGs (i.e., the resistome of a microbiome) is frequently observed to be related to changes in microbial community composition, that may, but must not necessarily, be related to specific AMR-selective pressures [\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. For instance, recent studies revealed that in response to disturbance, resistomes were structurally correlated with microbiomes in water compartments, such as freshwater [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e] or WW [\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. However, there is presently a lack of such analyses for river biofilms. There have been a few studies where certain groups of microbial taxa such as pathogenic or WW-related groups were correlated with selected ARGs [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. However, the magnitude and importance (i.e., relative to HGT) of this effect remains unexplored, highlighting the need for a study that simultaneously assesses both factors.\u003c/p\u003e\u003cp\u003eStudies exploring resistome rsponse to invasive microbial species are likewise currently not available. However, several studies explored the effects of invasion on the invaded microbial communities [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Many of these studies found a significant impact, for example transient or permanent increase in overall taxonomic diversity, with varied results depending upon the biotic or abiotic conditions experienced by the indigenous communities [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eGiven that changes in resistomes can be attributed to presence of invading species, HGT, changes in microbiomes, or all of these, this question necessitates further exploration. In an earlier study, we observed that invasions of river-grown biofilms with \u003cem\u003eE. coli MG1655\u003c/em\u003e (strain CM2372, derived from \u003cem\u003eE. coli\u003c/em\u003e EDCM367 [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e], ARB-\u003cem\u003eE. coli\u003c/em\u003e) were transient, regardless of temperature and flow conditions [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. However, in the previous study we did not monitor the persistence of the conjugative plasmid of this organism, which could potentially transfer to native bacteria persisting in the biofilm. Likewise, the ability of invasion and HGT to alter river biofilm microbiomes and thus indirectly their resistomes has not been assessed.\u003c/p\u003e\u003cp\u003eHere we performed experiments with a non-pathogenic ARB-\u003cem\u003eE. coli\u003c/em\u003e, possessing a plasmid (\u003cem\u003epG527)\u003c/em\u003e harbouring the \u003cem\u003enptII\u003c/em\u003e resistance gene invading indigenous biofilms obtained from four locations of the river Glatt, Switzerland. Biofilms from these locations differed in microbial community composition and diversity and their pre-exposure to WW.\u003c/p\u003e\u003cp\u003eThe goals of this study were to (1) quantify the survival of the ARB-\u003cem\u003eE. coli\u003c/em\u003e in the different biofilms and, (2) assess persistence of its resistance-conferring plasmid \u003cem\u003epG527\u003c/em\u003e to evaluate HGT. Further to (3) evaluate whether a transient invasion would drive changes in the overall community structure of the biofilm microbiomes and their resistomes. We hypothesized that previous exposure to WW and indigenous biofilm diversity are influencing factors. Specifically, we hypothesized that (1) the most upstream (not previously exposed to evident source of pollution) and least diverse biofilms would be the most vulnerable to the invasion and to (2) HGT by conjugative plasmid transfer. We further expected that (3) the ARB-\u003cem\u003eE. coli\u003c/em\u003e invasion would induce correlated changes of resistomes and microbiomes over time, indicating resistomes are linked to shifts in community structure. The fate and the invasion dynamics of ARB-\u003cem\u003eE. coli\u003c/em\u003e and its plasmid, as well as the biofilms total and native (\u003cem\u003eE. coli\u003c/em\u003e effect eliminated) resistomes and microbiomes were monitored through qPCR, shotgun metagenomics and amplicon sequencing, tracking temporal changes in resistomes and microbiomes, and their potential association, employing multivariate analysis.\u003c/p\u003e"},{"header":"2. Material and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e\u003cb\u003e2.1 Study locations and biofilm colonization\u003c/b\u003e\u003c/h2\u003e\u003cp\u003eThe study sites are located along the river Glatt, a 25 km-long tributary of the Thur in northeastern Switzerland (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The two \u0026ldquo;upstream\u0026rdquo; sites (\u003cem\u003eupL\u003c/em\u003e and \u003cem\u003eupH\u003c/em\u003e) are located about three kilometres upstream of the Oberglatt WWTP, one in the Glatt and one in a tributary - both sites and their mostly forested watersheds experience minimal human impact. The two \u0026ldquo;downstream\u0026rdquo; sites are located a few hundred meters up- (\u0026ldquo;before\u0026rdquo;, \u003cem\u003edwB\u003c/em\u003e) and downstream (\u0026ldquo;after\u0026rdquo;, \u003cem\u003edwA\u003c/em\u003e) of the WWTP discharge point. Both sites are downstream of a tributary (Dorfbach) draining the town of Gossau, and agricultural fields, indicating increased human impact. Sampling locations were chosen based on preliminary microbiome diversity screenings of native epilithic biofilms [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]: \u003cem\u003eupL\u003c/em\u003e and \u003cem\u003eupH\u003c/em\u003e had low and high native biofilm microbial diversity, respectively, while \u003cem\u003edwB\u003c/em\u003e and \u003cem\u003edwA\u003c/em\u003e had high and highest biofilm biodiversity. In May 2022, the river biofilms for the experiments were grown in situ for 5 weeks on glass slides in environmental exposure units (SI1 and Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e; [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Flumes experiment and biofilm collection\u003c/h2\u003e\u003cp\u003eTwelve flumes were set up in a 20\u0026deg;C climate chamber, with three flumes per sampling site: one control and two receiving the \u003cem\u003eE. coli\u003c/em\u003e invader (Fig. S3A). Each flume contained 6.5 L of filtered river water in a continuous recirculation system (8 L total water volume, details in SI2). To ensure sufficient biomass recovery and to minimize the impact of slide-to-slide variability, every biomass sample consisted of pooled biofilm of three randomly chosen microscope glass slides, gently scraped with a spatula and washed with sterile water from both sides (Fig. S3B). Samples were taken at time 0 and 1, 2, 4, 7, 10, and 14 days post-invasion. Biofilms were centrifuged, supernatant removed, weighed (wet biomass, mg), and 0.25 g stored at -20\u0026deg;C for DNA extraction.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Inoculum of antibiotic resistant \u003cem\u003eE. coli MG1655\u003c/em\u003e preparation\u003c/h2\u003e\u003cp\u003eA non-pathogenic \u003cem\u003eE.coli MG1655 ΔLacZY\u003c/em\u003e strain possessing a native conjugative IncPα plasmid carrying the \u003cem\u003enptII\u003c/em\u003e resistance gene, was grown overnight at 37\u0026deg;C on a shaker table operating at 160 rpm in three falcon tubes each with 10 mL of Luria-Bertani (LB) broth containing 50 \u0026micro;g/mL of kanamycin ([\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e], SI3). To start the invasion event (time 0), \u003cem\u003eMG1655\u003c/em\u003e cell suspension was mixed into the 10% water volume (800 mL) added during the weekly water renewal, in each flume apart from those of controls, to achieve a final concentration of 10\u003csup\u003e7\u003c/sup\u003e cells/mL in the recirculation system.