Environ. Sci. Technol. 2008, 42, 728–733
Fate of Prions in Soil: Interactions of RecPrP with Organic Matter of Soil Aggregates as Revealed by LTA-PAS A M A R A N T A P U C C I , * ,†,‡ L U I G I P A O L O D’ACQUI,‡ AND LUCA CALAMAI† Dipartimento di Scienza del Suolo e Nutrizione della Pianta, Università degli Studi di Firenze, Piazzale delle Cascine 28, 50144 Firenze, Italy, and Istituto per lo Studio degli Ecosistemi, CNR Via Madonna del Piano 10, 50019 Sesto Fiorentino (FI), Italy
Received June 4, 2007. Revised manuscript received November 9, 2007. Accepted November 13, 2007.
The contribution of soil organic matter (OM) to the adsorption of a recombinant prion protein (RecPrP) was studied in microcosm systems (soil aggregates from two different soils) before and after OM removal by low temperature ashing (LTA). The LTA technique allows a controlled removal of OM layer by layer, like a peeling of an onion skin, with minimal disturbance of the mineral matrix. Soil aggregates were selected as a representative model of the “in situ” conditions. Adsorption from batch vs percolation experiments were compared, and the aggregates were characterized by photoacustic Fouriertransform IR spectroscopy (PAS-FTIR). High affinity (H-type) adsorption isotherms were found with complete removal of RecPrP from solution for protein/soil ratios up to 1:62.5. OM removal from aggregates decreased the adsorbed RecPrP in amounts corresponding to 330–1000 µg mg-1 of soil organic carbon (OC) indicating that native OM has specific adsorption capacity comparable and/or superior to the mineral matrix. The coupled LTA-PAS-FTIR approach demonstrated that, albeit OM composition was homogeneous throughout the aggregates, its presence in the most external surfaces of the aggregates affects the diffusion dynamics of RecPrP within the aggregates during percolation.
Introduction Prion proteins (PrPs) have recently gain attention as they are indicated as the infectious agents in transmissible spongiform encephalopathies (TSE), according to Prusiner’s protein-only theory (1, 2). Environmental concern regarding the fate and the potential dissemination of prions requires knowledge on their degradation and direct evidence of the role of soil and of its constituents in the persistence of PrPs. Research carried out on prions adsorption related to soil (3–5) has been conducted either with the unglicosilated recombinant ovine (recPrP) or the infective “scrapie” form (PrPsc) of the proteins and has focused either on the clay fraction alone (6), on selected inorganic constituents (7, 8) * Corresponding author fax: +39 05552225920;
[email protected]. † Università degli Studi di Firenze. ‡ Istituto per lo Studio degli Ecosistemi. 728
9
e-mail:
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 3, 2008
or on the soil as a whole (9–12). To date, the persistence and/or of the retention of infectivity of prion proteins in soil was demonstrated (8, 10, 12–16) along with a different contribution of the soil constituents in fixing the prion. Rigou et al. (6) found that at least 95% of RecPrP was retained by homoionic Wyoming montmorillonite for protein/clay ratios varying from 500:1 to 1:1. Comparable maximal adsorbed amounts were measured onto mica sheets by Vasina et al. (7) who identified the electrostatic attraction between the protein and highly negatively charged mica surfaces as the main interaction mechanism. Johnson et al. (11) ascertained that either clay minerals and quartz were able to bind scrapie prion protein (PrPSc) from syrian hamster brain homogenate albeit montmorillonite and quarz had larger specific binding capacities in comparison with kaolinite. However, some discrepancies emerged when comparing the total retained PrPSc amounts of four soils differing in texture and mineralogical composition, if the specific adsorption of the individual inorganic constituents were considered. A competition among various adsorbing domains, and/or a hampered access to the montmorillonite surface sites by organic matter (OM) and/or mineral oxides was generally suggested to account for the different behavior while a specific sorption capacity of OM was generally poorly shown (8, 10, 14–16). The specific sorption properties of soil OM with respect to PrPs have been poorly investigated so far. Rao et al. (17), demonstrated an irreversible interaction with both synthetic humic-like and humic–mineral complexes (catechol