A sequential procedure for the quantification of biologically produced polyphosphate in sediment samples

Limnetica, 29 (2): x-xx (2011) Limnetica, 30 (1): 17-26 (2011) c Asociaci´ on Ib´ erica de Limnolog´ ıa, Madrid. Spain. ISSN: 0213-8409 A sequential procedure for the quantification of biologically produced polyphosphate in sediment samples Marta Reina1,∗, Laura Serrano1and Han L. Golterman2 1Department of Plant Biology and Ecology, University of Sevilla, P.O. Box 1095, 41080 Sevilla, Spain. 2Rouboslaan 66, 2252 TR Voorschoten, The Netherlands ([email protected]). ∗Corresponding author: [email protected] 2 Received: 4/10/10 Accepted: 25/11/10 ABSTRACT A sequential procedure for the quantification of biologically produced polyphosphate in sediment samples Polyphosphate (poly-P) is a phosphate storage compound widespread in prokaryotic and eukaryotic cells. Its quantification, however, is not yet well developed for sediment samples because of the difficulties associated with using a single extractant to extract a specific P compound without altering others. The present work uses an analytical procedure already developed to measure poly-P in algal cultures and later modified here for sediments to study poly-P in the sediment. The procedure uses the Ethylenediaminetetraacetic acid (EDTA) method for sequential P-fractionation. This approach ensures that the main inorganic phosphate compounds (i.e., Fe- and Ca-bound phosphate) are extracted prior to poly-P quantification. We applied this method to suspensions likely to contain biologically produced poly-P, such as laboratory cultures of Synechocystis PCC 6803 (Cyanobacteriaceae), Anabaena variabilis (Cyanobacteriaceae), and Chlorella sp. (Chlorophyceae) growing in a P- enriched medium, and activated sludge. According to this procedure, the concentrations of poly-P in Synechocystis PCC 6803, Anabaena variabilis and Chlorella cultures were 15.4, 6.2 and 7.2% of the sum of all P-fractions, respectively, whereas activated sludge showed a lower percentage (3.7%). Known volumes of these suspensions and a commercially available synthetic poly-P standard (Trimetaphosphate) were added to natural sediment. The P-composition of each sample was then compared in duplicate to the corresponding control samples (sediment alone). Poly-P in all control sediment samples was extremely low (0-2 µgg −1d.w.). It increased in all sediment suspensions to which biological samples had been added. The increase was particularly noteworthy for Synechocystis (up to 220 µgg −1d.w. of poly-P). The recovery of poly-P in sediment samples to which Anabaena cultures had been added showed some deviation from a 1:1 expected:observed ratio, particularly when 2 ml of the culture were added (ratio of 0.79). However, the recovery improved to 1.12:1 when a larger volume was added (5 ml), probably owing to the inhomogeneous nature of the cyanobacteria culture. In contrast, the addition of 500 mg of trimetaphosphate to the control sediment did not significantly increase (p>0.05) the percentage found for the observed poly-P fraction. Consequently, biologically produced poly-P can be associated with the fraction of poly-P extracted through use of this procedure. Key words: Polyphosphate, EDTA method, P-fractionation, sediment. RESUMEN An´ alisis secuencial para cuantificar el polifosfato biol´ ogico en muestras de sedimento El polifosfato (poly-P) es un compuesto de reserva de fosfato ampliamente distribuido entre c´ elulas procariotas y eucariotas. Sin embargo, a´ un no est´ abiendefinida la metodolog´ ıa para su cuantificaci´ on en muestras de sedimento, debido a la dificultad de encontrar un ´ unico extractante para compuestos espec´ ıficos de P que no altere a los dem´ as. En el presente trabajo se sigue un procedimiento anal´ ıtico para medir poly-P en cultivos de algas que posteriormente fue modificado para estudiar el poly-P en los sedimentos mediante el m´ etodo EDTA