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003e2.4 DNA extraction and quantitative PCR (qPCR)\u003c/h2\u003e\u003cp\u003eDNA extraction was conducted using the DNeasy PowerMax Soil Kit (Qiagen, Germany) following the manufacturer\u0026rsquo;s instructions. Extraction blanks confirmed the absence of DNA contamination. The quality and quantity of the extracted DNA were assessed with a NanoDrop One spectrophotometer (Thermo Fisher Scientific, USA). DNA was stored at -20\u0026deg;C for future analysis. \u003cem\u003eE. coli\u003c/em\u003e MG1655 and its conjugative IncPα plasmid were quantified via qPCR for samples collected on days 0, 1, 2, 4, 7, 10, and 14 (Fig. S3B).\u003c/p\u003e\u003cp\u003eThe plasmid pG527, containing the target sequences, served as the qPCR standard for \u003cem\u003eE. coli\u003c/em\u003e MG1655 ΔlacZY strain and its IncPα plasmid [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. It was linearized with XbaI (Thermo Fisher Scientific, USA), purified with the QIAquick PCR purification kit (Qiagen, Germany), and verified by gel electrophoresis. qPCR assays were performed using Taqman chemistry on a LightCycler 480 II (Roche, Switzerland) in a 10 \u0026micro;L reaction volume. Details on primers, probes, and conditions are provided in Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e, with further information on the qPCR procedure and standard curve efficiency in Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. \u003cem\u003eE. coli\u003c/em\u003e invader MG1655 was quantified using qPCR targeting the ΔlacZY chromosomal gene, while pG527 was used for the plasmid [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Bacterial abundance was assessed through qPCR of the 16S rRNA gene using universal primers [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. The qPCR results were expressed in units of gene copies per glass slide surface area (cm\u003csup\u003e2\u003c/sup\u003e), and relative abundances of \u003cem\u003eΔlacZY\u003c/em\u003e and pG527 were obtained by normalizing to 16S rRNA gene copy numbers (gene copies/16S).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e2.5 Shotgun metagenomics and amplicon sequencing workflows\u003c/h2\u003e\u003cp\u003eShotgun metagenomics and 16S rRNA gene amplicon sequencing analysis were performed using Illumina platforms, NovaSeq6000 with a paired-end (2\u0026times;150bp) strategy and MiSeq (2\u0026times;300bp), respectively. For shotgun metagenomics, DNA samples from each invasion treatment replicate and one control replicate corresponding to sampling times day 0, 1, 2, 4, 10, 14 (7 was omitted) were sequenced for each sampling site (n\u0026thinsp;=\u0026thinsp;72). Library construction and sequencing was performed by IKBM facility (Kiel University, Germany), yielding a median output of 21.9 giga-bases (interquartile range of 20.8\u0026ndash;23.6).\u003c/p\u003e\u003cp\u003eShotgun metagenomic reads were cleaned by IKBM by removing adapter sequences and low-quality reads (N\u0026thinsp;\u0026gt;\u0026thinsp;10% and quality score\u0026thinsp;\u0026le;\u0026thinsp;5). The read qualities were double checked by the authors using FastQC v0.11.4 [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. For \u0026lsquo;total\u0026rsquo; (i.e., including the indigenous community resistome and that of the \u003cem\u003eE. coli\u003c/em\u003e invader) resistome analysis, DeepARG short-sequence pipeline v1.0.2 was used to annotate ARGs using the consolidated DeepARG DB v2.0 database [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. This pipeline finds and quantifies ARG-like reads from short read data and normalizes to 16S rRNA gene abundance [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. For all downstream analysis, we used this normalization (i.e., ARG copies per 16S rRNA gene) and refer to it as relative abundance [\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. We further evaluated ARG-subtypes at more granular level (e.g., from OXA to OXA-1, 2 \u0026hellip;) using a customized workflow written in Python v3.5.3, which was developed previously [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eTo separate ARGs associated with the \u003cem\u003eE. coli\u003c/em\u003e invader from those associated with the invaded community, we followed a bioinformatics workflow modified from relevant references [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. We first mapped the trimmed reads against reference sequences for \u003cem\u003eE. coli\u003c/em\u003e invader chromosome (Strain MG1655∆lacZY) and for the plasmid (GenBank: JX469830.1) using Bowtie2 v2.2.8, with \u0026lsquo;--very-sensitive\u0026rsquo; option. Using Samtools v2.6.3., the unmapped reads were separated from the mapped reads, then subjected to further analysis using DeepARG to identify \u0026lsquo;indigenous\u0026rsquo; resistomes, as outlined above (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Using the mapped reads, we further calculated coverage for each of chromosome and plasmid, followed by normalization using 16S rRNA gene, as outlined above for total resistome analysis (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eFor the comparison between biofilms grown on natural and artificial substrates, 16S rRNA gene amplicon sequencing was performed on duplicates of DNA from rocks and glass slides biofilms for each site (n\u0026thinsp;=\u0026thinsp;16). To monitor changes in the microbiome, duplicates of DNA samples from both treatment flumes obtained on day 0, 1, 2, 4, 7, 10, and 14 of the experiment were sequenced (n\u0026thinsp;=\u0026thinsp;56). Amplicon sequencing of bacterial 16S rRNA genes was performed using primers targeting the V3-V4 region (V3F: 5\u0026prime;-CCTACGGGAGGCAGCAG-3\u0026prime;; V4R: 5\u0026prime;-GGACTACHVGGGTWTCTAAT-3) [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e] to generate paired-end reads, that were collectively analysed using the DADA2 algorithm [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e] in QIIME2 [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. On average, 9288 reads per sample were obtained, distributed across 56 samples with a minimum of 1553 and a maximum of 21127 reads per sample. Forward and reverse reads were merged into amplicon-sequence variants (ASVs), classified taxonomically, and used to build an ASV table (4,090 ASVs). Indigenous microbiomes were analysed after screening out reads not mapped against 16S rRNA gene sequences for \u003cem\u003eE. coli\u003c/em\u003e invader, following the same protocol for indigenous resistome. This resulted in 9230 reads per sample on average and 4,039 ASVs.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e2.6 Statistical analysis and visualization\u003c/h2\u003e\u003cp\u003eStatistical analyses and all plots were generated in R (v4.2.0, R Core Team). To evaluate microbiome diversity (Vegan v2.6-4, \u003cem\u003evegan\u003c/em\u003e R package) at sampling sites and composition differences by substrate (natural rock vs. glass slide), we calculated alpha and beta diversity from 16S rRNA gene sequences. Shannon-index differences were tested with one-way ANOVA (sampling sites as fixed factors) followed by Tukey test. Community composition differences among microbial communities grown on natural rocks or glass slides were analysed with NMDS based on Bray-Curtis dissimilarity [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e] and group significance was assessed with a permutation test (n\u0026thinsp;=\u0026thinsp;999). \u003cem\u003eE. coli\u003c/em\u003e invader and plasmid abundance differences over time were tested with one-way ANOVA (time as a fixed factor; package: \u003cem\u003estats\u003c/em\u003e package v4.1.2 in R) followed by Tukey test.