Response of three biofilm-forming benthic microorganisms to Ag nanoparticles and Ag+: the diatom Nitzschia palea, the green alga Uronema confervicolum and the cyanobacteria Leptolyngbya sp.
- 262 Downloads
Although the industrial use of nanoparticles has increased over the past decade, the knowledge about their interaction with benthic phototrophic microorganisms in the environment is still limited. This study aims to characterize the toxic effect of ionic Ag+ and Ag nanoparticles (citrate-coated silver nanoparticles, AgNPs) in a wide concentration range (from 1 to 1000 μg L−1) and duration of exposure (2, 5 and 14 days) on three biofilm-forming benthic microorganisms: diatom Nitzschia palea, green algae Uronema confervicolum and cyanobacteria Leptolyngbya sp. Ag+ has a significant effect on the growth of all three species at low concentrations (1–10 μg L−1), whereas the inhibitory effect of AgNPs was only observed at 1000 μg L−1 and solely after 2 days of exposure. The inhibitory effect of both Ag+ and AgNPs decreased in the course of the experiments from 2 to 14 days, which can be explained by the progressive excretion of the exopolysaccharides and dissolved organic carbon by the microorganisms, thus allowing them to alleviate the toxic effects of aqueous silver. The lower impact of AgNPs on cells compared to Ag+ can be explained in terms of availability, internalization, reactive oxygen species production, dissolved silver concentration and agglomeration of AgNPs. The duration of exposure to Ag+ and AgNPs stress is a fundamental parameter controlling the bioaccumulation and detoxification in benthic phototrophic microorganisms.
KeywordsAgNPs Ionic silver Growth inhibition Biosorption Complexation Biofilm
New technologies have been created over the past decade, exploiting and synthetising materials at a scale smaller than 1 μm (Klaine et al. 2008). These materials are called ‘nanomaterials’ or ‘nanoparticles’ and are defined as natural or artificial structures with at least one dimension between 1 and 100 nm (Nel et al. 2006). They are characterized by a high specific surface area (Navarro et al. 2008a; Nel et al. 2006), abundant and highly reactive surface sites and high mobility in aqueous solution (Navarro et al. 2008a). In the present manuscript we investigated the effect of silver nanoparticles (AgNPs) on microorganisms. Silver nanoparticles (AgNPs) are widely used in the industry for their high thermal and electrical conductivity, chemical stability and catalytic activity (Capek 2004; Frattini et al. 2005). All these properties provide a specific physico-chemical behaviour of AgNPs in the environment, controlling their interaction with biological matrices. The AgNPs are mainly applied in water treatment, surfaces (antifouling), food packaging (Perelaer et al. 2009), renewable energy (Pavasupree et al. 2006), environmental remediation (Tungittiplakorn et al. 2004; Zhang 2003) and medicine (Barnett et al. 2007). The AgNPs are partially released into natural waters as demonstrated for AgNP-treated textiles (Benn and Westerhoff 2008) and facade paints containing AgNPs (Kaegi et al. 2010). The concentration of AgNPs in the environment has been predicted to achieve ∼1 μg L−1 in solution and ∼14 mg kg−1 in sediments (Blaser et al. 2008; Gottschalk et al. 2009; Mueller and Nowack 2008).
The agglomeration of nanoparticles is affected by pH, ionic strength or salinity, coating agents, electrolyte valence and the concentration of natural organic matter (NOM) in solution (El Badawy et al. 2010) as well as by speciation in aqueous solution, which in turn is controlled by pH, redox state, ionic strength and NOM concentration (Marambio-Jones and Hoek 2010). All aforementioned factors lead to different degrees of AgNPs agglomeration, which result in different toxic effects on microorganisms (Auffan et al. 2009; Choi and Hu 2008; Jiang et al. 2009; Liu et al. 2010; Pal et al. 2007; Park et al. 2010). However, agglomeration does not always imply a decrease in the toxicity of nanoparticles for microorganisms (Arulvasu et al. 2014; Wei et al. 2010).
It is generally accepted that the toxicity of AgNPs in solution occurs due to their oxidative dissolution, which results in the release of Ag+ ions. In natural water, silver is oxidized following the typical Fenton-like reactions, where reactive oxygen species (ROS) are produced as superoxide radical, hydroxyl radical and hydrogen peroxide. ROS have toxic effects on the cell membrane of organisms (Fabrega et al. 2011a; Feng et al. 2000; Kittler et al. 2010) because of their accumulation outside the membrane followed by their internalization in the cell (Sondi and Salopek-Sondi 2004). In fact, ROS can damage the cell surfaces by peroxidation of lipids as demonstrated at transcriptome, proteome and cellular level for Chlamydomonas reinhardtii (Pillai et al. 2014).