polymers and birnessite–catechol polymers mixtures) and fulllength RecPrP. However, a higher level of complexity was obtained when RecPrP was copolymerized within the humiclike complexes, suggesting that PrPs should be strongly retained in soils rich in OM, with very low risks of release and subsequent dissemination away from the initial application site. However, no direct data on the specific sorption capacity (µg mg-1) of humic-like complexes were given. This literature points out the relevance of studying the adsorption properties of OM either alone and/or as a function of its interaction with other soil constituents. Generally, the studies on soil OM are carried out mainly using chemical extraction methods which modify both organic substances and mineral phases (18) resulting in possible artifacts when sorption (or interaction) properties of the purified fractions are investigated. Recently, a low-temperature ashing (LTA) by oxygen plasma has been used for removing soil OM from the mineral matrix of soil aggregates with minimal disturbance and damage to the inorganic constituents (19, 20). This technique allows a controlled oxidation, at low temperature, of organic substances on undisturbed specimens by removing organic material, layer by layer, without altering or damaging the residual layers, like peeling onion skins. In a LTA approach to the “in situ” study of OM, other analytical techniques are applied to the specimens before and after LTA treatment such as elemental analysis and photoacoustic Fourier transform IR spectroscopy (PAS-FTIR), allowing the spectroscopic characterization of the sample surfaces and providing insights on the nature of OM on the surface of organo-mineral soil constituents. In this work, undisturbed soil aggregates were used as a simplified microcosm model system of soil at laboratory scale, and considered as individual adsorption system rather than the summation of single constituents. In fact, soils are not simply mixtures of discrete organic and mineral phases but they are organized into aggregates of organo-mineral complexes (21, 22). Therefore, the adsorption process “in situ” is a complex phenomenon due only partly to the 10.1021/es071314q CCC: $40.75
2008 American Chemical Society
Published on Web 01/03/2008
properties of single components but, also due to properties resulting from of the interaction among constituents. The adsorption of a recombinant purified protein whose structure mimics the ovine prion (RecPrP) was studied on native OM of undisturbed soil aggregates. This model protein was chosen as a first step approach to the study of the behavior of infective ovine PrPsc as these proteins shares the same primary structure. Further future investigation with the use of a glycosylated RecPrP, and finally, to PrPsc might help elucidate the role of the glycosylation, and of further structural levels (secondary and tertiary) structure, to the adsorption. The experiments were performed on undisturbed and LTA-treated aggregates (where OM was removed) to assess the contribution on native OM to RecPrP adsorption. A further step toward the “in situ” condition was achieved by comparing the adsorption at equilibrium, vs that obtained with percolation.
Materials and Methods Model Protein. A purified full-length ovine ARQ genetic variant (Mr 23 kDa, residues 23–234) prion protein (RecPrP) (23), was supplied by the Virologie et Immunologie Moléculaires, INRA (Jouy-en-Josas, France) within the European Union research project “TSE-Soil Fate”. The RecPrP concentration was determined using Sigma’s P5656 Lowry essay kit based on Peterson’s modification (24, 25). Preliminary comparative experiments verified that this modified Lowry essay was superior over Bradford (26–28), Bicinconinic acid (29) and UV absorption at 280 nm (30) with respect to interferences. All spectrophotometric measurements at 700 nm were done by a Perkin-Elmer Lambda 3B UV–vis spectrophotometer with disposable polystyrene cuvettes (optical path 1 cm). A calibration curve was obtained with Bovine Serum Albumine (BSA) solutions in the concentration range 50–400 µg mL-1, and therefore the RecPrP concentrations were expressed as BSA equivalents. Soil Aggregates. Undisturbed aggregates (gently broken at field-moisture and sieved at 0.5–1 mm) were collected from the surface horizon of a sandy soil (SS) from central Scotland and a loamy soil (CM) from Campo Migliaio, (Tuscany, Italy) and