The Effect of Glyphosate And Ciprooxacin Eexposure On The Gut Bacterial Microbiota Diversity of Rhinella Arenarum (Anura: Bufonidae) Tadpoles

The high load of agrochemicals and antibiotics coexisting in aquatic environments within agroecosystems represents a risk for wildlife. As the gut microbiota plays a key role on its host’s functioning and is sensitive to a wide variety of pollutants, its study allows evaluating organisms’ health and therefore, the ecosystem. We studied toxic effects of commercial formulations of a glyphosate-based herbicide (GBH) and the antibiotic ciprooxacin (CIP) on gut bacterial microbiota diversity of the common toad (Rhinella arenarum) tadpoles, considered a sentinel species. The study was carried out by classic microbiological analysis and matrix-assisted laser desorption ionization-time of ight mass spectrometry. The microbiota from GBH treatment had greater taxa diversity and richness, including some genera, such as Proteus spp. and Yersinia spp. that were absent in control. In contrast, microbiota from CIP treatment registered a decrease of diversity indexes, dominance of Aeromonas spp. and presence of Leclercia spp. The GBH-CIP treatment showed changes in taxa composition, including decrease of Klebsiella spp. and Pseudomonas spp. and trends of individual pollutant treatments. For all cases, changes in the composition of bacterial community (dysbiosis) were linked to a signicant decrease in tadpoles’ weight. To the best of our knowledge, this is the rst report on the disruption of gut bacterial microbiota of tadpoles by a mixture of two contaminants of emerging concern worldwide. These ndings contribute to understanding how the presence of two co-occurring pollutants in freshwaters results in deleterious effects on the amphibian community and potentially affect the microbiota of those environments.


Introduction
Classic studies based on gut contents have classi ed tadpoles as herbivores, detritivores, microphagous, and suspension feeders (Lajmanovich 1997(Lajmanovich , 2000. For these reasons, tadpoles transfer nutrients, and energy between aquatic and terrestrial ecosystems (Capps et al. 2015). Those herbivorous tadpoles' species that mostly obtain signi cant nutrition from autotrophic sources are likely to be assisted by the gut microbiota (Altig et al. 2007). In this sense, Lajmanovich et al. (2001) carried out a quantitative description of all gut bacteria of Rhinella arenarum tadpoles and de ned them as reservoirs of bacteria of sanitary interest, indicators of contamination and health risk in aquatic environments.
At present, where the load of contaminant of emerging concern (CECs) (e.g. pharmaceuticals and pesticides) is increasing worldwide in water bodies every day as a result of intensive agriculture and livestock practices (Varol and Sünbül 2017), it is crucial to evaluate their effect on the bacterial community in freshwaters and organisms (Evariste et al. 2019). The stability of microbial community structure is important not only for the health of the host organisms but also for the overall functioning of freshwater ecosystems due to its primary productivity (Vera et al. 2010;Villeneuve et al. 2011;Lozano et al. 2020). Amphibians have a complex life cycle and the species-speci c physiologic and behavioural approaches can be used as a reliable model for evaluation of biological toxic effects of environmental pollutants (Hopkins 2007). Exposure to herbicides during the developmental stages of tadpoles can induce immediate and lasting alterations to gut microbiome (Shehata et al. 2013; Knutie et al. 2018).
Changes in the composition of these bacterial communities (called dysbiosis) could lead to a disruption of metabolic capacities, increase susceptibility to disease, pathogenic infections and lead to increased risk of amphibian decline (Jiménez and Sommer 2016). Furthermore, the occurrence of pharmaceutical residues (e.g. antibiotics) as CECs in the aquatic environment alarm scienti c community in the last decades (Milić et al. 2013, Godoy and Sánchez 2020). The presence of antibiotics in aquatic systems is an issue of heightened interest throughout the world, due to bacterial acquisition of resistance to antibiotics (Kurenbach et al. 2018). Several  Glyphosate (GLY) is the herbicide more used worldwide (Benbrook 2016). In aquatic environment, a concentration of 3.49 mg/L has been estimated in the worst-case scenario (Wagner et al. 2013). GLY performs as an inhibitor of 5-enolpyruvylshikimate-3-phospate synthase (EPSP synthase), not only in crop plants but also in bacteria. Microbial communities have been studied to be affected by GLY and other herbicide cocktails (Lozano et al. 2020). Inhibiting results on EPSP synthase from gut microbiota has been reported, affecting principally bene cial bacteria. Consequently, researchers have suggested that GLY can cause gut dysbiosis, a disturbance that is characterized by an imbalance between bene cial  Boccioni et al. 2020). This was the rst study to investigate how these two pollutants affect the health of R. arenarum tadpoles, causing morphological abnormalities, thyroid disruption, and delayed development. However, the effect of those pollutants on tadpoles' gut microbiota communities remains unstudied. Considering the importance of gut microbiota to amphibian tadpoles, common toad (R. arenarum) tadpoles were used to study the toxic effects of GBH and CIP exposures on community diversity and structure of gut bacterial microbiota of tadpoles by classic microbiological analysis and to identify the microbial species by matrix-assisted laser desorption ionization-time of ight mass spectrometry (MALDI-TOF MS). The effect of pollutants on the gut microbiota will help to elucidate the comprehensive impact of environmental relevant exposure to GBH and CIP in pond environments in which tadpoles are raised, on the amphibian community with implications of wildlife conservation.