The toxic effects of AgNPs on the microbial community of phototrophic biofilms are especially important because these biofilms represent the first link in the food chain (Fabrega et al. 2011a, b; Kalishwaralal et al. 2010). Phototrophic benthic microorganisms such as eukaryotic, bacteria, cyanobacteria, microalgae and fungi are the main biofilm constituents (Kroll et al. 2016). They are responsible for primary production in streams, thus controlling biogeochemical cycles of carbon and metals in terrestrial freshwater systems and likely to be damaged by the presence of soluble Ag. The toxic effects of AgNPs and Ag+ have been studied under different conditions but the impact of AgNPs on individual microorganisms is still insufficiently understood, mainly because of the diverse experimental methods, protocols and type of nanoparticles (Moreno-Garrido et al. 2015). Molecular mechanisms of the toxic effects of silver on algae have recently been revealed (Pillai et al. 2014). The toxicity of citrate-AgNPs has been demonstrated by different microorganisms as Chattonella marina, Pseudokirchneriella subcapitata and C. reinhardtii (He et al. 2012; Lee et al. 2005; Ribeiro 2014) as well as bacteria such as Candida glabrata, Candida tropicalis and Staphylococcus aureus (Wady et al. 2014). The covering agent also has to be taken into account in toxicity experiments as reported by Angel et al. (2013). On the other hand, the toxicity effects of AgNPs have been related to the photosynthetic activity (Navarro et al. 2008a, b). Moreover, the speciation of Ag+ in aqueous solution and AgCl°(aq) complexes can also affect the toxicity for microorganisms. In addition, AgNPs also counteract biofilm formation (Moreno-Garrido et al. 2015). This was demonstrated by Gonzalez et al. (2015), where the effect of Ag+ and AgNPs on the phototrophic biofilm was clearly pronounced at micrograms per level, similar to that reported in periphyton biofilm (Gil-Allué et al. 2015). Natural phototrophic biofilm, studied in our previous work (Gonzalez et al. 2015), was dominated by the diatom Nitzschia palea, the green alga Uronema confervicolum and the cyanobacterium Leptolyngbya sp. In this paper, we hypothesise that the difference in the impact of Ag+ vs AgNPs on the whole biofilm community will also be observed at the level of individual biofilm constituents. To this end, we studied the effect of ionic silver (Ag+) and particulate (AgNPs) on each species during incubation in monocultures. The tolerance of individual microorganisms in a wide range of added Ag was tested via toxicity studies in the range of 1 to 1000 μg L−1 of dissolved Ag. We used citrate-coated AgNPs because their properties under specific conditions (pH, light, dark, ionic strength and natural organic matter) are well known (Tolaymat et al. 2010). In addition, as the exposure time is a key factor controlling possible effects of silver on microorganisms (Kroll et al. 2016), we studied toxic effects from 2 to 14 days of exposure, which is comparable with the growth cycle (exponential-stationary-decay) of algae.
In accord with the hypothesis and chosen experimental setup, the following specific questions were addressed in this work: (i) How significant is the effect of exposure time between 2 and 14 days on the degree of Ag toxicity? (ii) Does Ag+ and AgNPs speciation in solution affect the degree of Ag toxicity? (iii) Which phototrophic species are most resistant to the presence of Ag+ and AgNPs?
Material and methods
AgNO3 was used as the source of ionic silver (Ag+) (Prolabo, purity = 99.8 %). Commercial AgNPs were obtained as dispersed solution of 20 mg L−1 (Sigma-Aldrich). The stock solution was kept in the fridge <1 month before use; the typical shelf life for storage of commercial AgNPs solution is >6 months. The AgNPs in this solution are spherical, have an average diameter of 20 ± 0.2 nm as measured by transmission electron microscopy (TEM, JEOL JEM-1400 HC and camera Gatan Orius 11Mpixels) and are stabilized with sodium citrate coating to prevent the formation of aggregates. AgNPs were sonicated during 10 min before use, allowing their better dispersion (Fabrega et al. 2009, 2011a). TEM images were taken by immersing grids coated with a carbon film for 10 s in prepared algal suspension. Slightly dried grids were used for TEM imaging. An average of 16 images were collected for each session to minimize subjective bias in imaging the nanoparticles. Fig. S1 shows that the nanoparticles are well dispersed in the solution and exhibit homogeneous size distribution in the stock solution as it is seen for AgNPs stock solution (Fig. S1a) suspended in the Combo media during 1 h at 50 μg L−1 (Fig. S1b) and 1000 μg L−1 (Fig. S1c). Sodium citrate dehydrated Na3C6H5O7 (LABOSI, purity = 99.973 %) tests were performed to elucidate the effect of this possible nutrient on silver toxicity for the microorganisms. All reactants were chemical grade level and they were prepared with Milli-Q water (Millipore Direct-Q 3 UV, 18.2 MΩ cm, 25 °C).
Maintenance of microorganism strains
The three benthic prototrophic microorganisms originate from the algae collection of the EcoLab facilities (Toulouse). Individual cultures of the three species were prepared every 14 days in 250-mL Erlenmeyer flasks with nutrient media adapted to each species. For N. palea and U. confervicolum, Combo was used as the culture medium (Kilham et al. 1998) and for Leptolyngbya sp., the BG-11 medium was used (Stanier et al. 1971). The ionic strength of both media was 2.7 mM. The strains were maintained at 18 °C under white light of 40 ± 10 μmol s−1 m−2 with light/dark periods of 16 h:8 h. All preparations and samplings were performed under a UV-sterilized laminar flow hood (Jouan MSC 12).
Silver toxicity tests
The toxicity tests were performed using sterile 12-well (6 mL) microplates from Fisher Scientific kept in a culture chamber (Friocell) at 18 °C, illuminated under white light of 40 ± 10 μmol s−1 m−2 with light/dark periods of 16 h:8 h. The initial concentration of wet biomass (WB) was always kept constant as 1 gWB L−1 for N. palea and Leptolyngbya sp. and 5 gWB L−1 for U. confervicolum. A higher initial WB was used for U. confervicolum in order to increase the sensitivity of the analysis. The cell count was not possible because of biofilm or filament formation. For this reason and for consistency with previous studies, we expressed the cell concentration in grams wet biomass per litre. The aliquots of stock cultures were centrifuged at 17,000g for 10 min (centrifuge Heraeus Multifuge) and rinsed three times with the experimental solution to remove all possible exudates, cell debris and exopolysaccharides (González et al. 2010). After removal of the supernatant, the biomass was weighed and the culture medium was added into the wells to reach the expected volume. Afterwards, the Ag+ or AgNPs were added, setting the initial conditions for the experiments, from 0 to 1000 μg L−1 of both forms of silver, except for the 2-day exposure with maximum concentration of 5000 μg L−1. The use of the culture media guarantees that any toxicity effect is due to the addition of Ag+ or AgNPs and never due to the lack of nutrients. The stability of both Ag+ and AgNPs in supernatant at conditions similar to those of biotic experiments (pH, DOC, Cl−) was verified in a series of cell-free tests and was found to be sufficiently high (max 10 % decrease over several days of exposure) up to 1000 μg L−1 of Ag. These tests allowed quantifying the adsorption of Ag+ and AgNPs to the walls of the wells, Ag+ hydrolysis and precipitation and AgNPs aggregation and removal from solution. The Ag+ released from AgNPs was measured after centrifugation 25,000g for 2-h and 2- and 5-day exposure.