\u003c/p\u003e\u003cp\u003eResistome Shannon-index was calculated from ARG relative abundances, following the same method as microbiome analysis, using \u003cem\u003eVegan v2.6-4\u003c/em\u003e (\u003cem\u003evegan\u003c/em\u003e package). Metagenome beta diversity was analysed by Bray-Curtis dissimilarity and visualized with NMDS. Group comparison by treatment used ANOSIM (\u003cem\u003evegan\u003c/em\u003e), and ARG subtype correlation was visualized with biplot analysis in \u003cem\u003evegan\u003c/em\u003e (p\u0026thinsp;\u0026le;\u0026thinsp;0.05). The top 20 significantly correlated ARGs associated with the sample of interest were further identified, based upon the residual sum of squares, representing the deviation between regression lines and the sample (i.e., visualized as vectors on the ordination plot) [\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. The top 20 genes were further visualized in heatmaps using gplots v3.1.3. Procrustes analysis was conducted to correlate resistome and microbiome ordination spaces using \u003cem\u003evegan\u003c/em\u003e, visualizing deviations as blue vector arrows, with significance assessed by a permutation test (n\u0026thinsp;=\u0026thinsp;999). Spearman correlations between qPCR and metagenomics data (in copies per 16S rRNA gene) for \u003cem\u003eE. coli\u003c/em\u003e and the \u003cem\u003epG527\u003c/em\u003e plasmid were tested using \u003cem\u003estats\u003c/em\u003e package v4.1.2 in R, excluding an outlier (the 2nd replicate of D14-\u003cem\u003edwA\u003c/em\u003e). Trendlines with confidence interval were calculated using the \u003cem\u003escales\u003c/em\u003e package (method\u0026thinsp;=\u0026thinsp;linear model) and plotted in R.\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Characterization of biofilm microbiome diversity\u003c/h2\u003e\u003cp\u003eShannon diversity varied by sampling location, regardless of samples originating from natural rocks [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. or grown on glass slides (n\u0026thinsp;=\u0026thinsp;16; Fig. S4). Both substrate types exhibited similar diversity patterns, with upstream locations being less diverse than downstream ones. The difference between the two downstream sites was not statistically significant (Fig. S4).\u003c/p\u003e\u003cp\u003eBeta diversity analysis revealed significant differences between microbial communities growing on natural rocks compared to glass slides (PERMANOVA, p\u0026thinsp;=\u0026thinsp;0.004; Fig. S5A). \u003cem\u003eCyanobacteria\u003c/em\u003e and \u003cem\u003eBacillota\u003c/em\u003e were notably more abundant in natural rock biofilms (Fig. S5B). The number of taxa (ASVs) on glass slides was not significantly different from that on natural rocks (Kruskal-Wallis pairwise, p\u0026thinsp;=\u0026thinsp;0.174), with both types predominantly composed of \u003cem\u003ePseudomonadota\u003c/em\u003e, \u003cem\u003eBacteroidota\u003c/em\u003e, and \u003cem\u003eActinomycetota\u003c/em\u003e (Fig. S5B).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Tracking ARB-\u003cem\u003eE. coli\u003c/em\u003e invader and potential HGT transfer\u003c/h2\u003e\u003cp\u003eThe abundance of the ARB-\u003cem\u003eE. coli\u003c/em\u003e invader and its plasmid within indigenous biofilm communities were tracked independently with specific qPCR assays. The abundance of the invader and its plasmid varied among replicates and decreased over time regardless of biofilm origin (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, B). The chromosomal ΔlacZY target indicating \u003cem\u003eE. coli\u003c/em\u003e abundance peaked 24 hours post-invasion, then declined but remained detectable until the end of sampling (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). The resistance plasmid abundance was consistently\u0026thinsp;~\u0026thinsp;2-log higher than that of the invader (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB), but abundance patterns were overall consistent with \u003cem\u003eE. coli\u003c/em\u003e. The plasmid and strain abundances showed considerable variability among replicates of the same origin (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA, B). We therefore analysed replicates individually. However, all sites and replicates showed the same transitory invasion pattern and the negative slope of the trendline of the plasmid-to-strain suggests that plasmid abundance was decreasing faster than \u003cem\u003eE. coli\u003c/em\u003e chromosomal DNA, indicating a low potential conjugation rate and/or loss of the plasmid from the donor strain (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC).\u003c/p\u003e\u003cp\u003eThe ratio decreased significantly from the beginning to the end of the invasion in most of the invaded biofilms (ANOVA, p\u0026thinsp;=\u0026thinsp;0.003 \u003cem\u003eupL\u003c/em\u003e; p\u0026thinsp;=\u0026thinsp;0.038 \u003cem\u003eupH\u003c/em\u003e; p-value\u0026thinsp;=\u0026thinsp;1.88e-05 \u003cem\u003edwB\u003c/em\u003e; Fig. S6). However, the rate of the decrease varied. For instance, it was more stable in \u003cem\u003eupH\u003c/em\u003e than in \u003cem\u003eupL\u003c/em\u003e from day 7 to the end of the experiment. In the downstream \u003cem\u003edwA\u003c/em\u003e location, situated after the WW effluent, the ratio did not decrease significantly (ANOVA p-value\u0026thinsp;=\u0026thinsp;0.0891 \u003cem\u003edwA\u003c/em\u003e) indicating a higher stability of the plasmid and/or possible horizontal gene transfer (Fig. S6). At both downstream locations, ratios were initially lower than upstream but then decreased only slightly and were higher than in the upstream by the end of the experiment. Metagenomics supported the qPCR results: we obtained significant correlations between the \u003cem\u003eE. coli\u003c/em\u003e and plasmid relative abundance measured by qPCR and calculated from metagenome data (p\u0026thinsp;\u0026le;\u0026thinsp;0.008, Fig. S7).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eIn addition, we quantified \u003cem\u003eE. coli\u003c/em\u003e using both metagenomic read mapping and 16S rRNA amplicon sequencing. Estimates obtained from metagenomic data were highly consistent with those derived from 16S-based \u003cem\u003eE. coli\u003c/em\u003e recounts, confirming that both approaches captured similar temporal patterns of invasion across treatments and biofilm origins (Fig. S8). The 16S- and metagenomic derived \u003cem\u003eE. coli\u003c/em\u003e recounts reflected the same dynamics observed in the relative abundances (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA) with a sharp increase in \u003cem\u003eE. coli\u003c/em\u003e on days 1\u0026ndash;2 followed by a rapid decline to background levels by day 4\u0026ndash;7. This agreement across independent molecular methods strengthens the robustness of our invasion assessments and supports the reliability of the observed differences between biofilm origins.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Responses of indigenous resistomes to ARB-\u003cem\u003eE. coli\u003c/em\u003e\u003c/h2\u003e\u003cp\u003eWe applied a custom bioinformatics workflow that allowed to determine the contribution of the invading \u003cem\u003eE. coli\u003c/em\u003e to the biofilm metagenomes and to extract information about the indigenous metagenome and resistome. We found that the indigenous resistomes were highly diverse, and contained some ARGs of potential public health concern [\u003cspan additionalcitationids=\"CR32 CR33\" citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e], such as genes for class-A (NMC), class-B (GIM, and IMP), and class-D (OXA) carbapenemases, and the \u003cem\u003eermB\u003c/em\u003e gene for macrolide-lincosamide-streptogramin (MLS) resistance (Fig. S9). Among them, some OXA genes (e.g, OXA-3, 5, 53, -119, -129, -198, and \u0026minus;\u0026thinsp;205) and \u003cem\u003eermB\u003c/em\u003e occurred abundantly across samples but their relative abundances did not exhibit substantial temporal dynamics (Fig. S9).