Chemicals
Bioassays solutions were prepared using a commercial formulation of a GBH (74.7 % active ingredient, Nphosphonomethyl glycine and inert adjuvant quantum satis; Roundup Ultra-Max®, Monsanto© Argentina;), and CIP (Sigma-Aldrich, Germany). Ultra-Max® GBH formulation and Sigma-Aldrich CIP were analytically evaluated in previous works, and for both cases, the error did not exceed 5% of the nominal concentrations The experiments followed the regulations of the American Society of Ichthyologists and Herpetologists (ASIH 2004). As the study of the gut microbiota required the isolation of fresh intestinal content, the individuals had to be sacri ced and immediately processed for morphological and microbiological analysis. Individuals were sacri ced by immersion in a solution of 0.1% tricainemethanesulfonate (TMS, MS-222) buffered at pH 7.8 with NaHCO3, following the protocol of the Animal Euthanasia Guide proposed by the Institutional Committee for the Care and Use of Animals (IACUC), and that of the bioethics committee of the FBCB-UNL (Res. Nº 388/06).

Morphological parameters
At the end of the bioassay, 15 individuals from each treatment (5 from each replica) were randomly selected for morphological evaluation. The organisms were xed in 10% Formaldehyde and preserved in 70% alcohol. The individuals were dry weighed (W, in mg) on a Pioneer Ohasus ® digital scale with 0.0001 g precision, and evaluated under an Arcano® stereoscopic magnifying glass with a Moticam® digital camera attached, to determine the total length (TL, in mm) and the stage of development (GS) according to Gosner (1960)

Microbiota sampling and analysis
Intestinal tracts of 15 individuals per treatment were aseptically removed, weighted and pooled in 3 samples (one from each replica, consisting of 5 individuals each) due to the low tissue volumes. Pool samples were homogenized in 500 µL of sterile peptone water using sterile glass beads (425-600 µm diameter) for intestinal walls rupture. Serial dilutions (up to 1/10000) of homogenized intestines were plated onto nutrient agar plates (0.5% pluripeptone; 0.3% meat extract; 0.8% NaCl) and incubated at 37°C for 24 h. The plate count method was used to calculate the amount of colony colony-forming units per gram of intestine (CFU/g) for evaluating quantitative differences between treatments.
In order to study diversity of species, 20 CFUs were randomly selected from the plates of each pool sample (Total N per treatment = 60 CFUs). Each isolated CFUs was individually re-suspended in 1 mL nutrient broth medium (0.5% pluripeptone; 0.3% meat extract; 0.8% NaCl, agar 2%) and incubated at 35-37 ° C overnight. Each CFU was rst morphologically characterized by gram staining reaction and seven biochemical tests: triple sugar iron agar (TSI), citrate, indole, motility, urease, fenilalanine and lysine-iron. For the strains that it was possible, the identi cation was made using phenotypic pro ling (according to Hawkey 2006, Lopardo et al. 2016, Ochoa and Ochoa 2017), but the unusual and di cult-to-identify strains required further analysis. Those strains were identi ed by matrix-assisted laser desorption/ionization-time-of-ight (MALDI-TOF) using the VITEK MS system (bioMérieux). MALDI-TOF based identi cation was also used to con rm the identi cation of 10% of the strains identi ed by classical biochemical tests.

Data analysis
All data regarding morphological biomarkers are reported as the mean ± SD. Effects of treatments on TL and W were analyzed with ANOVA followed by post-hoc Dunnett's Comparison Test. Differences in GS and CFU/g of tissue were evaluated by Kruskal-Wallis test followed by Dunn's post hoc test. Plate count results are expressed as CFU and were used to calculate CFU/g of tissue (gut) as follow: [N° colonies/ml plated * DF] / [ grams tissue/ml original homogenate] = N° colonies/gram tissue = CFU/g tissue. Data is shown as log (CFU/g of intestine).
Alpha diversity parameters (Chao1 richness estimator, Dominance index, Shannon and Simpson diversity index) were calculated using PAST 3.22 Software (Hammer et al. 2001), and are expressed as the mean ± SD (from replicas). Taxa richness is expressed by total genera per treatment. MANOVA (Wilks' lambda multivariate test statistic) was performed to determine whether there were signi cant overall differences in microbiotic diversity parameters among treatments, and subsequent univariate analysis of variance (ANOVA) test followed by Dunnett's post hoc-tests were done for each parameter.

