The impact of engineered cobalt, iron, nickel and silver nanoparticles on soil bacterial diversity under field conditions

Our understanding of how engineered nanoparticles (NPs) migrate through soil and affect microbial communities is scarce. In the current study we examined how metal NPs, including those from the iron triad (iron, cobalt and nickel), moved through pots of soil maintained under winter field conditions for 50 days, when mesophilic bacteria may not be dividing. Based on total metal analysis, cobalt and nickel were localized in the top layer of soil, even after exposure to high precipitation and freeze–thaw cycles. In contrast, a bimodal distribution of silver was observed. Due to high endogenous levels of iron, the migration pattern of these NPs could not be determined. Pyrosequence analysis of the bacterial communities revealed that there was no significant engineered NP-mediated decline in microbial richness. However, analysis of individual genera showed that Sphingomonas and Lysobacter were represented by fewer sequences in horizons containing elevated metal levels whereas there was an increase in the numbers of Flavobacterium and Niastella. Collectively, the results indicate that along with the differential migration behavior of NPs in the soil matrix, their impact on soil bacterial diversity appears to be dependent on environmental parameters.


Introduction
The class of engineered nanoparticles (NPs) is one of the most promising material science developments in recent times. NPs find practical applications in many fields of industry and daily life, with an expected market value for nano-products in 2011 to 2015 of around $1 trillion yearly [1]. We already use NPs in electronics, materials engineering, food, transportation, cosmetics, energy, pharmaceuticals, biomedical products, agriculture, fishing, manufacturing, remediation, security, and consumer goods [2,3], and by 2014, more than 15% of all marketed products are estimated to have some Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI. kind of nanotechnology incorporated into their manufacturing process [4]. This will inextricably lead to an increase in their release into the environment. From an environmental perspective, it is important to ascertain that upon release into soil, NPs will not be toxic to microorganisms. Any change in the microbial community can influence the soil quality and health with microorganisms involved in decomposition, geochemical and nutrient cycling, plant protection and symbiosis, and bioremediation. Good soil quality within natural or managed ecosystems is essential to sustain plant and animal production, maintain adequate water and air quality, and support human health and habitation [5,6].
There are only few studies reported on the influence of metal NPs on soil microbial diversity. While magnetic Fe 2 O 3 NPs stimulated the growth of actinobacteria in soil, Ag NPs, TiO 2 and ZnO NPs significantly reduced the overall soil microbial biomass [7][8][9][10]. Other studies have shown that CdSe/ZnS quantum dots, gold nanorods, TiSiO 4 , and Fe/Co magnetic fluid NPs altered soil communities [11]. In particular, nitrogen fixers appeared susceptible to TiO 2 , ZnO, CeO 2 and ZnO NPs [9,10,12]. As well, our earlier studies similarly showed that the organisms involved in the nitrogen cycle seemed to be particularly sensitive to metal NPs [13][14][15], arguing that it is important to continue to investigate the impact of these emerging contaminants on soils.
In our previous field study on zero valent Cu and ZnO NPs, we had been careful not only to investigate the effects of metallic NPs on soil chemistry and biota, but also considered migration through the soil matrix and the parallel leaching of ions when evaluating the results [15]. The microbial community was indeed altered by the NPs/metal ions as the particles were transported through the soil. Here we have focused our interest on metallic NPs from the iron triad (iron, cobalt and nickel), so called because they are located beside one another on the periodic chart. The iron triad share properties associated with their unpaired electrons including their strong magnetic pole and their utility in industrial processes due to their ready association with carbon. In addition, we have also used a non-iron triad metal particle, Ag NP, for comparison. Thus, here we investigate the influence of these four metal NPs on soil bacterial diversity under field conditions. Based on their chemical properties, our hypothesis was that the iron triad metal NPs would react more strongly to the soil matrix and thus possibly move through the soil more slowly than the Ag NPs. Thus, we considered it possible that these NPs would have a greater impact on soil diversity at least in the upper layers than would the Ag NPs.