\u003c/p\u003e\u003cp\u003eAlpha-diversity of the indigenous resistomes (i.e., ARG subtypes identified in the metagenome), and total resistome (i.e. including both ARB-\u003cem\u003eE. coli\u003c/em\u003e and indigenous ARGs) was quantified using the Shannon-index. The alpha diversity of the total resistomes increased markedly shortly after invader addition (Day 0 to 2) in the invaded biofilms (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Both replicates (Rep1 and 2) exhibited generally similar temporal dynamics. However, only three of four downstream biofilms (\u003cem\u003edwA\u003c/em\u003e Rep 1, and \u003cem\u003edwB\u003c/em\u003e Rep 1 and 2) showed still elevated resistome diversity on day 4 compared to upstream sites where it had already dropped to values only slightly above day 0 values. In contrast, only minor changes were observed in the control biofilms. The dynamics thus generally mirrored the transient ARB-\u003cem\u003eE. coli\u003c/em\u003e invasion (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). In the indigenous resistome, alpha diversity was less dynamic, only replicate 1 of \u003cem\u003edwA\u003c/em\u003e and replicate 2 of \u003cem\u003edwB\u003c/em\u003e appear to show a small transient response (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Some changes and a divergence of samples was observed at later timepoints (day 1 and 14) in both invaded and control biofilms.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eSimilar to alpha-diversity, beta-diversity analysis also showed very different patterns for the total resistomes of invaded biofilms in comparison to indigenous resistomes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA and B). For the total ARGs, clear temporal changes connected to the transitory invasion are apparent while differences between biofilm origins were not significant (i.e., sampling sites; ANOSIM r\u0026thinsp;=\u0026thinsp;0.032, p\u0026thinsp;=\u0026thinsp;0.141; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). Specifically, the total resistomes shifted a few days after the invasion (Day1\u0026thinsp;~\u0026thinsp;2) along the NMDS1 axis, indicating the ARB-\u003cem\u003eE. coli\u003c/em\u003e impact on biofilm resistomes during the invasion events is captured on this axis. Total resistomes then returned to the pre-disturbance configuration (i.e., \u003cem\u003eupH, dwB\u003c/em\u003e, and \u003cem\u003edwA\u003c/em\u003e in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA; \u003cem\u003eupL, upH\u003c/em\u003e, and \u003cem\u003edwA\u003c/em\u003e in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). Controls did not exhibit pronounced temporal changes along the ordination axis (Fig. S10).\u003c/p\u003e\u003cp\u003eThe indigenous resistomes significantly differed by biofilm origin (ANOSIM r\u0026thinsp;=\u0026thinsp;0.79, p\u0026thinsp;=\u0026thinsp;0.001, Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB) but not over time. Some outliers, such as \u003cem\u003edwA\u003c/em\u003e-Rep2 D1, \u003cem\u003edwB\u003c/em\u003e-Rep1 D2, and \u003cem\u003eupL\u003c/em\u003e-Rep2 D14 were noted (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB and S10). Similarly, controls mostly did not change much over the course of monitoring, forming tight clusters by biofilm origin, with one exception, \u003cem\u003eupH\u003c/em\u003e-D14 (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB and S10).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec13\" class=\"Section2\"\u003e\u003ch2\u003e3.4 Correlation between resistome and microbiome\u003c/h2\u003e\u003cp\u003eBetween-group comparison using ANOSIM revealed that dissimilarity of total microbiomes significantly differed by sampling locations (r\u0026thinsp;=\u0026thinsp;0.79, p\u0026thinsp;=\u0026thinsp;0.001), not exhibiting pronounced temporal variation (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). This indicates that biofilm origin rather than transient ARB-\u003cem\u003eE. coli\u003c/em\u003e invasion mainly determined microbiome composition throughout the experiment. Results for indigenous microbiomes mirrored those for total microbiome, indicating that the presence of the single invader species had only a minor impact on the overall community structure (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB).\u003c/p\u003e\u003cp\u003eProcrustes analysis revealed significant association between indigenous resistome and indigenous microbiome at a community level at 5% significance level (p\u0026thinsp;=\u0026thinsp;0.025) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC), indicating that resistomes and microbiomes were structurally correlated, suggesting that differences by origin (i.e., sites) in microbial community composition might explain those observed in the indigenous resistome.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003e4.1 Contrasting microbiome diversity characterized the study sites\u003c/h2\u003e\u003cp\u003eEssential for the intended experiments was to obtain riverine biofilms on substrates suitable for experimentation yet representing natural contrasts in diversity and community composition. Our experimental procedures were in line with the standardization proposed by Freixa and colleagues [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e] for the use of artificial substrates in freshwater biofilms studies, and the experimental approach using exposure units to grow biofilm in-situ (Fig. \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e) proved viable and achieved the desired result. The comparison between biofilms grown on natural substrates (i.e., river rocks) and artificial substrates (i.e., glass-slides deployed in the river) ensured that the natural contrasts of microbial diversity and composition of the fresh waters biofilms were retained when using the artificial substrates (glass slides) for experimental purposes [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. Differences between rocks and glass-slides biofilm diversity indices and community compositions were likely due to differences in light exposition, nutrient status and water flow they might be exposed to during the colonization period [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e, \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e], and the comparatively young age of glass-slide biofilms.\u003c/p\u003e\u003cp\u003e\u003cb\u003e4.2 The presence of the invading ARB-\u003c/b\u003e\u003cb\u003eE. coli\u003c/b\u003e \u003cb\u003ewas short-lived, and conjugative transfer of its plasmid was limited\u003c/b\u003e\u003c/p\u003e\u003cp\u003eThe fate of the invading ARB-\u003cem\u003eE. coli\u003c/em\u003e was overall independent of the origin and composition of the invaded biofilm. Similarly to what was observed in our previous study using the same \u003cem\u003eE. coli\u003c/em\u003e strain [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e], invasion was only transitory, referring to only low levels of the invader DNA remaining after 14 days. However, the degree of invasion during the first 2 days, indicated by the abundances of ARB-\u003cem\u003eE. coli\u003c/em\u003e in the biofilm communities, was considerable and differed by the origin of biofilm. Although invaders did not establish long-term dominance, this could still be relevant under chronic exposure scenarios, where a continuous external source may sustain their presence in the biofilm despite their limited ability to naturally establish. For successful invasion, the invader must persist and be able to exploit the resources of the given habitat, overcoming the indigenous community [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Theoretically, healthy environments with high microbial diversity make the invasion process more difficult as a community with great diversity enhances its capacity to exploit the habitat resources [\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Consistent with this hypothesis, the upstream low diversity biofilm (\u003cem\u003eupL\u003c/em\u003e) allowed larger abundances of ARB-\u003cem\u003eE. coli\u003c/em\u003e invader to establish than those of the upstream high diversity biofilm, potentially indicating a greater availability of ecological niches and less intraspecific competition among the low diversity community.