Morphological parameters
The mean value of CO tadpoles' weight at the end of the assay was 69.98 ± 0.87 mg. Tadpoles treated with GBH, CIP and GBH-CIP showed a signi cant decrease (between 13-9%) on weight respect to CO (F = 5.8, p < 0.01; Fig. 1a). Mean weight in tadpoles from GBH treatment was 61.06 ± 0.62 mg, while in CIP was 60.06 ± 0.85 mg, and in GBH-CIP treatment, 63.78 ± 0.97 mg. The mean value of CO tadpoles' length was 9.36 ± 0.56 mm, while the values from treatments (9.58 ± 0.73, 9.8 ± 0.75, and 9.71 ± 0.42 mm for GBH, CIP and GBH-CIP, respectively) did not result in signi cant differences (F = 2.6, p > 0.05; Fig. 1b). On average, tadpoles from CO and all treatments excepted for GBH, ended the assay in a mean GS 27, while those from GBH showed lower GS (mean GS 26, KW = 8.559, p < 0.05, Fig. 1b).
were identi ed as the predominant taxa in the CO group. After treatment, the relative prevalence of Aeromonas spp. increased in detriment of the other named genera. This genus was dominant in both CIP and mixed treatments. In contrast, Aeromona spp. dominance was lower in the GBH treatment group (Table 1). Yersinia spp. and Proteus spp. were isolated in these groups and not in the others. The PCA for Beta diversity resulted in four groups corresponding to treatments samples (Fig. 3a). Treatments showed to be scattered through components axis, indicating that microbial communities of gut samples from CO, GBH, CIP and GBH-CIP varied among taxa. In the X axis (where most of variation was represented, λ PC1 = 59,2%), CO community were closely related to GBH and GBH-CIP ones, while CIP community was the most different treatment in according to the taxa present in each one (Fig. 3a).
Similarly, in UPGMA the GBH-CIP treatment was closely related to CO, and while GBH was closer to them than CIP, based on the relative abundance of taxa in each treatment (Fig. 3b).