Nanoparticles
NPs were purchased from Sun Innovations, USA in powder form and had no surface coating. The properties of the NPs were provided by the manufacturer and the particles were used as received. Cobalt NPs (Co NPs) were 99.8% pure, with average size of 28 nm and size range of 2-60 nm. Zero valent iron NPs (Fe NPs) were 99.9% pure, with a size range of 2-58 nm (average 25 nm). Nickel NPs (Ni NPs) were 99.9% pure, with average size of 20 nm and. Silver NPs (Ag NPs) were 99.8% pure, with a size range of 2-50 nm (average 35 nm). All NPs were spherical in shape.

Experimental design
The experimental set up was similar to the one previously described [15]. Briefly, the soil was obtained from an agricultural soil-processing center in New York (located at 40.76 • N, 73.27 • W) and debris (plant matter, rocks, and wood chips) were as removed manually. Plastic pots (23.6 cm × 26.7 cm; Ames True Temper) with a rolled rim and a bottom saucer were filled with soil to a height of 20 cm and transported to the field site located at 40.72 • N, 73.09 • W. The soil was allowed to stabilize for seven days and 550 mg of the appropriate NPs were sprinkled on the soil surface. The NPs were distributed only in a 5 cm diameter area at the center of the soil surface. Control pots received no NPs. After 50 days outside (Jan 19-March 9), the pots were transported to the lab and a core sample from the center was removed using poly vinyl chloride (PVC) pipe (5 cm diameter). The recovered soil was separated into three samples based on the depth: the soil surface to 2.5 cm depth was identified as the top horizon, the mid horizon was from 2.5 to 7.6 cm, and the bottom horizon was from 7.6 to 20 cm. Metal analysis was carried out as previously described [15]. Soils from each horizon were homogenized prior to sampling for molecular studies.

DNA isolation and 454-pyrosequencing
Soil samples were first treated with ethidium monoazide (EMA) using the protocol described in the literature [16]. This allows the preferential amplification of viable bacteria. DNA was extracted from the soil using PowerSoil TM DNA isolation kits (MO BIO Laboratories Inc, Carlsbad, CA). Bacterial tag-encoded FLX amplicon pyrosequencing (bTE-FAP) were performed as described previously using Gray28F 5 TTTGATCNTGGCTCAG and Gray519r 5 GTNTTAC-NGCGGCKGCTG primers [17]. Amplifications consisting of one-step polymerase chain reaction (PCR) of 30 cycles, were accomplished using a mixture of Hot Start and HotStar high fidelity Taq polymerases. Amplicons originating and extending from the primers were used for initial generation of the sequencing library. Pyrosequencing analyses utilized Roche 454 FLX instrument with Titanium reagents and Titanium procedures and were performed at the Research and Testing Laboratory (Lubbock, TX) based upon RTL protocols (www.researchandtesting.com). After sequencing, all failed sequence reads, low quality sequence ends, tags and primers as well as any non-bacterial ribosome sequences and chimeras were removed using B2C2 [18] as previously described [17]. To curate the short (<150 bp) reads, sequences with ambiguous base calls, and sequences with relatively long homopolymers (>6 bp) were also removed. To determine the identity of bacteria in the remaining sequences, sequences were denoised, assembled into OUT clusters (96.5% identity), and queried using a distributed .NET algorithm via Blastn+ (KrakenBLAST www.krakenblast.com) against a database of high quality 16S rRNA bacterial gene sequences. Using a .NET and C# analysis pipeline, the resulting BLASTn+ outputs were compiled and data reduction analysis performed as described previously [19].
Based upon the above BLASTn+ derived sequence identities (pe rcent of total length query sequence which aligns with a given database sequence), the bacteria were classified at the appropriate taxonomic levels as previously described [15]. The percentage of sequences assigned to each bacterial phylogenetic level were individually analyzed for each pooled sample providing relative abundance information within and among the individual samples, based upon relative numbers of reads within each [18].