The biomass production was evaluated after 2, 5 and 14 days of incubation both in the absence (control) and in the presence of different Ag+ or AgNPs concentrations. For sampling, the supernatant was removed from the wells and centrifuged at 17,000g for 10 min, filtered through a pore size of 0.45 μm, acidified with suprapure bidistilled HNO3 (pH 2) and stored in a refrigerator at 4 °C until the analysis of total dissolved Ag concentration in solution. Aliquots were also removed and filtered to measure the dissolved organic carbon (DOC) during the experiments. After sampling the dissolved components, the biofilm which had formed on the bottom of the well was collected and quantitatively transferred to a 1-mL vial with fresh culture medium. The optical density (OD) was immediately measured with the spectrophotometer (Biochrom Asys UVM 340) at a wavelength of 680 nm, which corresponds to the peak absorbance of chlorophyll-a. The samples were examined with an optical microscope after the experiments and no visible changes were observed in the cell structure and shape. The supernatant was acidified with bidistilled HNO3 and kept at 4 °C, under darkness, until analysis. All the experiments were performed in triplicates.
The dissolved silver concentration in supernatant solution was measured using graphite atomic absorption spectroscopy (GF-AAS, PerkinElmer AAnalyst 600) with a detection limit of 1 μg L−1 and an uncertainty of ±5 %. DOC was analysed with a total organic carbon metre (Shimadzu TOC-CSN), which has a detection limit of 0.1 mg L−1 and an uncertainty of ±3 %. The OD followed a linear relationship (R2 > 0.995) with the wet biomass (WB) for all the species (Fig. S2).
Effective concentration 50 (EC50) (concentration of silver that yields 50 % of growth inhibition) was calculated using a polynomial fit to the growth curve as a function of Ag concentration.
Ag+ speciation was calculated by Visual Minteq codes (Gustafsson 2012) as presented in the Supporting Information S1 and Fig. S4. For this speciation modelling, the effect of DOM was taken into account using the Stockholm Humic Model (SHM) database. SHM considers a discrete site distribution to explain the proton and metal binding by organic ligands such as humic or fulvic acids. This model-ligand can be used to study the aggregation in gel-like structures (van Schaik 2008). In the absence of any direct measurements of Ag+ cell exometabolites, soluble EPS and protein complexes, the complexation of Ag+ with DOM was approximated by natural organic matter.
Statistical data analyses were performed with Statgraphics Centurion XVI in order to compare the growth of each species (N. palea, U. confervicolum and Leptolyngbya sp.) in the presence of Ag+/AgNPs and each exposure time (2, 5, 14 days). Nine statistical tests were performed to discern if the effects of Ag+ and AgNPs were significantly different. Appropriate statistical tests were applied in each case for the available data. The first step was to check the normal distribution of the data set for each species, exposure time and metal species (Ag+/AgNPs) with a Saphiro-Wilk test. Those data groups that did not follow a normal distribution were compared by performing a Mann-Whitney test in order to compare the medians. Only Leptolyngbya sp. for 14-day exposure time followed a normal distribution. For this species, the homogeneity of variance between Ag+ and AgNPs was verified by performing first a Levene’s test (α = 0.05) and then a two-way ANOVA (analysis of variance) of two fixed and crossed factors (form of silver and concentration; α = 0.05) was performed. Considering also the biomass, we compared N. palea with U. confervicolum by using the two-way ANOVA because they followed a normal distribution, but this analysis was less rigorous because the biomass was a function of time. The comparison between species demonstrated that the results were always statistically significant (p value <0.05).
Growth of studied microorganisms in Ag-free solutions
Growth rate, dissolved organic carbon (DOC) and EC50 for N. palea, U. confervicolum and Leptolyngbya sp. in the presence of Ag+ and AgNPs, at 2, 5 and 14 days of exposure time
Ag+ (μg L-1)
EC50 μg L−1
EC50 μg L−1
EC50 μg L−1
(g L−1 day−1)
(g L−1 day−1)
(g L−1 d−1)
AgNPs (μg L-1)
EC50 μg L−1
EC50 μg L−1
EC50 μg L−1
(g L−1 day−1)
(g L−1 day−1)
(g L−1 day−1)
Taking into account the control experiments, U. confervicolum yielded the highest wet biomass after 14 days (141 ± 16 gWB L−1 for Ag+, 116 ± 13 gWB L−1 for AgNPs), followed by Leptolyngbya sp. (50 ± 12 gWB L−1 for Ag+, 73 ± 8 gWB L−1 for AgNPs) and N. palea (33.5 ± 0.8 gWB L−1 for Ag+, 34 ± 2 gWB L−1 for AgNPs). As the biomass evolution with time was practically linear and did not reach a constant value, it follows that the selected incubation time (2, 5 and 14 days) was within the active, linear or exponential growth stage of these species. Therefore, the chosen experimental conditions were optimal and any observed growth inhibition has to be directly related to the presence of Ag+ or AgNPs and not to the deficiency of nutrients or the cell number saturation in the culture. In order to assess the ability of microorganisms to form biofilm via extraction of soluble extracellular polymeric substances (EPS) to alleviate the potential toxicity of Ag, DOC production was monitored as a function of cell growth period in experiments without silver (Table 1). N. palea and Leptolyngbya sp. increased DOC concentration in solution during growth whilst U. confervicolum decreased the DOC production over time. Leptolyngbya sp. yielded the highest DOC concentration after 14 days of growth (∼15 mg L−1). The cells were examined using an optical microscope and no visible differences were observed in the cell structure and shape. Further investigation should be carried out in order to study the enzymatic response of these microorganisms in the presence of silver.