de fraccionamiento secuencial de P y que permite extraer los compuestos de P inorg´ anico (P adsorbidos a Fe y Ca) antes de cuantificar el poly-P. Hemos utilizado esta metodolog´ ıa en suspensiones que conten´ ıan poly-P producido biol´ ogicamente, como cultivos de laboratorio enriquecidos en P de Synechocystis PCC 6803 (Cianobacteria), Anabaena variabilis (Cianobacteria), y Chlorella sp. (Clorofita), as´ ıcomo en fangos activados de depuradora. Siguiendo este procedimiento, las concentraciones de poly-P en los cultivos de Synechocystis PCC 6803, Anabaena 18 Reina et al. variabilis,yChlorella sp. supusieron el 15.4, 6.2 y 7.2% de la suma de todas las fracciones de P, respectivamente, mientras que los fangos activados alcanzaron porcentajes menores (3.7%). Se enriquecieron suspensiones de sedimento natural con vol´ umenes conocidos de las anteriores suspensiones y con un poly-P est´ andar comercialmente disponible (Trimetafosfato); se compar´ o, en muestras duplicadas, la composici´ on de P de estas suspensiones con sus correspondientes muestras de sedimento control. La concentraci´ on de poly-P determinada en las muestras de sedimento control fue extremadamente baja (0-2 µgg −1p.s.), pero ´ esta aumentaba en todas las suspensiones enriquecidas con las muestras biol´ ogicas, en particular con Synechocystis (m´ as de 220 µgg −1p.s.). La ratio entre las concentraciones de poly-P esperadas y observadas en las muestras enriquecidas no fue exactamente 1:1, sino 0.79 al a˜ nadir 2 ml y 1.12 al incrementar el volumen a˜ nadido de cultivo de Anabaena (5 ml) debido, probablemente, a la falta de homogeneidad del cultivo de cianobacterias. Sin embargo, la adici´ on de 500 mg de trimetafosfato, un compuesto sint´ etico, a la muestra control de sedimento no increment´ o significativamente (p>0.05) el porcentaje de la fracci´ on de poly-P. Por tanto, el poly-P producido biol´ ogicamente se puede asociar a la fracci´ on de poly-P extra´ ıda siguiendo el procedimiento aqu´ ıpresentado. Palabras clave: Polifosfato, m´ etodo EDTA, fraccionamiento de P, sedimento. Abbreviations used: Abbreviation Description Tot-P Total P i-Pdiss Dissolved inorganic phosphate org-P Organic phosphate compounds FeOOH≈P Iron-bound phosphate CaCO3≈P Calcium-bound phosphate org-P→acid Organic phosphate compounds soluble in acid org-P→EDTA Sum of organic phosphate compounds soluble in EDTA solutions org-P→NaOH Organic phosphate compounds soluble in hot NaOH org-Pres Residual organic phosphate compounds Tot-P→TCA Total P in the hot TCA extract TCA Trichloroacetic acid INTRODUCTION Polyphosphate (poly-P) is a linear orthophosphate polymer containing a variable number of phosphate residues that are linked by energyrich phosphoanhydride bonds. This compound is widespread in living organisms, particularly microorganisms, and plays a multifunctional role in phosphate and energy storage, cation sequestration, cell envelope formation, membrane transport, and metabolic regulation (Kornberg 1995). Low- and high-molecular weight poly-P occur in a variety of locations within microbial cells, from cell membrane to nucleus, as do poly-P-utilising enzymes (Kulaev & Kulakovskaya 2000). The function of poly-P as a phosphate reserve in prokaryotic and eukaryotic cells is well known. It has long been established that the intracellular concentration of poly-P depends strongly on the phosphate concentration of the medium in laboratory cell cultures (Kulaev et al. 2004). The extrapolation of this relationship to natural systems remains problematic. Attempts to use the concentration of intracellular phosphate reserve compounds in algae as an indication of ecosystem trophic level and P-availability to phytoplankton have produced contradictory results (Kenney et al. 2001, Waters et al. 2005, Selig et al. 2006). The storage of poly-P in cells is, however, caused by the transfer of cells from conditions of starvation to a phosphate-rich culture