Soil and Biota of Serpentine: A World View 2009 Northeastern Naturalist 16(Special Issue 5):215–222 Using Chelator-buffered Nutrient Solutions to Limit Ni Phytoavailability to the Ni-Hyperaccumulator Alyssum murale Rufus L. Chaney1, Guido Fellet2, Ramon Torres3, Tiziana Centofanti1, Carrie E. Green1, and Luca Marchiol2 Abstract - Nickel (Ni) is essential for all plants due to its role in urease activation. Demonstration of Ni essentiality has required exceptional effort to purify nutrient solutions to remove Ni; thus, an improved technique would make study of Ni deficiency more available to diverse researchers. As part of our research on Ni hyperaccumulation by plants, we developed chelator-buffered nutrient solutions with very low buffered activity of free Ni2+, and tested growth of Alyssum murale (Goldentuft Madwort), A. corsicum (Madwort), A. montanum (Mountain Alyssum) and Lycopersicon esculentum (Tomato). We used a modified Hoagland nutrient solution with 2 mM Mg and 1 mM Ca to simulate serpentine soil solutions. We could use hydroxyethyl-ethylene-diaminetriacetate (HEDTA) to achieve Ni2+ activity levels as low as 10-16 M, and cyclohexane-ethylenediamine-tetraacetate (CDTA) to supply higher activities of buffered Ni2+ compared with HEDTA; however, we were unable to obtain proof of induced Ni-deficiency, even with urea-N supply in a 6-week growth period, apparently because seeds supplied enough Ni for growth. Yields were somewhat reduced at lower Ni activity by the end of the test period, but strong deficiency symptoms did not occur, apparently due to the supply of Ni from hyperaccumulator species seeds (contained 7000–9000 mg Ni kg-1). Chelator buffering supplied controlled levels of Ni2+ for all test species; very low plant Ni levels were attained when seed Ni was low. Reaching clear and strong Ni deficiency appears to require longer growing periods, using seed with exceedingly low initial endogenous Ni, or species possessing higher Ni requirements. Introduction Nickel is an essential micronutrient for higher plants (Brown et al. 1987). One key function is activation of urease for hydrolysis of urea and ureides for release of NH3 (Eskew et al. 1983, Welch 1981); however, these researchers found no evidence of field deficiency of Ni. Because urea accumulates in Ni-deficient tissues, Ni deficiency can result in toxic accumulation of urea within plant tissues. More recently, Wood et al. (2005) showed that severe Ni deficiency of Carya illinoinensis (Wangenh.) K. Koch (Pecan) on Georgia Coastal Plain soils could kill trees. Ruter (2005) found economically significant Ni deficiency in Betula nigra L. (River Birch) grown in low-Ni potting media, and Wood et al. (2006b) report observations of potential 1USDA-Agricultural Research Service, Environmental Management and Byproduct Utilization Laboratory, Beltsville, MD, USA. 2Department of Agronomy and Environmental Sciences, Udine University, Udine, Italy. 3University of Puerto Rico, Mayaguez, PR, USA.*Corresponding author - Rufus.Chaney@ars.usda.gov. 216 Northeastern Naturalist Vol. 16, Special Issue 5 Ni deficiency in other crop plants on low Ni soils. Wood et al. (2006a) reported that ground biomass and water extracts of biomass of Ni hyperaccumulators such as Alyssum murale Waldst & Kit. (Goldentuft Madwort) were effective Ni fertilizers to prevent deficiency in Pecan. Certain Alyssum spp. can accumulate higher Ni concentrations than most plant species. This fact causes us to question whether their Ni requirement is higher than non-hyperaccumulators. For example, Li et al. (1995) showed that growth of Thlaspi caerulescens J&C Pres. (Alpine Pennycress) was reduced when foliar Zn was below 100 mg kg-1 dry weight (DW), while most higher plant species do not show Zn deficiency until leaf Zn falls below 12–15 mg kg-1 DW. This finding indicates that similar results might occur for Ni in regards to Ni accumulators and non-accumulators. Additionally, because we used chelator-buffering techniques to induce Zn deficiency, we demonstrated that T. caerulescens requires Zn2+ concentration in solution of 10-7 mol L-1 compared to Brassica oleraceae L. (Cabbage) and Thlaspi arvense L. (Field Pennycress), which required 10-11.4 mol Zn2+ L-1. In chelator-buffering, a metal chelator is used at higher concentrations than the sum of strongly chelated microelement cations. For dicots, it is easy to supply low but adequate levels of buffered micronutrient cations using several chelating agents (Parker et al. 1995). The theory for using chelatorbuffering is described in a review paper by Chaney et al. (1989). In addition to use with higher plants, chelator-buffering has also been used to support growth of soil microbes (Angle et al. 1991, Buyer et al. 1989). An important consideration when using metal chelators in nutrient solutions is to maintain soluble Fe. A ferric-specific chelator such as FeHBED prevents other added metals from displacing Fe (Chaney 1988). When