Statistical analysis
DNA isolation was performed in duplicate for each sample and the two isolated DNA samples were pooled prior to pyrosequencing. To ensure that minor changes in abundance of Genus that are present in lower percentage did not influence the results, only those present at ≥1% abundance were used in the statistical analysis. As the assumption that the data being analyzed would be normally distributed may not always hold true, nonparametric statistical analysis were performed. Wilcoxon matched pair test was used to compare the pyrosequencing data between untreated control versus treated samples. A value of ≤0.05 indicates a significant difference between the compared values. Analysis was performed using Statistica (Release 8.0) software.

Results and discussion
The agricultural soil used is rich in organic matter with total organic carbon (TOC) and total Kjeldahl nitrogen (TKN) of 13 100 mg kg −1 dry weight and 980 mg kg −1 dry weight, respectively. The pH of the soil was 8.0. During the 50 day winter period, the temperature ranged from −10 to 10 • C with a total precipitation of 166 mm (figure 1). Due to soil freezing and thawing, deep cracks were observed at the conclusion of the incubation period. We considered it possible that some NPs would travel with melted water through the cracks into the lower horizon soil. Total metal analysis carried out for top, mid, and bottom horizons allowed an assessment of the migration of the added metal irrespective of its form as NPs or ions. Despite the presence of cracks in the soil, in accordance to our hypothesis, relative to control soils there was a 650-fold and a 1000-fold higher concentration of total Co and Ni, respectively in the top horizon (table 1). Smaller quantities (∼4-fold for Co and ∼2-fold for Ni over control levels) migrated to the lower layers. Ag appeared to have a bimodal distribution and was mostly found in the top and bottom layer only, with an increase relative to control levels of 100-fold and a 79-fold, respectively (table 1). Ag did not appear to remain in at the mid horizon (1.2-fold increase). We were unable to distinguish the addition of Fe to the soils since the endogenous levels of this metal were already high (table 1). There was no change in the TOC, TKN or pH levels in any of the soil samples analyzed post-incubation (not shown).
While the current study does not provide information on the ionization state of metal in different layers, it is evident that the metal transport through soil under field conditions obviously varies depending upon the metal. Two types of distribution curves of the NPs within the three layers were observed. Co and Ni NPs remained localized in the top layer of soil. In contrast, Ag NPs showed a bimodal distribution of metal and similar to previous reports that noted that NPs on the top horizon appear to act as a source for Ag+ ions, which relocate with heavy rainfall. Since we did not detect elevated levels of Ag in the middle horizon, this suggests a fast migration of the metal through the soil matrix. Notwithstanding our inability to follow Fe, these results are consistent with our hypothesis that metals of the iron triad would be more reactive and thus be more likely retained in the top horizon compared to less reactive Ag.
To evaluate the impact of NPs on the bacterial community, we carried out pyrosequencing having first treated the soil so as to decrease the proportion of DNA derived from dead or dying cells. Control soil samples showed rich diversity with species richness ranging from 43 to 49 for all three layers of soil (table 2). Upon incubation with engineered NPs, there was no major decline in the species richness. In the top horizon of soils treated with Ag and Ni NPs, richness in fact increased by 16%. However, there was a marginal decrease (∼10%) in richness in the middle and bottom horizons of the Ag NP-treated soil (table 2). To compare the community composition between the control soil samples and samples exposed to NPs, a Wilcoxon matched pair test was used. Exposure to engineered NPs resulted in no statistically significant difference in the population composition across all the three horizons ( p > 0.05, table 2). This implies that the metal NPs did not result in acute changes in this soil microbial community, and thus there was no obvious association of the presence of NPs and the distribution of genera, under these conditions and at the tested concentrations.
We believe that in order to appreciate the environmental behavior of all the NPs currently being produced, understanding the relationship between the changes in chemical properties of NPs as a function of time, environmental parameters, and the rate of transport is highly critical. Based on this knowledge, we may be able to group NPs based on their similarity and study in detail one representative particle from each group. Without such approach, the innovation aspect of nanotechnology will always be a couple of steps ahead of our understanding of their environmental impacts.