\u003c/p\u003e\u003cp\u003eConsidering that biofilms provide conditions facilitating plasmid transfer through conjugation (i.e., close contact between cells as well as minimal shear forces), it has been assumed that conjugative gene transfer in microbial biofilms occurs with high efficiency [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e]. As we observed overall decreasing plasmid-to-host ratios, the evidence suggests absence or a limited impact of horizontal gene transfer (HGT) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC and Fig. S6), while indicating that the remaining ARB-\u003cem\u003eE. coli\u003c/em\u003e populations carries increasingly lower numbers of plasmids over time. Plasmid loss is a phenomenon known to occur in non-selective environments, but generally as an evolutionary process during population growth [\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e] and often persistence is higher than expected [\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e]. Thus, the mechanism of the observed excess plasmid loss remains unclear. Biofilm communities located upstream, far from known pollution sources showed a more rapid loss of plasmid in comparison to the downstream biofilms (Fig. S6). This slower plasmid loss downstream may reflect selection pressures favoring plasmid retention in polluted environments, although contributions from conjugative transfer cannot be excluded.\u003c/p\u003e\u003cp\u003eHowever, these results do not provide sufficient evidence that biofilm communities grown in the upstream \u0026ldquo;non-impacted\u0026rdquo; locations differ from the more impacted downstream locations in their natural resilience against the spread of antibiotic resistance, since an increase in the plasmid-to-host ratio was never observed.\u003c/p\u003e\u003cp\u003eA limitation of this study is that only one type of Inc plasmid carrying resistance genes has been tested. The fate of AMR plasmids is the result of the complex interactions between the plasmid, its hosts and the environment [\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e]. For instance, abiotic factors that directly select for plasmid-encoded traits, such as antibiotics or heavy metals, can inhibit the growth of plasmid-free cells. Other abiotic factors such as pH, temperature or nutrients, may directly affect some of the mechanisms that determine whether a plasmid will persist in a population: plasmid fitness cost, replication and segregation rate, and horizontal transfer rate [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e, \u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. Therefore, the \u0026lsquo;invasibility\u0026rsquo; of the biofilm by plasmids, or the proportion of biofilm cells that acquired plasmids within a few hours, could depend critically on the type of plasmid, on the time of biofilm exposure to the donor, on the biofilm age and on the ability of the plasmid donor to attach to the biofilm [\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e].\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003e4.3 Does biofilm diversity and composition matter for the fate of AMR?\u003c/h2\u003e\u003cp\u003eThe presence of the invader profoundly impacted the resistomes diversity and structure, but this was primarily traced to the direct presence of ARB-\u003cem\u003eE. coli\u003c/em\u003e itself using our method of observing the indigenous resistome. Nevertheless, this finding is remarkable in showing that an invasion that did not affect the overall microbiome composition resulted in pronounced changes to the resistome.\u003c/p\u003e\u003cp\u003eWe further reported on changes in the indigenous community resistomes that were not associated with the invasive ARB-\u003cem\u003eE. coli\u003c/em\u003e. As introduced earlier, both HGT and shifts of community are known drivers for changes in resistome [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. The extent of HGT does not appear to have been extensive enough to affect overall indigenous resistome structure. This is in line with the results of qPCR-based analysis (3.2, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e, 4.2). The shifts in microbiome composition over the course of monitoring were likewise limited, as evidenced by the microbial community structure clustering mainly by biofilm origin rather than time (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA, B). The significant structural correlation between microbiomes and resistomes therefore mostly confirms that differences in community composition between biofilm origins influenced the resistomes. The early fluctuations observed in the indigenous resistome (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB), although minor compared to the temporal trends observed in the total resistome, raise the question of whether they reflect true biological responses or technical noise introduced during indigenous metagenome filtering and normalization. By manually crosschecking our data, we did not observe evidence that the bioinformatic filtering workflow introduced artifacts that would be problematic, i.e. incomplete filtering of resistance genes of \u003cem\u003eE. coli\u003c/em\u003e MG1655. The consistency across replicates also suggests that part of this variation may represent genuine transient community adjustments following the initial disturbance from exposure to the invading organism.\u003c/p\u003e\u003cp\u003eFrom day 7 onward, the plasmid-to-strain ratio remained markedly more stable in upH than in upL, indicating a greater persistence of the plasmid in the upH community (Fig. S6). Likewise, in the downstream dwA biofilms, located immediately after the WWTP effluent, the ratio did not significantly decrease (Fig. S6), supporting a higher persistence of the introduced plasmid and/or increased HGT potential in this environment. This stability, despite the decline in \u003cem\u003eE. coli\u003c/em\u003e abundance, aligns with the notion that downstream environments receiving continuous anthropogenic inputs may sustain persistent alterations of their resistomes even when invaders fail to establish long-term. In both locations, the higher diversity site appeared to support a slightly higher persistence of the plasmid over the duration of the experiment. This was not expected according to our initial hypothesis, and we can currently not speculate regarding the cause.\u003c/p\u003e\u003cp\u003eThe magnitude of community shift and HGT, has been known to depend strongly upon biotic interactions (i.e., composition of introduced and indigenous communities) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Our study in this respect confirms that the composition of the indigenous community is an important factor - anthropogenically impacted indigenous biofilm communities (i.e., \u003cem\u003edwA\u003c/em\u003e) exhibited the lowest ecological resilience in response to ARB-invasion, at least in one of the repeated cases. WW-impacted biofilms evolve in highly specialized environments, where a chronic perturbation (i.e., wastewater effluents) exists. The effluents contain high biomass with a stable microbial and resistome composition that is distinct from undisturbed waters [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e]. Thus, the locations receiving high volumes of wastewater effluents can be expected to be more favourable environments for colonization by organisms with a wastewater origin [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. Exposed to chronic perturbation or constant invasion pressure, communities may reach alternative stable states adapted to function in the new environment [\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e, \u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e]. Removal of chronic perturbation does not necessarily lead to the reversal of community assembly because resilience can be reduced. In this sense, \u003cem\u003edwA\u003c/em\u003e might have exhibited reduced resilience under the absence of a chronic perturbation in the case of our experiment setting (i.e., using filtered river waters, where biomass particles were removed), allowing certain organisms to colonize successfully in the later stage of ecological succession.