normal Murashige and Skoog medium is used, Fe from FeEDTA (Fe3+-ethylenediaminetetraacetate) precipitates and the activity of trace element cations is much lower than the total amount present. Thus, because it seems possible that Ni hyperaccumulator species might require higher levels of Ni than normal plant species, we developed chelator-buffered nutrient solutions to supply low Ni2+ activity, and compare responses of two hyperaccumulators species of the Alyssum genus (A. murale and A. corsicum Duby [Madwort] and two non-hyperaccumulators (A. montanum L. [Mountain Alyssum] and Lycopersicon esculentum Mill. [Tomato]) when grown with different levels of Ni2+ activity, in solutions containing urea as a nitrogen source, in an attempt to define their Ni requirement. Experimental Methods The experiments tested the use of chelator-buffering to control Ni2+ activity for examining Ni requirements and response of Alyssum spp. and tomato to Ni activity while growing plants in hydroponics at very low, low, and moderate Ni2+ activities under controlled conditions (temperature: 24 °Cday /20 °Cnight; humidity: 70%day, 50%night; PAR: 300 μmol s-1m-2, 16 h day-1). 2009 R.L. Chaney, G. Fellet, R. Torres, T. Centofanti, C.E. Green, and L. Marchiol 217 Figure 1 shows the relationship between total solution Ni and Ni2+ activity in the HEDTA- and CDTA-buffered nutrient solutions. Seeds of test plants were germinated in vermiculite irrigated with macronutrient solution (<1 μg Ni L-1) until plants were large enough to use in experiments. Four-week-old seedlings were transferred into the hydroponics solution, without Ni or urea, to allow the root system to recover from possible transplant damages. After two weeks, 4 replicates for each species were collected, separated into roots and shoots and prepared for analysis to measure the starting biomass and endogenous Ni concentration. The nutrient solution was then replaced with the test nutrient solutions to expose plants to varied Ni supply for one month. The simulated serpentine nutrient solution with HEDTA buffering contained 2 mM MgSO4, 1 mM CaCl2, 4 mM KCl, 4 mM urea, 0.1 mM K2HPO4, 1 mM KOH, and 2 mM HEPES buffer. The microelements were added at 20.0 μM Zn, 5.0 μM Mn, 5.0 μM Cu, 20.0 μM Fe, 15 μM H3BO3, and 0.10 μM Mo with 150 μM HEDTA plus the amount of NiHEDTA added, yielding Zn2+= 10-10.35 mol L-1, Mn2+= 10-7.76 mol L-1, Cu2+= 10-13.75 mol L-1, Fe3+= 10-19.05 mol L-1, and Ni at the levels shown in Table 1. Nutrient solution pH was checked every two days and adjusted to 7.0 by KOH addition when needed, and completely replaced after two weeks. The complete nutrient solution contained 0.001 mg Ni L-1 in the absence of Ni additions. Seed of A. murale (from Kotodesh, Albania) contained 7480 mg Ni kg-1 (3.1 μg/seed), while seed of A. corsicum (from Greece) contained 9090 mg Ni kg-1 (9.2 μg/seed), and A. montanum (commercial) contained 0.97 mg Ni kg-1 (0.0017 μg/seed). Rutgers variety tomato seed (from Burpee Seeds) contained 0.30 mg Ni kg-1 DW. We used HEDTA and CDTA and levels of total Ni to give Ni2+ activity levels near the suspected requirement (<10-13 mol L-1). As shown in Figure 1, CDTA binds Ni less strongly, thus providing a Ni2+ activity of about 2 log units of Ni2+ activity higher with CDTA than with HEDTA buffering. One way to look at the selectivity of a particular ligand for Ni is to look at the relative binding of Ca and Ni by the compound. Ca is the major cation in the solution; thus, Ca fills most of the chelator not filled with strongly chelated microelement cations and competes for free ligands. The chelators used and Ni2+ levels set by chelator-buffering are shown in Table 1. Four replications of each treatment were grown. Also, we included a treatment without Ni or chelator-buffering and without urea, but with 1.00 mM Ca(NO3)2 and 2.50 mM KNO3 as sources of nitrogen. The nutrient solution contained 2 mM Mg and 1 mM Ca to simulate serpentine soil fertility conditions. The highest Ni treatment, both for CDTA and HEDTA, was expected to provide a higher activity to assure adequacy and see the response (symptoms and accumulation) of the plants to adequate Ni. During growth, the plants on CDTA treatment at pH 7 began to show Fe-deficient interveinal chlorosis in all test species, so we added 20 μM FeHBED to the CDTA treatments to supply higher phytoavailable Fe at that pH, thus preventing chlorosis. 