medium and N shortage (Kuesel et al. 1989). The degree of poly-P storage in algae is not significantly affected by variations in phosphate con- Quantification of polyphosphate in sediment samples 19 centration in the medium (Eixler et al. 2005). Incorporation of external phosphate into the cell poly-P pool proceeds in various steps: transport through the cell membrane, phosphorylation, and polymerisation or elongation of the poly-P chain (Plaetzera et al. 2005). Theoretically, this uptake mechanism enables the organisms to maintain continuous growth, but under natural conditions it is more likely that P uptake and growth are decoupled when growth is limited by the lack of another nutrient, often N. Nuclear magnetic resonance spectroscopy (31P-NMR), developed for the determination of poly-P in cell cultures, has been applied to the detection of this compound in soil and sediment. However, it has long been known that great care must be taken to interpret quantifications of organic phosphate compounds because the extractant used will greatly affect both the concentration and the forms of P obtained in a 31P-NMR analysis (Cade-Menun & Preston 1996, Golterman 2006). The complexity of the mixture of phosphate compounds in sediment makes it impossible for a single extractant to dissolve all P forms or to extract a specific P compound without alteration of others. 31P-NMR results for these substances in the sediment are only correct if other interfering compounds can be excluded, e.g., by a preceding purification extraction (Golterman, 2006). Thus, the quantification of organic phosphate compounds in the sediment first requires a sequential procedure to ensure that main inorganic phosphate compounds have been previously extracted (i.e., Fe- and Ca-bound phosphate). This fact notwithstanding, sediment organic phosphate compounds are often quantified by using 31P-NMR after a direct extraction with a mixture of NaOH with or without EDTA (Ahlgren et al. 2006, Reitzel et al. 2006, Hupfer et al. 2007). It is difficult, if not impossible, to compare results from different studies with NMR in which different extractants were used when identifying and quantifying organic P-compounds in the sediment. The present work follows an analytical procedure developed to measure poly-P in algal cultures (Golterman 1960) and later modified for sediment (Golterman 2006). Several samples of biologically produced poly-P were also analysed to provide positive controls for the extraction of poly-P from the sediment. This biologically produced poly- P came from different sources: activated sludge, a natural phytoplankton bloom, and cell cultures of Synechocystis PCC 680, Anabaena variabilis and Chlorella sp. Additionally, sediment samples were augmented with cell cultures, activated sludge, and a commercially available synthetic poly-P standard compound (Trimetaphosphate) to test for possible interference with the determination of poly-P in the sediment. MATERIAL AND METHODS Two cultures of cyanobacteria (Synechocystis strain PCC 6803 and Anabaena variabilis)were grown at 30 ◦C with BG-11c medium (Rippka et al. 1979) in 750 ml flasks. The cultures were bubbled with a continuous stream of 1 % (v/v) CO2in air under continuous illumination (50 µEm −2s−1). Chlorella sp. was cultured with Sueoka medium in 1000 ml flasks (Sueoka et al. 1967). Therefore, these culture media contained a P buffer. Once exponential growth ended, cells were incubated for 12-24 h in light with extra phosphate added to their culture medium to a final concentration of 5 mg l−1of P by adding the corresponding volume of a K2HPO4solu- tion (1.6 mM). Then, cultures were centrifuged at 12000 rpm for 15 min to recover the cells in pellets that were washed with tap water (<0.1 µgl −1 of i-Pdiss,0.3mScm −1) after each centrifugation. Afinal cell suspension of about 200 ml was obtained for each culture. In order to remove i-Pdiss and phosphate associated