Discussion
Amphibians face anthropogenic pollution threats from different sources, and their associated gut microbiomes are highly sensitive to these pollutants (Jiménez and Sommer 2016). Pesticides are likely to co-occur with antibiotics that are worldwide used in human and veterinary medicine and arrive to aquatic environments from agroecosystems and swine, chicken and cow feedlots without any treatments (Rico et al. 2014). The interactive impact of antibiotics and herbicides on gut microbiota disturbance has rarely been studied (Zhan et al. 2018). The present study reinforced that GBH and CIP, individually and in mixture, altered gut microbiota composition of R. arenarum tadpoles. A decrease in weight mass and stage of development in GBH treatment-was also observed in treatments with microbiota alteration.
It is known that gut microbiota plays key roles in host vital functions as immune-system modulation, digestion, biotransformation and protection against pathogens (Sommer et al. 2013). Therefore, dysbiosis of microbiota, as well as alteration of its functional optimization can lead to severe health problems for hosts (Claus et al. 2016). In this study, decrease on tadpoles' weight from both individual and mixture pollutants treatments was observed, together with a disruption of normal microbiota community structure. Moreover, in GBH treatment, tadpoles also showed a delay on development. Similar to previous studies that described gut microbiota shifts related to weight and changes on development Chai et al. 2018), these results suggest a relation between microorganism community and the animals' physiology (Nehra et al. 2016) and enhance the importance of study gut microbiota to estimate impact of environmental chemicals on health and tness of wildlife, environments and humans (Nguyen et al. 2015).
The taxa that have been found in this study as part of the microbial communities of R. arenarum tadpoles mostly belong to the bacilli Gram (-) Enterobacteriaceae family, also Aeromonas and Pseudomonas species. The genera of bacteria found in the microbial communities of CO tadpoles coincided with those described in pioneering works in the area that share the culture sowing methodology Another explanation for the shift in bacteria community composition in GBH treatment may be related to the ability of some bacteria to transform GLY (Sviridov et al. 2015). Some bacteria transform GLY into aminomethylphosphonic acid (AMPA) by the enzyme glyphosate oxidoreductase, and use this metabolite or directly the GLY molecule to obtain phosphate for their metabolism by C-P bond break down (Imparato et al. 2016). Thus, it can be inferred that GBH application may induce to an arti cial selection that stimulates existing bacteria capable of degrading the herbicide (Villarreal-Chiu et al. 2017). In accordance, some of the taxa increased in GBH treatment of our study, such as Enterobacter spp. and Providencia spp., are known to have associative traits to GLY degradation or use as a substrate (Nourouzi et al. 2011, Kryuchkova et al. 2014. From an ecological point of view, the decrease of some bacterial species could release ecological niches that would be occupied by others (Blot et al. 2019). Regarding this, the importance of deepening the study of the relationships between the different taxa that make up the intestinal microbiota here is highlighted, and the imbalance that could be generated between bene cial and pathogenic bacteria for amphibian tadpoles.
Concerning the quanti cation of CFU, GBH treatment showed a signi cant increase compared to CO. This effect could be related, among various factors, to host transcriptional changes enriched for lipid and carbon metabolism, as is suggested by Suppa et al. (2020). Moreover, studies on dynamics of bacteria communities of the soil and rhizosphere, associated the increase of fast-growing bacteria abundance with the availability of carbon compounds in GLY presence (Imparato et al. 2016). The increase in the amount of CFU in the GBH treatment probably corresponds to the increase in those taxa capable of degrading GLY. Consequently, the production of AMPA would increase, together with its potential risks to hosts animals and human health (e.g. impairment of DNA reparation and mRNA synthesis, Allemann Similarly, it is important to pay attention to the effects that drug residues may have on the bacterial communities of non-target organisms, since these are frequent in water bodies ). In our study, CIP treatment induced dysbiosis on gut microbiota of R. arenarum tadpoles by reduction of taxa diversity and increase dominance of a single genus. Aeromonas spp. represented more than 50% of relative taxa abundance on microbiotal gut community, assuming its resistance to CIP. This result is consistent with other studies that reported multidrug-resistance (including CIP) of Aeromonas spp. from wild animals (Dias et al. 2018 In the present study, richness and diversity index in the gut microbiota from GBH-CIP treatment were similar to CO, but the taxa composition showed to be different. Some genera from CO as Klebsiella spp. and Pseudomonas spp. were decreased or absent in the mixture treatment. Additionally, some trends observed for individual pollutant treatments were repeated in CHB-CIP: increase of Enterobacter spp. and presence of Proteus spp. (as in GBH), and increase of Aeromonas spp. and presence of Leclercia spp. (similar to CIP treatment). It is more than clear that pressure of both xenobiotics interacts to in uence microbiotal community structure. Results observed in GBH-CIP mixture treatment not only con rmed the susceptibility of gut bacterial microbiota in R. arenarum tadpoles to different type of pollutants individually, but also enhance their effects in mixtures, as they are more likely happen in the environment (Ramakrishnan et al. 2019). To the best of our knowledge, this is the rst report on disruption of gut microbiota of amphibian tadpoles by a mixture of an antibiotic and herbicide. As it is clearly aimed on a recent review, CECs affect gut bacteria and have great imbalance on host health (Tsiauossis et al. 2019).
More studies are need to elucidate how real-life scenarios with complex CECs mixtures can affect tadpole microbiota and ultimately, life aquatic health.
Overall, gut bacterial microbiota demonstrated to be a key endpoint for evaluating the effects of pollutants on non-target animals as amphibians' tadpoles. Last years, there has been a growing interest and concern about its diversity and structure variation due to changes in environmental conditions and pressures in order to understand its complex symbiotic relations with hosts' life (Evariste et al. 2019), and the bacterial resistance due to exposure to antibiotics such as CIP (Jørgensen et al. 2013).

Conclusion
The results of our study suggest that commercial formulations of GBH and CIP, individually and in mixture, caused severe gut bacterial microbiota dysbiosis of R. arenarum tadpoles affectingindividuals' weight. Apart from the already known direct effects of these pollutants on tadpoles, the dysbiosis they cause may lead to additional physiological problems through alteration of the gut microbiota normal functioning, including metabolic activities related to nutrients and energy recovery. Further studies about the gut microbiota on tadpoles exposed to pollutants mixtures such as herbicides (e.g. GLY) and antibiotics (e.g. CIP) and the potential use of microbiota composition as biomarker to apply not only to environmental risk assessment but also on wild animals health are urgent needed. Furthermore, the bacterial communities' dynamics in face of CECs is essential for understanding bacterial resistance, a highly-complex and growing concern nowadays for human health.

Declarations
Funding This study was supported by National Agency for Promotion of Science and Technology (PICT Nº 1069).

Con icts of interest / Competing interests
The authors have no con icts of interest to declare that are relevant to the content of this article.
Availability of data and material (data transparency) Data presented in this study are available on request from the corresponding author.