Citrate effect on the growth of microorganisms
The commercial AgNPs used in this investigation were stabilized by coating with sodium citrate; thus it was necessary to rule out the possible effect of un-bound citrate on cell growth. The effect of sodium citrate was studied in the concentration range from 1 μg L−1 to 10 mg L−1 to cover the maximal possible range of released ligand after 2 days of exposure and all other conditions similar to those of Ag-bearing experiments (Fig. S3). These studies were carried out in 0.01 M NaNO3 in order to allow testing the most elevated citrate concentration whilst keeping the ionic strength constant. No statistically significant effect of increasing sodium citrate concentrations on the growth of N. palea, U. confervicolum and Leptolyngbya sp. was observed (p value >0.99). The biomass production was independent of the added citrate concentration, although it varied significantly among the three species (p value <0.01) according to a Kruskal-Wallis test (α = 0.05): U. confervicolum, N. palea and Leptolyngbya sp. produced 24 ± 4, 7.7 ± 0.6 and 3.9 ± 0.3 gWB L−1, respectively, after 2 days of growth. Thus, it can be concluded that the effect observed in AgNP-bearing experiments was due to the presence of AgNPs without any interferences of aqueous citrate that could be released from AgNPs coating on the cell growth.
Silver speciation in solution
Silver speciation plays a key role in toxicity effects on the microorganisms in natural settings. It can be seen from Fig. S4 that Ag+-DOM complexation in Ag(I)-rich systems is relatively low as the proportion of Ag-DOM complexes is a factor of 10 lower than that of inorganic forms. Low complexation of Ag+ with DOC is indirectly confirmed by available voltammetric measurements in the seawater with aqueous metal/organic ligand ratio similar to those of the present study (Miller and Bruland 1995). A comparative inhibitory test (Ag+ and AgNPs) for N. palea in 0.01 M NaNO3 solution and in Combo medium for 2 days of exposure was indistinguishable for both forms of Ag within the reproducibility of three replicates (Fig. S5), suggesting that the effect of Ag complexation with nutrients of cell growth media is negligible and the degree of biomass growth inhibition is similar for NaNO3 and Combo media.
The effect of silver on the growth of microorganisms and the release of DOC as a function of exposure time
Effect of ionic silver (Ag+)
The three species considered in this study showed a progressive decrease in wet biomass in response to an increase of Ag+ concentration in solution. For 2-day exposure (Fig. 2a), the N. palea biomass was reduced by 65 % compared to the control culture when 20 μg L−1 of Ag+ was added. After 5 days of exposure, N. palea decreased its biomass from 13.3 ± 0.6 gWB L−1 in the control to 1.1 ± 0.2 gWB L−1 at 1000 μg L−1 Ag+ loading (Fig. 3a). Furthermore, N. palea exhibited a constant biomass of 34 ± 1 gWB L−1 which decreased significantly above 10 μg L−1 of Ag+ until 1.5 ± 0.7 gWB L−1 at 100 μg L−1 of Ag+ after 14 days of exposure (Fig. 4a). In this regard, the maximum growth inhibition of this diatom (>91 %) was achieved at 100 μg L−1 Ag+.
U. confervicolum showed a higher sensitivity to the presence of Ag+, as its biomass decreased from 18.8 ± 0.9 gWB L−1 (control) to 11.3 ± 0.8 gWB L−1 already at a concentration of 1 μg L−1 of Ag+ after 2 days (Fig. 2a). After 5 days, U. confervicolum growth was significantly inhibited by the presence of 100 μg L−1 of Ag+, since its biomass decreased from 57 ± 6 gWB L−1 in the control to 17 ± 4 gWB L−1 (Fig. 3a). However, for longer exposure time, U. confervicolum exhibited a stable biomass of 135 ± 8 gWB L−1 up to 100 μg L−1 Ag+, decreasing it to 13 ± 2 gWB L−1 only for 1000 μg L−1 Ag+ (Fig. 4a).
The cyanobacterium Leptolyngbya sp. exhibited an 88 % decrease of its biomass in the presence of 10 μg L−1 Ag+ after 2 days of exposure (Fig. 2a). Moreover, above 50 μg L−1 of Ag+, this cyanobacterium did not grow at all, exhibiting a clear mortality effect compared to the control experiments. This was the only case where mortality was reached. The biomass slightly increased when they were exposed to 500 and 1000 μg L−1 of Ag+, presumably due to the partial precipitation of Ag hydroxide. Leptolyngbya sp. biomass demonstrated a 66 % growth inhibition at 100 μg L−1 of Ag+ and an 88 % inhibition at 1000 μg L−1 Ag+ after 5 days (Fig. 3a), whilst for 14-day exposure, the biomass descended gradually from 50 ± 12 gWB L−1 (control) to 32 ± 8 gWB L−1 (100 μg L−1 Ag+) and achieved the maximum inhibition at 1000 μg L−1 Ag+ (14 ± 5 gWB L−1, or 72 % inhibition growth), as shown in Fig. 4a.