with the cell surface, 25 ml of trichloroacetic acid (TCA 5 %) was added to 10 ml of each cell suspension in several replicates and kept at 3-4◦C for 0.5-2 h. Then, cell suspensions were centrifuged at 12000 rpm for 15 min, and supernatants were collected in a 200 ml flask for determination of i-Pdiss concentrations. Each pellet was resuspended in 25 ml of 70 % ethanol at 90◦C for 10-20 min to inactivate enzyme activity and to remove lipidassociated P (Golterman 1960) and later centrifuged (12000 rpm, 15 min). Supernatants were 20 Reina et al. collected in a 200 ml flask for determination of i-Pdiss concentrations. Pellets were resuspended in distilled water to obtain the final cell suspensions. The dry weight of these suspensions was determined gravimetrically in 3-6 replicates of 1 ml-aliquot each. Tot-P concentration was measured in 2-6 replicates of 1 ml-aliquot each as i-Pdiss following the method of Murphy & Riley (1962) after acid digestion with 0.5 M H2SO4and K2S2O8(0.5-1 g) at 120◦C for 4 h (De Groot & Golterman 1990). The optical density at 882 nm was read in a HITACHI U-2000 Spectrophotometer using either a 1 cm or 10 cm cuvette. The P-fractionation of the final cell suspension was performed according to the EDTA method for sequential extraction of P in the sediment (Golterman 1996) and later modified to measure poly-P (Golterman 2006). The results are the average of a duplicate of 10 ml aliquots for Chlorella sp., 2 ml aliquots for Synechocystis PCC 6803 and 5 ml aliquots for Anabaena.Inthefirst extraction step, 25 ml of a solution of 0.05 M Ca-EDTA (pH 7.8) containing 1 % dithionite was added to each aliquot (pellet 0, Fig. 1). This extraction was performed during 1-2 h over continuous shaking in a rotor. The suspensions were then centrifuged (12000 rpm, 15 min) and the concentration of i-Pdiss was measured in each supernatant in order to determine the Fe(OOH)≈P. Next, 25 ml of 0.1 m Na2-EDTA (pH 5) was added to the previous pellet and shaken continuously during 2-6 h. The suspensions were centrifuged again. The measured concentration of i-Pdiss in the supernatant corresponds to the fraction of CaCO3≈P. Figure 1. The EDTA method for sequential P-fractionation of sediment (Golterman 1996) with some modifications for the quantification of poly-P as proposed by Golterman (2006). M´ etodo EDTA para la extracci´ on secuencial de las fracciones P sedimentario (Golterman 1996) con algunas modificaciones para la cuantificaci´ on de poly-P propuestas por Golterman (2006). Quantification of polyphosphate in sediment samples 21 The concentration of Tot-P was also measured in the supernatants extracted with EDTA in order to calculate the organic P-fraction dissolved by the reagent itself (org-P→EDTA) as the difference between the Tot-P and i-Pdiss concentrations. Next, 25 ml of 0.5 m H2SO4was added to the previous pellet (1 h, continuous shaking). The suspensions were centrifuged as usual, and the concentration of i-Pdiss in the supernatant was measured to obtain the organic P-fraction soluble in H2SO4(org-P→acid) and to ensure that the extraction of the previous P-fractions was completed. Then, 25 ml of TCA (5 %) were added to the previous pellet, kept at 3-4◦C for 45 min, centrifuged as usual and the concentration of i- Pdiss measured in the supernatant (cold TCA extraction). The same suspension was then heated in a water bath at 90◦C for another 45 min. The concentration of i-Pdiss in the supernatant was measured after centrifugation (hot TCA extraction). The difference between the i-Pdiss concentration in the hot and cold TCA extractions equals the poly-P fraction, according to Golterman (1960), who followed the acid-soluble specific method of MacFarlane (1936 in Kulaev et al., 2004). The concentration of Tot-P in the supernatant with hot TCA was determined after digestion (Tot-P→TCA). Next, 25 ml of 2 M NaOH was added to the previous pellet and kept at 90◦C for 2 