218 Northeastern Naturalist Vol. 16, Special Issue 5 After 30 additional days of treatment, the plants were harvested. The roots were separated from the shoot biomass, rinsed, and oven dried together with the shoot biomass at 70 °C for 24 h. The dried samples, including the samples collected for the background Ni level, were ashed at 480° C for 16 h and then digested with concentrated HNO3, dissolved in 3 M HCl, filtered, and diluted to 25 ml with 0.1 M HCl. The Ni concentrations were determined using atomic absorption spectrometry with background correction and inductively coupled plasma atomic emission spectrometry. Samples of Alyssum seeds underwent the same procedure to measure the Ni seed supply. An Figure 1. Activity of free Ni2+ in HEDTA- and CDTA-buffered serpentine nutrient solution. Table 1. Effect of buffered Ni activity and chelator on shoot yield (g DW pot-1) in test species grown for 30 days in chelator-buffered nutrient solutions. Means within columns (i.e., shoot tissue of each species) followed by the same letter are not significantly different at P ≤ 0.05 (Waller-Duncan test). Alyssum Lycopersicon Treatment -log Ni2+ corsicum A. murale A. montanum esculentum NO3-N-Control - 3.41 a 3.52 a 4.42 a 3.79 a 0.01 μM Ni CDTA 14.3 1.85 c 2.21 b 2.25 c 1.51 cd 5.0 μM Ni CDTA 11.6 2.95 bc 1.93 b 2.67 bc 2.24 bc 31.6 μM Ni CDTA 10.8 2.44 b 1.98 b 2.81 bc 2.18 bc 1.0 μM Ni HEDTA 14.2 2.04 bc 1.76 b 2.74 bc 0.89 d 10.0 μM Ni HEDTA 13.2 2.13 bc 1.71 b 2.73 bc 1.41 d 130. μM Ni HEDTA 11.9 2.51 b 2.41 b 3.20 b 2.32 b 2009 R.L. Chaney, G. Fellet, R. Torres, T. Centofanti, C.E. Green, and L. Marchiol 219 internal standard (Y) was used to improve analysis reliability, and standard plant reference materials were analyzed for quality assurance. Data were analyzed using SAS procedures GLM for ANOVA and Duncan- Waller Means Separation test. Results By the end of the test growth period, shoot yields were lower for A. corsicum in solutions with the lowest Ni activity with CDTA. However, strong Ni deficiency (the main symptom of which is leaf tip burn) was not reached, apparently because seeds supplied enough Ni for considerable growth of all four species. Shoot yields are shown in Table 1. Nickel treatments did not necessarily reduce shoot yield significantly, and growth with urea-N caused lower shoot yield than growth with nitrate-N. Figure 2 is a photograph of the plants grown with HEDTA buffering at harvest, which shows that growth was reduced at the lowest Ni activity tested. Plant Ni levels are shown in Table 2. At time of transplanting, we analyzed seedling roots and shoots for Ni and other elements. At the start of the test growth period, the test species roots and shoots, respectively, contained: 14.7 and 205 mg Ni/kg DW (A. corsicum), 52 and 75 mg Ni/kg DW (A. murale), 12.5 and 0.1 mg Ni/kg DW (A. montanum), and 4.1 and <0.1 mg Ni/kg DW (L. esculentum). Note that the hyperaccumulators had translocated most of the seedling Ni to their shoots, in strong contrast with the non-hyperaccumulators. Table 2 shows the average Ni concentration found in the shoots and roots for A. murale, A. corsicum, A. montanum, and L. esculentum for the seven Figure 2. Alyssum corsicum plants growing in the HEDTA-buffered solutions at harvest. 220 Northeastern Naturalist Vol. 16, Special Issue 5 treatments. For all the treatments for A. murale and A. corsicum, the average Ni concentration of the aboveground biomass is always greater than the corresponding root Ni concentration, confirming the natural ability of the two species to translocate and accumulate the element into their shoots, even at low Ni supply. On the other hand, A. montanum and L. esculentum show a higher Ni concentration in their roots compared to the shoots, with the exception of treatment 10 μM Ni HEDTA for A. montanum and treatment 130 μM Ni HEDTA for L. esculentum. In contrast with the goal, no clear visible symptoms of Ni deficiency or urea toxicity were induced, although yield at the lowest Ni levels were lower than the higher Ni level. These results indicate that Ni deficiency indeed occurred, but was not sufficient to trigger obvious visual symptoms of defi- ciency. Demonstration of recovery of plants after supply of Ni to apparently deficient plants would clearly prove that deficiency had occurred. This result may be explained by the seed Ni supply and by assuming that the Ni requirement level for the two hyperaccumulators is below the concentrations measured after 10 weeks of growth. The analysis of plants harvested before the beginning of the treatments to define the background Ni content of the plants and the analysis of the seeds showed that average Ni content (in μg) is similar for the hyperaccumulators seedlings and the corresponding seeds (respectively: 9.15 μg and 9.81 μg for A. corsicum, 3.10 μg and 1.83 μg for A. murale). The location of Ni within seeds of Alyssum murale is primarily within the embryo, and is retained in the living seedling during germination in chelator-buffered solutions possessing very low Ni2+ activity (T. Centofanti, R.V. Tappero, and R.L. Chaney, unpubl. data). Lower Ni seeds have been produced on Alyssum murale plants growing in MD and VA Table 2. Effect of buffered Ni activity and chelator on Ni concentration (mg Ni kg-1 DW) in shoots and roots of test species grown for 30 days in chelator-buffered nutrient solutions. Means within columns (i.e., shoot tissue of each species) followed by the same letter are not signifi- cantly different at P ≤ 0.05 (Waller-Duncan test).