The inhibitory effect could be further illustrated in terms of EC50 values listed in Table 1. The diatom N. palea was the most resistant after a 2-day exposure with EC50 = 18.1 μg L−1 and the green alga U. confervicolum was the least resistant with an EC50 of 4.2 μg L−1 of Ag+. The sensitivity of microorganisms to Ag+ was a direct function of exposure time. Thus, after a 5-day exposure to Ag+, the most affected species was N. palea with EC50 of 17.3 μg L−1, followed by U. confervicolum and Leptolyngbya sp. with EC50 of 33.6 and 43.7 μg L−1, respectively. In contrast, after 14 days, U. confervicolum became the most resistant species with EC50 of 394 μg L−1, following Leptolyngbya sp. with an EC50 of 300 μg L−1.
The concentration of DOC over time is plotted for 2, 5 and 14 days of exposure in Figs. 2c, 3c and 4c, respectively. Generally, the DOC either decreased or remained constant with the increase of ionic silver concentration for all the species considered in this study, with the tendency of increasing DOC up to 100 μg L−1 of Ag+, then DOC either decreased or remained constant with the increase of Ag+. An exception was Leptolyngbya sp., for which a marked DOC decrease was observed up to 10 μg L−1 of Ag+.
Effect of Ag nanoparticles
The toxic effect of AgNPs on all three species was pronounced at much higher concentrations than Ag+. The measurements after 2 days of exposure showed that the wet biomass of N. palea remained constant at 2.8 ± 0.4 gWB L−1 until 1000 μg L−1 of AgNPs and at higher concentrations, >80 % inhibition occurred (Fig. 2b). For this diatom, the EC50 value was >1000 μg L−1 AgNPs (Table 1). The green alga U. confervicolum exhibited a constant biomass around 12.3 ± 0.5 gWB L−1 up to 20 μg L−1 of AgNPs (Fig. 2b) after 2 days of exposure. There was a ∼50 % inhibitory effect at 100 μg L−1; this effect reached 76 % at 2000 μg L−1 of AgNPs. The EC50 for U. confervicolum was >500 μg L−1 of AgNPs (Table 1). In the case of the cyanobacterium Leptolyngbya sp., the cell growth was significantly different in the presence of AgNPs than in the presence of Ag+ (p value <0.01). In addition, the toxic effect was not observed at the conditions of this study after 2 days of exposure: the biomass slightly increased until 100 μg L−1 of AgNPs and slightly decreased above 500 μg L−1 of AgNPs (Fig. 2b). For this cyanobacterium, the EC50 was >5000 μg L−1 of AgNPs (Table 1).
The 5-day experimental cultures of all three species were not significantly affected by AgNPs addition (Fig. 3b), with EC50 being >1000 for all three microorganisms (Table 1). N. palea biomass remained constant at around 22 ± 2 gWB L−1 up to 1000 μg L−1 of AgNPs, whereas U. confervicolum showed a small biomass reduction up to 10 μg L−1 of AgNPs, although this decrease was within the standard deviation of triplicates. The growth of N. palea and U. confervicolum was statistically significant in the presence of AgNPs compared to Ag+ (p value <0.01). The cyanobacterium Leptolyngbya sp. exhibited a constant biomass until 100 μg L−1 of AgNPs with a growth inhibition of 35 % at 1000 μg L−1 of AgNPs, although the cell growth was not significantly different than in Ag+ exposure (p value = 0.975).
After 14 days in AgNPs-bearing solutions, the growth of N. palea and U. confervicolum was constant at 34.1 ± 0.2 and 122 ± 12 gWB L−1, respectively (Fig. 4b). Leptolyngbya sp. demonstrated a slight growth inhibition (29 %) at 10 μg L−1 of AgNPs without further change up to 1000 μg L−1 of AgNPs (significantly different compared to Ag+, p value <0.01). The EC50 values for the three species were >1000 μg L−1 of AgNPs (Table 1).
In contrast to Ag+, the exposure of all three microorganisms to AgNPs produced a much smaller variation in DOC concentration (Figs. 2d, 3d and 4d). For N. palea and Leptolyngbya sp., after 2 days of exposure, the DOC only increased in 1000 μg L−1 of AgNPs solution by ca. 40 and 110 % above the control, respectively (Fig. 2d), whereas for U. confervicolum, the DOC decreased from 11.3 mg L−1 in the control to 7.5 mg L−1 for 1000 μg L−1 of AgNPs. After 5 days (Fig. 3d), the DOC remained constant around 4.0 ± 0.1 and 2.6 ± 0.3 mg L−1, respectively, for N. palea and Leptolyngbya sp. and it decreased for U. confervicolum from 13.1 mg L−1 in the control to 9.3 mg L−1 in the 10 μg L−1 of AgNPs solution. Finally, after 14 days (Fig. 4d), the DOC was constant for N. palea (6.1 ± 0.4 mg L−1) and U. confervicolum (9.0 ± 0.4 mg L−1 until 100 μg L−1 of AgNPs). For Leptolyngbya sp., the DOC was 1.5–2 times higher than the control value, ranging from 14 to 22 mg L−1.
Silver concentration in solution and its effect on growth inhibition
Silver toxicity as a function of exposure time
Long-term experiments are useful to understand the interaction between silver (Ag+ and AgNPs) and microorganisms. During long-term experiments, the AgNPs are subjected to changes in size (agglomeration) but the concentration of released Ag+ can also increase. On the other hand, microorganisms are able to produce organic ligand to alleviate the toxic effects of metals during longer periods (Rico et al. 2013).