\u003c/p\u003e\u003cp\u003eThis mainly shows that perturbation can have rather unpredictable effects on the resistome of a microbial community. The response of the indigenous resistome was also stochastic - only one of the replicates exhibited significant alteration in the later stage of ecological succession (Day14). As discussed earlier, the composition of microbial community itself and the perturbation, can be important factors; the composition of the microbiome in the two replicates for \u003cem\u003edwA\u003c/em\u003e may not have been identical, leading to varying responses to the invader between the replicates. These observations warrant further study, including on biofilms affected by multiple sources of contamination and multiple ARB-invasions.\u003c/p\u003e\u003c/div\u003e"},{"header":"Conclusions","content":"\u003cp\u003eFrom a One-Health perspective, biofilms are considered potential hotspots where transferrable AMR could evolve and thrive. In this study, we reported that the indigenous freshwater biofilm microbiomes were largely resilient and did not respond to the invasion of an ARB-\u003cem\u003eE. coli\u003c/em\u003e whereas resistomes were temporally affected. The limited colonization success of the invader and limited extent of indirect impacts of the invasion (i.e., HGT and changes in indigenous resistomes) within freshwater biofilms, indicates that the potential health risks posed by freshwater biofilms, might not be as high as we initially hypothesized. However, specific groups of indigenous ARGs, including those of public health concern \u0026lsquo;can\u0026rsquo; still increase in biofilms from wastewater-impacted sites. This suggests that the potential public health risks can persist locally, particularly in highly wastewater-impacted freshwater environments. Therefore, future studies involving additional wastewater-impacted sites are required. Once confirmed, these results could motivate measures to maintain highly resilient microbial biofilms in rivers as one way to reduce the spread of resistances in natural environments.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eDeclaration of competing interest\u003c/h2\u003e\u003cp\u003eAll authors declare no financial or non-financial competing interests.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis study was supported by the ANTIVERSA project (BiodivERsA, European Union) and the Swiss National Science Foundation (Switzerland) grant 186531. GG was supported by a Junior Leader Incoming Fellowship contract (LCF/BQ/PI23/11970040) funded by La Caixa Foundation (ID 100010434).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eGG : Conceptualization, Investigation, Formal analysis, Visualization, Writing - Original Draft, Review and Editing; JL : Investigation, Formal analysis, Visualization, Writing - Original Draft, Review and Editing; OH : Investigation, Review and Editing; KB : Investigation, Review and Editing; HB : Conceptualization, Funding acquisition, Supervision, Writing - Review and Editing.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eWe acknowledge our ANTIVERSA colleagues for the joint development of the experimental design used in this study and specifically X. Bellanger and C. Merlin from LCPME (Nancy, France) for providing E. coli MG1655 (CM2372) ∆lacZY and pG527 IncPα plasmid, used to study microbial invasion and for providing the standards for their quantification through qPCR. We thank Eawag workshop team for their help building the exposure units and the flumes, as well as with the experimental setup. Finally, we thank the Oberglatt Fishery Department for their support and permission to access to the Glatt River for colonizing biofilm on the glass slides and sampling.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eAll the sequences were deposited in the NCBI (National Center for Biotechnology Information) under SRA accession (sequence read archive) number: PRJNA1070575 (Microbiome data) and PRJNA1129054 (Metagenomics data). All data are presented in the manuscript and supplementary material. Raw data that support the findings of this study are available from the corresponding author upon request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eHernando-Amado S, Coque TM, Baquero F, Mart\u0026iacute;nez JL. Defining and combating antibiotic resistance from One Health and Global Health perspectives. \u003cem\u003eNat Microbiol\u003c/em\u003e 2019; 4: 1432\u0026ndash;1442.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSmalla K, Cook K, Djordjevic SP, Kl\u0026uuml;mper U, Gillings M. Environmental dimensions of antibiotic resistance: assessment of basic science gaps. \u003cem\u003eFEMS Microbiology Ecology\u003c/em\u003e 2018; 94.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWoods LC, Gorrell RJ, Taylor F, Connallon T, Kwok T, McDonald MJ. Horizontal gene transfer potentiates adaptation by reducing selective constraints on the spread of genetic variation. \u003cem\u003eProc Natl Acad Sci USA\u003c/em\u003e 2020; 117: 26868\u0026ndash;26875.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMatviichuk O, Mondamert L, Geffroy C, Gaschet M, Dagot C, Labanowski J. River Biofilms Microbiome and Resistome Responses to Wastewater Treatment Plant Effluents Containing Antibiotics. \u003cem\u003eFront Microbiol\u003c/em\u003e 2022; 13: 795206.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eReichert G, Hilgert S, Alexander J, Rodrigues De Azevedo JC, Morck T, Fuchs S, et al. Determination of antibiotic resistance genes in a WWTP-impacted river in surface water, sediment, and biofilm: Influence of seasonality and water quality. \u003cem\u003eScience of The Total Environment\u003c/em\u003e 2021; 768: 144526.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBagra K, Bellanger X, Merlin C, Singh G, Berendonk TU, Kl\u0026uuml;mper U. Environmental stress increases the invasion success of antimicrobial resistant bacteria in river microbial communities. \u003cem\u003eScience of The Total Environment\u003c/em\u003e 2023; 904: 166661.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJu F, Beck K, Yin X, Maccagnan A, McArdell CS, Singer HP, et al. Wastewater treatment plant resistomes are shaped by bacterial composition, genetic exchange, and upregulated expression in the effluent microbiomes. \u003cem\u003eThe ISME Journal\u003c/em\u003e 2019; 13: 346\u0026ndash;360.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee J, Ju F, Maile-Moskowitz A, Beck K, Maccagnan A, McArdell CS, et al. Unraveling the riverine antibiotic resistome: The downstream fate of anthropogenic inputs. \u003cem\u003eWater Research\u003c/em\u003e 2021; 197: 117050.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLuo G, Li B, Li L-G, Zhang T, Angelidaki I. Antibiotic Resistance Genes and Correlations with Microbial Community and Metal Resistance Genes in Full-Scale Biogas Reactors As Revealed by Metagenomic Analysis. \u003cem\u003eEnviron Sci Technol\u003c/em\u003e 2017; 51: 4069\u0026ndash;4080.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eJones ML, Rivett DW, Pascual-Garc\u0026iacute;a A, Bell T. Relationships between community composition, productivity and invasion resistance in semi-natural bacterial microcosms. \u003cem\u003eeLife\u003c/em\u003e 2021; 10: e71811.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMawarda PC, Lakke SL, Dirk Van Elsas J, Salles JF. Temporal dynamics of the soil bacterial community following Bacillus invasion. \u003cem\u003eiScience\u003c/em\u003e 2022; 25: 104185.