h. The suspensions were centrifuged, and each supernatant was brought to a final volume of 50 ml with H2O. The concentration of Tot- P after digestion of the supernatant yielded the organic P-fraction soluble in hot NaOH (org- P→NaOH). The next pellet was digested with 0.5 m H2SO4and1gK 2S2O8at 120◦C for the determination of Tot-P. This fraction was termed the residual organic P-fraction (org-Pres). Owing to the high concentrations of i-Pdiss found in the Fe(OOH)≈P and in the org-P→NaOH, both steps were repeated several times to ensure a complete extraction. The difference between the sum of P- fractions and the Tot-P concentration of these cell cultures was lower than 15 %. A water sample of 1.5 L was collected from the shore of Dulce pond (Do˜ nana National Park, SW Spain) during a phytoplankton bloom in September 2007. An aliquot was preserved with lugol in situ for taxonomic determination with an optical microscope (1000×of magnification). Aphanocapsa aff. holsatica (Cyanobacteria) was the dominant taxon in this bloom. A cell preparation was obtained following the same procedure as that applied to the laboratory cultures (centrifugation, 5 % TCA and ethanol pretreatments), except that no P incubation was performed. Then, dry weight, P-fraction and Tot-P were determined as mentioned above in duplicate with 1 ml aliquots (Fig. 1). The difference between the sum of P-fractions and the Tot-P concentration of the phytoplankton suspension was always lower than 10 % and often lower than 5 %. Dry-weight and Tot-P were also determined for the original water sample before obtaining the cell suspension. A sample of activated sludge was provided by the laboratory staff of the municipal sewagetreatment plant (E.D.A.R. Tablada, Sevilla) and analysed a few hours after collection. A suspension of this activated sludge was obtained and pretreated prior to P-fractionation (in duplicate with 5 ml aliquots) as mentioned above (Fig. 1). Again, the difference between the sum of P- fractions and the Tot-P concentration of the activated sludge was less than 10 %. A 200 ml sediment suspension was made with H2O and dry sediment collected from a temporary pond (Dulce), once the pond had naturally dried up, to represent the control sediment samples (SED). A volume of 10 ml of the final cell suspension of Synechocystis PCC 6803 was added to 10 ml of sediment suspension in duplicate (SED+SYN). Duplicate sediment samples were also augmented with 5 ml of the final cell suspension of Chlorella (SED+CHL), 1.5 ml of the activated sludge (SED+ACT), and Na-trimetaphosphate (SED+TMP) in increasing quantities (100, 250, and 500 mg). The dry weight of each final suspension was evaluated gravimetrically (3 replicates of 1 ml each). A more careful recovery test was also made with sediment samples augmented using 2 different volumes of Anabaena cell culture: 2 and 5 ml of cell culture, with an average dry weight of 0.0101 g ml−1, were added, in duplicate, to a known amount of control sediment (2.0135 g on 22 Reina et al. Table 1. Average concentration of each P-fraction (µgg −1d.w.) in laboratory cell cultures (Synechocystis PCC 6803, Anabaena variabilis,Chlorella sp.), a natural phytoplankton bloom, and activated sludge. Concentraci´ on media de cada fracci´ on (µgg −1p.s.) en los cultivos celulares de laboratorio (Synechocystis PCC 6803, Anabaena variabilis,Chlorella sp.), en un bloom de fitoplancton y en fangos activados. P-fraction Synechocystis PCC 6803 Anabaena variabilis Chlorella sp. Phytoplankton Activated sludge FeOOH≈P 2539 1677 1750 46 757 CaCO3≈P 683 37 51 63 46 org-P→EDTA 12836 3249 2211 761 46602 org-P→acid 135 26 6 2 25 Tot-P→TCA 4012 9651 6385 2108 9490 org-P→NaOH 1304 2042 1295 1419 1815 org-Pres 56 222 277 57 69 P-fractions 21566 16904 11975 4456 58805 average). The final dry weight of each augmented suspension was the sum of the dry weight in each control sediment sample plus the dry weight that