It is considered that the toxicity of AgNPs is primarily related to their size, since smaller nanoparticles are more toxic than the larger ones (Albanese et al. 2012; Ivask et al. 2014; Tedesco et al. 2010). In this regard, the agglomeration effect may become important. The presence of exopolysaccharides produced by microorganisms or biofilm can induce a size increase of AgNPs (Kroll et al. 2014). In fact, these agglomerates may remain in solution for more than 7 days (Piccapietra et al. 2011) as also shown for citrate-coated AgNPs exposed to humic substances (Cumberland and Lead 2009; Huynh and Chen 2011). According with Zhang et al. (2011), the size of AgNPs increases very fast during the first 6 h, the agglomeration rate decreases and the size of AgNPs is 50-fold during the first 2 days, but only 10-fold after 14 days in solution. On the other hand, the organic ligands (DOM) act as strong inhibitors of AgNPs agglomeration in solution (Lodeiro et al. 2016). Accordingly, the toxic effects of AgNPs in our studies are likely to be affected by agglomeration, which could explain the lower toxicity of AgNPs than Ag+. According to Fig. S6, only 1–10 % of Ag is released from the AgNPs. The degree of growth inhibition in the presence of Ag+ varied between 83 and 92 % relative to the control for all species at 1000 μg L−1 of Ag+. For the same conditions, the inhibition of cell growth by AgNPs ranged from 0 to 43 % for all species. For the same added concentrations, Ag+ generated stronger inhibitory effects than AgNPs, especially during short (<5 days) periods of exposure.
The toxic effects of AgNPs are generally related to the concentration of Ag+ released into the solution, but it is still not clear whether the whole toxic effect is due to this release (Egorova 2011; Yin et al. 2011). In the case of benthic phototrophic microorganisms, the significantly lower toxicity of AgNPs relative to Ag+ can be explained by the fact that the majority of AgNPs are not directly bioavailable to microorganisms because they either have to be dissolved to release Ag+ or they have to penetrate the cell to generate toxic effects (Miao et al. 2010). Consistent with our macroscopic results, it is possible that the negative charge of AgNPs and the high concentration of DOC excreted by phototrophic microorganisms, also negatively charged, produced a repulsion effect between citrate-coated AgNPs and phototrophic microorganism cells, which resulted in only slight inhibition of growth in the presence of AgNPs. Indeed, surface properties of AgNPs coated by different compounds have been studied under different conditions (El Badawy et al. 2010). Citrate-coated AgNPs have negative ζ-potential, as has been reported previously, ranked from −13 to −41.8 mV, depending on the solution properties (pH, ionic strength and NOM; El Badawy et al. (2010); Gil-Allué et al. (2015)). The magnitude of negative charge for citrate-coated AgNPs was higher than that of AgNPs traditionally used for toxicity studies such as polyvinylpyrrolidone coated AgNPs or branched polyethyleneimine-coated AgNPs. As the AgNPs was coated with citrate molecules, this potential nutrient could alleviate the toxic effect of AgNPs, thus leading to underestimation of the true inhibiting effect. Results of culturing in the presence of citrate, however, dismiss this possibility as the biomass production was not affected by un-bound citrate presence for all three studied species.
Another accepted mechanism to explain the toxic effects of AgNPs is the production of ROS during the oxidation of the released Ag. This implies an oxidative dissolution of nanoparticles releasing Ag+ in the vicinity of the cells. However, the complexation of minor amounts of released Ag+ with DOC would decrease the overall toxicity of AgNPs with the increase of exposure time. Experiments on reactivity of AgNPs in culture medium demonstrated that less than 2 % of AgNPs was converted into Ag+ after 24 h of reaction (Fabrega et al. 2009). These authors used minimal Davies Medium at an ionic strength of 0.055 M which contained a mixture of phosphate and sulphate salts, citrate and glucose, comparable with the nutrient solution of the present study (I = 0.027 M). Wady et al. (2014) also studied the toxicity effects of AgNPs on biofilm by using the culture media and found that AgNPs decreased the biofilm-forming capacity for concentrations up to 4 mg L−1. In addition, El Badawy et al. (2010) reported that the constant conductivity for AgNPs for 3 weeks may imply a lower dissolution of AgNPs into Ag+. Note that the citrated-coated AgNPs are less prone to Ag+ release compared with PVP-AgNPs or Tan-AgNPs (Dobias and Bernier-Latmani 2013). Finally, Li and Lenhart (2012) studied the agglomeration and dissolution of AgNPs in natural waters in the presence and absence of sun light over 15 days and demonstrated that only 3 % of silver was released from citrate-coated AgNPs and this amount was not significantly affected by sunlight. These authors also reported that aggregated AgNPs did not easily release ionic silver and the released Ag+ is complexed with Cl−. The presence of organic ligands excreted by the microorganism should play a similar role of humic substances on the AgNPs reactions, where the presence of humic substances increases the stability of colloidal AgNPs (Fabrega et al. 2009; Lodeiro et al. 2016). This result is consistent with the general knowledge that organic ligands decrease the reactivity of metal ions in natural waters via complexation (López et al. 2015; Rico et al. 2013). Taking into account the measured Ag+ yield from AgNPs, the Ag+ concentration in our nutrient solution should range from 0.02 to 20 μg L−1 for AgNPs concentrations between 1 and 1000 μg L−1. This is consistent with a lower toxicity by a factor of 10 to 100 of AgNPs compared to Ag+ for phototrophic cell growth. Presumably, the microgram per litre-level concentration of toxic Ag+ produced in AgNP-bearing experiments is quickly complexed with soluble and surface EPS and cell exometabolites and thus does not affect live cells.