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAlbright MBN, Sevanto S, Gallegos-Graves LV, Dunbar J. Biotic Interactions Are More Important than Propagule Pressure in Microbial Community Invasions. \u003cem\u003emBio\u003c/em\u003e 2020; 11: e02089-20.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMerlin C, McAteer S, Masters M. Tools for Characterization of \u003cem\u003eEscherichia coli\u003c/em\u003e Genes of Unknown Function. \u003cem\u003eJ Bacteriol\u003c/em\u003e 2002; 184: 4573\u0026ndash;4581.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGionchetta G, Snead D, Semerad S, Beck K, Pruden A, B\u0026uuml;rgmann H. Dynamics of antibiotic resistance markers and \u003cem\u003eEscherichia coli\u003c/em\u003e invasion in riverine heterotrophic biofilms facing increasing heat and flow stagnation. \u003cem\u003eScience of The Total Environment\u003c/em\u003e 2023; 893: 164658.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKl\u0026uuml;mper U, Gionchetta G, Cat\u0026atilde;o E, Bellanger X, Dielacher I, Elena AX, et al. Environmental microbiome diversity and stability is a barrier to antimicrobial resistance gene accumulation. \u003cem\u003eCommun Biol\u003c/em\u003e 2024; 7: 706.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSagrillo C, Changey F, Bellanger X. Bacteriophages vehiculate a high amount of antibiotic resistance determinants of bacterial origin in the Orne River ecosystem. \u003cem\u003eEnvironmental Microbiology\u003c/em\u003e 2022; 24: 4317\u0026ndash;4328.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHeuer H, Smalla K. Manure and sulfadiazine synergistically increased bacterial antibiotic resistance in soil over at least two months. \u003cem\u003eEnvironmental Microbiology\u003c/em\u003e 2007; 9: 657\u0026ndash;666.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBustin SA, Benes V, Garson JA, Hellemans J, Huggett J, Kubista M, et al. The MIQE Guidelines: Minimum Information for Publication of Quantitative Real-Time PCR Experiments. \u003cem\u003eClinical Chemistry\u003c/em\u003e 2009; 55: 611\u0026ndash;622.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTakai K, Horikoshi K. Rapid Detection and Quantification of Members of the Archaeal Community by Quantitative PCR Using Fluorogenic Probes. \u003cem\u003eAppl Environ Microbiol\u003c/em\u003e 2000; 66: 5066\u0026ndash;5072.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAndrews, Simon. FastQC: a quality control tool for high throughput sequence data. \u003cem\u003eBabraham Bioinformatics, Babraham Institute, Cambridge, United Kingdom\u003c/em\u003e.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eArango-Argoty G, Garner E, Pruden A, Heath LS, Vikesland P, Zhang L. DeepARG: a deep learning approach for predicting antibiotic resistance genes from metagenomic data. \u003cem\u003eMicrobiome\u003c/em\u003e 2018; 6: 23.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYang Y, Jiang X, Chai B, Ma L, Li B, Zhang A, et al. ARGs-OAP: online analysis pipeline for antibiotic resistance genes detection from metagenomic data using an integrated structured ARG-database. \u003cem\u003eBioinformatics\u003c/em\u003e 2016; 32: 2346\u0026ndash;2351.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee, J., Xiang, K., Au, E. et al. Longitudinal monitoring of sewershed resistomes in socioeconomically diverse urban neighborhoods. \u003cem\u003eCommun Med\u003c/em\u003e 5, 7 (2025).\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTarracchini C, Alessandri G, Fontana F, Rizzo SM, Lugli GA, Bianchi MG, et al. Genetic strategies for sex-biased persistence of gut microbes across human life. \u003cem\u003eNat Commun\u003c/em\u003e 2023; 14: 4220.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu Y, Elworth RAL, Jochum MD, Aagaard KM, Treangen TJ. De novo identification of microbial contaminants in low microbial biomass microbiomes with Squeegee. \u003cem\u003eNat Commun\u003c/em\u003e 2022; 13: 6799.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCaporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Lozupone CA, Turnbaugh PJ, et al. Global patterns of 16S rRNA diversity at a depth of millions of sequences per sample. \u003cem\u003eProc Natl Acad Sci USA\u003c/em\u003e 2011; 108: 4516\u0026ndash;4522.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCallahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJA, Holmes SP. DADA2: High-resolution sample inference from Illumina amplicon data. \u003cem\u003eNat Methods\u003c/em\u003e 2016; 13: 581\u0026ndash;583.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBolyen E, Rideout JR, Dillon MR, Bokulich NA, Abnet CC, Al-Ghalith GA, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. \u003cem\u003eNat Biotechnol\u003c/em\u003e 2019; 37: 852\u0026ndash;857.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLegendre L, Legendre P. Numerical ecology. 2012. \u003cem\u003eElsevier\u003c/em\u003e, Amsterdam.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eParkins M, Lee J, Xiang K, Au E, Sarabi S, Acosta N, et al. Longitudinal monitoring of sewershed resistomes in socioeconomically diverse urban neighbourhoods. 2024.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eQueenan AM, Bush K. Carbapenemases: the Versatile β-Lactamases. \u003cem\u003eClin Microbiol Rev\u003c/em\u003e 2007; 20: 440\u0026ndash;458.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRossolini GM, D\u0026rsquo;Andrea MM, Mugnaioli C. The spread of CTX-M-type extended-spectrum β-lactamases. \u003cem\u003eClinical Microbiology and Infection\u003c/em\u003e 2008; 14: 33\u0026ndash;41.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEvans BA, Amyes SGB. OXA β-Lactamases. \u003cem\u003eClin Microbiol Rev\u003c/em\u003e 2014; 27: 241\u0026ndash;263.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMugnier P, Dubrous P, Casin I, Arlet G, Collatz E. A TEM-derived extended-spectrum beta-lactamase in \u003cem\u003ePseudomonas aeruginosa\u003c/em\u003e. \u003cem\u003eAntimicrob Agents Chemother\u003c/em\u003e 1996; 40: 2488\u0026ndash;2493.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSabater S, Guasch H, Ricart M, Roman\u0026iacute; A, Vidal G, Kl\u0026uuml;nder C, et al. Monitoring the effect of chemicals on biological communities. The biofilm as an interface. \u003cem\u003eAnal Bioanal Chem\u003c/em\u003e 2007; 387: 1425\u0026ndash;1434.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFreixa A, Ortiz-Rivero J, Sabater S. Artificial substrata to assess ecological and ecotoxicological responses in river biofilms: Use and recommendations. \u003cem\u003eMethodsX\u003c/em\u003e 2023; 10: 102089.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBiggs BJF. Artificial substrate exposure times for periphyton biomass estimates in rivers. \u003cem\u003eNew Zealand Journal of Marine and Freshwater Research\u003c/em\u003e 1988; 22: 507\u0026ndash;515.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKinnunen M, Dechesne A, Proctor C, Hammes F, Johnson D, Quintela-Baluja M, et al. A conceptual framework for invasion in microbial communities. \u003cem\u003eThe ISME Journal\u003c/em\u003e 2016; 10: 2773\u0026ndash;2779.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHector A, Dobson K, Minns A, Bazeley-White E, Hartley Lawton J. Community diversity and invasion resistance: An experimental test in a grassland ecosystem and a review of comparable studies. \u003cem\u003eEcological Research\u003c/em\u003e 2001; 16: 819\u0026ndash;831.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMallon CA, Le Roux X, Van Doorn GS, Dini-Andreote F, Poly F, Salles JF. The impact of failure: unsuccessful bacterial invasions steer the soil microbial community away from the invader\u0026rsquo;s niche. \u003cem\u003eThe ISME Journal\u003c/em\u003e 2018; 12: 728\u0026ndash;741.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMolin S, Tolker-Nielsen T. Gene transfer occurs with enhanced efficiency in biofilms and induces enhanced stabilisation of the biofilm structure. \u003cem\u003eCurrent Opinion in Biotechnology\u003c/em\u003e 2003; 14: 255\u0026ndash;261.