air-dried. These soils were selected and characterized as reference soil samples within the TSE-Soil Fate research project. Soil particle-size distribution and pH measurements were carried out according of SISS methods (31). Total organic carbon (OC) and nitrogen (N) content were determined by a Carlo Erba (Milan, Italy) NA 1500 CHNS Analyzer (SI Table S1). LTA Technique. The LTA procedure was performed with a self-assembled apparatus as described by D’Acqui et al. (14). The samples were placed in the LTA reactor and evacuated to 45 Pa under an oxygen a 20 mL min-1 flow rate. The generation of oxygen plasma was obtained with a radiofrequency of 13.56 MHz using a power input of 100 W and a reflected power of 5 W. In these conditions the surface temperature was maintained below 100 °C (32). Soil OM Removal. In pilot experiments, aliquots of 100 mg of sample were exposed to LTA for different times. The treatment was stopped after 5, 12, and 24 h to withdraw a 30 mg aliquot for elemental analysis. No further C and N removal was obtained beyond 24 h at the selected LTA conditions (19). Total C content was measured at each step to assess the dynamics and the extent of C and N removal. In these conditions LTA treatment had different efficiency depending on specimen shape, dimension, porosity, and on the amount of sample subjected to the treatment. Therefore, to obtain comparable results, it was decided to treat 100 mg sample aliquots for 5 or 24 h for each specimen.
Batch Adsorption Experiments of RecPrP with Untreated or LTA-Treated Aggregates. Batch adsorption isotherms of RecPrP were determined at 25 °C on CM and SS aggregates before and after LTA treatment (LTACM and LTASS, respectively). Aliquots of RecPrP concentrated solution (such to yield final concentrations from 0.3 mg mL-1 to 1 mg mL-1) were added to double-distilled H2O to a final volume of 1.5 mL, mixed with 40 mg of soil aggregates, and agitated in a end-overend shaker at 15 rev min-1 for 24 h. This time was largely sufficient to reach equilibrium as established in preliminary kinetic experiments (Figure S1 in Supporting Information (SI)). Protein essays with aggregates without protein or with laboratory-ware in the presence of RecPrP run contextually as controls ruled out any interferencies or adsorption other than by the aggregates. Percolation Adsorption Experiments of RecPrP on Aggregates. The aggregates soaking in the RecPrP solutions and the prolonged shaking required by the batch technique resulted in an almost complete aggregates breakdown. This fragmentation is unlikely to occur in soils below the surface, where a gentle aggregate wetting generally occurs. The adsorption experiments were then performed also with undisturbed soil aggregates by a percolation method to simulate the in situ conditions. Due to the limited availability of RecPrP the comparisons among the untreated and LTAtreated specimens, at all the initial concentrations used for the adsorption isotherms, were not possible. Therefore, only the concentrations corresponding to the plateau of the adsorption isotherms (for either untreated or LTA-treated aggregates) were used. Forty mg of aggregates were placed in Pasteur pipettes and 1.5 mL of RecPrP solution (0.5 mg mL-1 for SS and LTASS and 0.75 mg mL-1 for CM and LTACM) was slowly added from the top and let percolate through the aggregates. Since LTACM and LTASS aggregates were partially dehydrated during LTA-treatment, an initial prewetting of all aggregates was done to have the same starting conditions on both untreated and LTA-treated specimens. The leachate solutions were recovered and again added to the pipettes from the top. This step was repeated three times which was necessary to have no further depletion of RecPrP from solution occurred. The RecPrP concentration was determined in the final leachates. The experiments were done in quadruplicate and repeated twice. The data were expressed as means ( standard errors. Specific Surface Area and Pore Specific Volume. Specific surface area (SSA) of aggregates before and after soil OM removal was estimated using the BET equation after multiplestep adsorption–desorption of N2 at 77°K (27) with a 1990 Carlo Erba Instrument (Italy). Enough sample material (0.5–2 g) was degassed at 70 °C for 12 h and used for measurement to ensure a total surface area >5 m2. Pore specific volume (PSV) was calculated at partial pressure of 0.95 (pores