The observed difference between Ag+ and AgNPs is consistent with results of previous investigations. Gubbins et al. (2011) found that AgNPs were able to exert toxic effects on the growth of an aquatic plant (Lemma minor) for an exposure time of 14 days, whilst for the Ag+, the growth inhibition for the same concentration and period was much higher. He et al. (2012) found that the exposure to Ag+ for 1 h decreased in the cell viability of the marine raphidocyte C. marina significantly more than that of AgNPs. Experiments with Daphnia magna in the presence of AgNPs demonstrated that the aggregation of AgNPs can affect the EC50 but not the toxic effect and that the toxicity of Ag+ is higher than that of AgNPs (Park et al. 2014).
The species that was most resistant to short-term Ag+ exposure was N. palea, as reflected in its highest EC50 (18.1 μg L−1 of Ag+, Table 1). The much higher values of EC50 for AgNPs relative to Ag+ recorded in this study for all species are consistent with results of other studies. Thus, 50 % growth inhibition of Phaeodactylum tricornutum (marine diatom) was at 400 ± 110 μg L−1 of Ag+ and at 2380 ± 1880 μg L−1 of AgNPs (Angel et al. 2013). The toxic effects of both forms of Ag in the present study were mostly pronounced at shorter exposure times. After 2 days of culturing, the microorganisms do not produce significant amounts of biomass and DOM, and thus they are directly affected by the adsorption of Ag onto the cell walls and intracellular Ag assimilation via the production of ROS causing cellular damage. On the other hand, with a longer exposure time and at lower concentrations of silver, the metal becomes complexed with DOM, allowing the growth of microorganisms. Similar mechanisms have been suggested for the alleviation of high copper levels in cultures of marine diatom (P. tricornutum) and marine green algae (Dunaliella tertiolecta) (López et al. 2015; Rico et al. 2013). Consequently, the duration of metal exposure may be an important factor in interpreting the resistance of microorganisms to silver. Ecotoxicological assays are often performed at short exposure times. If less than 100 % of the cells are impacted during the first 2 days of exposure, the resistant cells fully recover and after several generation times, the difference in cell number is no more observable. During biofilm growth in the presence of Ag, the number, type and amount of organic ligands are closely related to the concentration of Ag in solution (González et al. 2015). Therefore, the duration of exposure to metal stress is a fundamental parameter controlling the bioaccumulation and detoxification of microorganisms.
For 2 days of exposure, N. palea showed the highest resistance to Ag+ stress, whereas for 5 and 14 days of exposure, Leptolyngbya sp. was the most resistant. It is possible that the intrinsic properties of the studied microorganisms were responsible for such a contrasting behaviour of diatoms and cyanobacteria. The freshwater periphytic diatoms possess thick siliceous frustule with a relatively thin EPS layer (Gélabert et al. 2004). The diffusion of toxic Ag+ through the mineral frustule should be slow and thus, despite the low biomass and high accumulation of Ag, the metal becomes less toxic to diatoms compared to other species. With time, the cyanobacteria and green algae develop sufficient biomass with thick EPS capsules and sheaths (González et al. 2014). This helps external layers, virtually absent in diatoms, to efficiently protect green algae and cyanobacteria from metal stress. The diatoms, due to slower growth and eventual diffusion of Ag+ through the frustule, finally become more vulnerable to aqueous Ag. In this regard, 2 weeks rather than 2 days should be taken as a favourable period for assessing the Ag toxicity to phototrophic benthic microorganisms.
Influence of silver speciation in solution on its toxicity
Metal speciation is a key factor controlling its toxicity in natural waters. In the course of the experiments at hand, all solutions remained undersaturated with respect to amorphous phases of Ag+ that are likely to precipitate from aqueous solutions such as AgOH(am), AgOCl(s) and AgCl(s). Usually, Ag toxicity is considered to be directly linked to free Ag+ concentration in solution (Luoma et al. 1995). In addition, AgCl°(aq) complexes can be adsorbed to the cell’s surfaces rendering the microorganism as a vector for AgCl transfer to other trophic levels (Baker et al. 2014). Neutral species AgCl°(aq) were present in the experimental solution at concentrations higher than that of Ag+ (Fig. S4b). These chloro-complexes may be directly available to the microorganisms being toxic to the cells in the form of Ag+ and AgCl° (Fig. S5). According to the mass balance calculations performed at Ag total concentration equalled to 1 % of added AgNPs, the free Ag concentration decreased from 5 to 8 % in the circumneutral pH range during DOC concentration rise from 5 to 15 mg L−1. Under these conditions, the concentration of Ag-DOM ranged between 3 and 10 % for the same range of DOC. This result reflects that in the case of AgNPs where the concentration of released Ag+ is quite low, the toxic effect may become low due to organic ligand complexation. In contrast Ag+-bearing experiments, when N. palea was exposed to AgNPs in the culture media and NaNO3, there was no difference in the growth inhibition effect compared to the AgNP-free control (Fig. 2b). In this case, a relative difference between the growth in the control and in 1000 μg L−1 AgNP solution was 8 and 10 % for 0.01 M NaNO3 solution and Combo culture medium, respectively.
The difference in cell response to Ag+ concentration after 2 and 14 days strongly suggests the low importance of AgCl(am) precipitation, AgNPs agglomeration and Ag adsorption onto the walls of the wells. These metal adsorption/removal processes are usually very fast, on the order of minutes to several hours and thus they should not be pronounced at longer exposure periods in the present study.