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKr\u0026oacute;l JE, Wojtowicz AJ, Rogers LM, Heuer H, Smalla K, Krone SM, et al. Invasion of \u003cem\u003eE. coli\u003c/em\u003e biofilms by antibiotic resistance plasmids. \u003cem\u003ePlasmid\u003c/em\u003e 2013; 70: 110\u0026ndash;119.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSmith MA, Bidochka MJ. Bacterial fitness and plasmid loss: the importance of culture conditions and plasmid size. \u003cem\u003eCan J Microbiol\u003c/em\u003e 1998; 44: 351\u0026ndash;355.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBellanger, X., Guilloteau, H., Bonot, S. \u0026amp; Merlin, C. Demonstrating plasmid-based horizontal gene transfer in complex environmental matrices: A practical approach for a critical review. \u003cem\u003eSci. Total Environ.\u003c/em\u003e 2014; 493, 872\u0026ndash;882.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCarroll AC, Wong A. Plasmid persistence: costs, benefits, and the plasmid paradox. \u003cem\u003eCan J Microbiol\u003c/em\u003e 2018; 64: 293\u0026ndash;304.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCasta\u0026ntilde;eda-Barba S, Top EM, Stalder T. Plasmids, a molecular cornerstone of antimicrobial resistance in the One Health era. \u003cem\u003eNat Rev Microbiol\u003c/em\u003e 2024; 22: 18\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSmalla K, Jechalke S, Top EM. Plasmid Detection, Characterization, and Ecology. In: Tolmasky ME, Alonso JC (eds). \u003cem\u003ePlasmids\u003c/em\u003e. 2015. ASM Press, Washington, DC, USA, pp 445\u0026ndash;458.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMansfeldt C, Deiner K, M\u0026auml;chler E, Fenner K, Eggen RIL, Stamm C, et al. Microbial community shifts in streams receiving treated wastewater effluent. \u003cem\u003eScience of The Total Environment\u003c/em\u003e 2020; 709: 135727.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang Z, Zhang G, Ju F. Using Culture-Enriched Phenotypic Metagenomics for Targeted High-Throughput Monitoring of the Clinically Important Fraction of the β-Lactam Resistome. \u003cem\u003eEnviron Sci Technol\u003c/em\u003e 2022; 56: 11429\u0026ndash;11439.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eShade A, Peter H, Allison SD, Baho DL, Berga M, B\u0026uuml;rgmann H, et al. Fundamentals of Microbial Community Resistance and Resilience. \u003cem\u003eFront Microbio\u003c/em\u003e 2012; 3.\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePotts LD, Douglas A, Perez Calderon LJ, Anderson JA, Witte U, Prosser JI, et al. Chronic Environmental Perturbation Influences Microbial Community Assembly Patterns. \u003cem\u003eEnviron Sci Technol\u003c/em\u003e 2022; 56: 2300\u0026ndash;2311.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-antimicrobials-and-resistance","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjamar","sideBox":"Learn more about [npj Antimicrobials and Resistance](http://www.nature.com/npjamar/)","snPcode":"44259","submissionUrl":"https://submission.springernature.com/new-submission/44259/3","title":"npj Antimicrobials and Resistance","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Antibiotic resistance, antibiotic resistance genes, disturbance, mobile genetic elements, diversity, pollution","lastPublishedDoi":"10.21203/rs.3.rs-8270852/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8270852/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eRiver biofilms are exposed to invasion by antibiotic resistant bacteria (ARB) due to episodic or chronic exposure to wastewater, yet the ecological processes determining the fate of invaders and their resistance plasmids remain poorly understood. We experimentally exposed river-grown biofilms, originating from sites with contrasting microbial diversity and wastewater influence, to invasion by a genetically tagged ARB-\u003cem\u003eE. coli\u003c/em\u003e carrying a transferable IncPα plasmid with the \u003cem\u003enptII\u003c/em\u003e resistance gene. Over two weeks, we quantified the dynamics of the invader and its plasmid, using qPCR and the plasmid-to-strain genome ratio in the biofilm as an indicator of horizontal gene transfer (HGT). We further characterized microbiomes and resistomes via 16S rRNA gene sequencing and metagenomics. Independent quantification methods provided highly consistent estimates of invasion dynamics: the invader established transiently across all biofilms, with ARB-\u003cem\u003eE. coli\u003c/em\u003e abundance peaking within 48h and subsequently declining to near-background levels within 14 days. Plasmid-to-strain genome ratios decreased, indicating limited HGT and progressive plasmid loss. Wastewater-impacted biofilms showed slower declines, suggesting higher plasmid persistence potential in disturbed environments. The total community resistome exhibited pronounced but short-lived shifts whereas indigenous resistomes and microbiome composition remained stable. Yet, in one replicate from the wastewater-impacted site, specific indigenous ARGs of public health relevance increased, indicating that disturbance can promote localized ARG proliferation even without sustained invader establishment. Our results show that interactions between invaders and indigenous biofilm are dynamic and strongly shaped by community composition. This supports the One Health concept, highlighting how environmental context modulates AMR propagation risks in freshwater ecosystems.\u003c/p\u003e","manuscriptTitle":"Invasion dynamics of antimicrobial-resistant E. coli in river biofilms: impacts on the resistome, microbiomes, and horizontal gene transfer","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-09 12:57:27","doi":"10.21203/rs.3.rs-8270852/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-02-17T17:22:07+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-18T14:51:12+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-01-05T00:47:14+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"269479355238656016923185338700837031823","date":"2025-12-23T14:14:07+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"160592732485391025134089005255248943381","date":"2025-12-22T16:23:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"290531083551443740713183818972480637078","date":"2025-12-20T14:16:39+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"135565258361071046050387021543645793886","date":"2025-12-18T14:37:01+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-12-18T13:53:56+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-12-05T15:35:11+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-12-05T03:45:44+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Antimicrobials and Resistance","date":"2025-12-03T13:17:50+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-antimicrobials-and-resistance","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"npjamar","sideBox":"Learn more about [npj Antimicrobials and Resistance](http://www.nature.com/npjamar/)","snPcode":"44259","submissionUrl":"https://submission.springernature.com/new-submission/44259/3","title":"npj Antimicrobials and Resistance","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"NPJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"862c140d-296e-4c92-9840-2313ec7371d3","owner":[],"postedDate":"December 9th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":59260349,"name":"Biological sciences/Ecology"},{"id":59260350,"name":"Earth and environmental sciences/Ecology"},{"id":59260351,"name":"Earth and environmental sciences/Environmental sciences"},{"id":59260352,"name":"Biological sciences/Microbiology"}],"tags":[],"updatedAt":"2026-04-14T13:23:17+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-09 12:57:27","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8270852","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8270852","identity":"rs-8270852","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}