The interactions of the metal with dissolved organic ligands have a direct influence on silver speciation and on the metal bioavailability for the microorganisms (Sikora and Stevenson 1988). Miao et al. (2009) found that AgNPs reduce the growth rate of the diatoms Thalassiosira weissflogii, affecting the efficiency of photosystem II and chlorophyll-a through the release of Ag+. However, these authors observed that the exopolysaccharides (EPS) production increased in the presence of AgNPs, thus providing an additional detoxification mechanism of this microorganism. Because EC50 increased for all three studied species as a function of exposure time (Table 1), the silver complexation with dissolved organic matter was mostly pronounced at lower concentrations of metal, long exposure time, and highest concentration of DOC (Figs. 2, 3, 4). The ratio DOC/wet biomass increased when inhibitory effects were higher, thus illustrating a strong protective role of DOC as a complexing agent for silver via reducing its bioavailability.
The toxicity of Ag+ and AgNPs was studied for three benthic phototrophic microorganisms, representing three main taxa of the periphyton biofilm, diatoms, green algae and cyanobacteria. All three microorganisms were severely affected by Ag+ at a significantly lower concentration than AgNPs. The EC50 (Ag concentration where 50 % of growth inhibition was achieved), suggests that Ag+ was significantly more toxic at concentrations higher than 4.2 μg L−1 after 2 days of exposure, whereas AgNPs became toxic at concentrations higher than 500 μg L−1. This EC50 increased with exposure time; the highest value for Ag+ was 394 μg L−1 after 14 days of exposure, whilst for AgNPs, this value never decreased below 500 μg L−1. This difference can be explained by slow dissolution of AgNPs producing toxic Ag+, electrochemical repulsion of negatively charged nanoparticles and microbial cells and the release of DOC capable of complexing a low amount of ROS-produced Ag+ by the microorganisms. At the beginning of exposure, the microorganisms do not produce a stable biofilm, which would protect the cells from Ag+(aq). For longer exposure time, the cells release DOC and EPS and the ionic silver becomes strongly complexed with organic ligands, both in solution and in the vicinity of the cells, thus allowing the growth of microorganisms in the form of biofilm.
This research was supported by the Midi-Pyrénées Regional Council (France) within the programme Gagilau (No. DAER-R93 90173). Partial support from BIO-GEO-CLIM grant No. 14.B25.31.0001 and ANR CITTOXIC-Nano are also acknowledged. Finally, we thank Katrin Meier for the English revision of the manuscript.
- Barnett BP, Arepally A, Karmarkar PV, Qian D, Gilson WD, Walczak P, Howland V, Lawler L, Lauzon C, Kraitchman DL, Bulte JWM (2007) Magnetic resonance-guided, real-time targeted delivery and imaging of magnetocapsules immunoprotecting pancreatic islet cells. Nat Med 13:986–991. doi:10.1038/nm1581 CrossRefGoogle Scholar
- Gélabert A, Pokrovsky OS, Schott J, Boudou A, Feurtet-Mazel A, Mielczarski J, Mielczarski E, Mesmer-Dudons N, Spalla O (2004) Study of diatoms/aqueous solution interface. I. Acid-base equilibria and spectroscopic observation of freshwater and marine species. Geochim Cosmochim Acta 68:4039–4058. doi:10.1016/j.gca.2004.01.011 CrossRefGoogle Scholar
- Gustafsson JP (2012) Visual MINTEQ, ver. 3.0. Compiled in visual basic. NET 2005. KYH, Dept of Land and Water Resources Engineering, StockholmGoogle Scholar
- Ivask A, Kurvet I, Kasemets K, Blinova I, Aruoja V, Suppi S, Vija H, Käkinen A, Titma T, Heinlaan M, Visnapuu M, Koller D, Kisand V, Kahru A (2014) Size-dependent toxicity of silver nanoparticles to bacteria, yeast, algae, crustaceans and mammalian cells in vitro. PLoS One 9:e102108. doi:10.1371/journal.pone.0102108 CrossRefGoogle Scholar
- Kroll A, Matzke M, Rybicki M, Obert-Rauser P, Burkart C, Jurkschat K, Verweij R, Sgier L, Jungmann D, Backhaus T, Svendsen C (2016) Mixed messages from benthic microbial communities exposed to nanoparticulate and ionic silver: 3D structure picks up nano-specific effects, while EPS and traditional endpoints indicate a concentration-dependent impact of silver ions. Environ Sci Pollut Res 23:4218–4234. doi:10.1007/s11356-015-4887-7 CrossRefGoogle Scholar
- Lodeiro P, Achterberg EP, Pampín J, Affatati A, El-Shahawi MS (2016) Silver nanoparticles coated with natural polysaccharides as models to study AgNP aggregation kinetics using UV-visible spectrophotometry upon discharge in complex environments. Sci Total Environ 539:7–16. doi:10.1016/j.scitotenv.2015.08.115 CrossRefGoogle Scholar
- Ribeiro FV (2014) Processos extremos na constituição da cidade. da crise à emergência nos espaços mundializados 2014. 18:2 doi:10.11606/issn.2179–0892.geousp.2014.81105Google Scholar
- Stanier R, Kunisawa R, Mandel M, Cohen-Bazire G (1971) Purification and properties of unicellular blue-green algae (order Chroococcales). Bacteriol Rev 35:171Google Scholar
- Tolaymat TM, El Badawy AM, Genaidy A, Scheckel KG, Luxton TP, Suidan M (2010) An evidence-based environmental perspective of manufactured silver nanoparticle in syntheses and applications: a systematic review and critical appraisal of peer-reviewed scientific papers. Sci Total Environ 408:999–1006CrossRefGoogle Scholar
- van Schaik JWJ (2008) Binding of metals to macromolecular organic acids in natural waters vol 2008.Google Scholar