corresponded to the added volume of the cell culture. P-fractionation was carried out at the same time for each amended sample and the cell culture. The concentration of Tot-P was carried out as described above using 2 replicates of 1 ml for each suspension. On average, the difference between the sum of P-fractions and the Tot- P concentration of the amended sediment suspensions was less than 11 %. RESULTS The concentration of Tot-P was highest in the sample of activated sludge, had intermediate values for the cell cultures of Synechocystis sp. PCC 6803, Anabaena variabilis and Chlorella sp. growing in a P-enriched culture medium, and was lowest in the natural phytoplankton bloom (Table 1). As expected, the sum of organic P- fractions constituted the largest portion in these samples, ranging from 78 % in the Chlorella culture to 99 % of the sum of all P-fractions in the activated sludge, but each suspension had a different proportion of organic P-fractions (Table 1). The activated sludge showed the highest percentage of org-P→EDTA, whereas the cultures of Chlorella and Anabaena and the phytoplankton bloom were richer in Tot-P→TCA. Once main inorganic P-fractions (FeOOH≈P and CaCO3≈P) and organic phosphate compounds soluble in acid (org-P→acid)hadbeen extracted, the concentration of poly-P was determined as the difference between a cold and a hot TCA extraction (Table 2). The percentage of poly-P was the highest for the culture of Synechocystis (15.4 %) followed by the culture of Chlorella (7.2 %) and Anabaena (6.2 %), whereas the phytoplankton bloom and the activated sludge showed lower proportions (3.8 and 3.7 %, respectively). This P-fractionation procedure was performed on sediment alone (as a control) and on sediment suspensions augmented with biological samples (cell cultures and activated sludge) and with a standard syn- Table 2. Average concentration of i-Pdiss (µgg −1d.w.) extracted with TCA, a procedure that yielded measurements of the poly-P pool from several biological sources: three laboratory cell cultures (Synechocystis PCC 6803, Anabaena variabilis, Chlorella sp.), a natural phytoplankton bloom, and activated sludge. Concentraci´ on media de i-Pdiss (µgg −1p.s.) extra´ ıda con TCA y su correspondiente fracci´ on de poly-P en distintos materiales biol´ ogicos: tres cultivos celulares (Synechocystis PCC 6803, Anabaena variabilis,Chlorella sp.), un bloom de fitoplancton y en fangos activados. cold TCA hot TCA poly-P Synechocystis PCC 6803 48 3375 3327 Anabaena variabilis 26 1067 1041 Chlorella sp. 76 936 860 Phytoplankton 25 197 172 Activated sludge 132 2331 2199 Quantification of polyphosphate in sediment samples 23 Table 3. Average concentration of P-fractions (µgg −1d.w.) in sediment augmented with Synechocystis PCC 6803 (SED+SYN), Anabaena variabilis (SED + ANA), Chlorella sp. (SED+CHL), activated sludge (SED+ACT), and trimetaphosphate (SED+TMP) and their corresponding control with sediment alone (SED). Concentraci´ on media de las fracciones de P (µgg −1p.s) en el sedimento enriquecido con Synechocystis PCC 6803 (SED+SYN), Anabaena variabilis (SED + ANA), Chlorella sp. (SED+CHL), fango activado (SED+ACT), y trimetafosfato (SED+TMP), as´ ıcomo sus correspondientes sedimentos control (SED). SED SED+SYN SED SED+ANA SED SED+CHL SED SED+ACT SED SED+TMP FeOOH≈P + CaCO3≈P 89 220 28 71 89 154 28 40 61 409 org-P→EDTA 10 1164 14 72 10 33 14 161 6 24608 org-P→acid 9913941445 Tot-P→TCA 10 648 1 217 10 76 1 44 2 6 org-P→NaOH +org-P res 114 261 37 123 114 104 37 94 95 110 thetic poly-P compound (Trimetaphosphate) to test for interference with the determination of biologically-produced poly-P in the sediment (Table 3). The largest P-fraction was the sum of org-P→NaOH and org-Pres in all control sediment samples. The augmentation of the control sediment suspensions changed the P-fraction composition (Table 3), and this change was particularly large for the percentages of org-P→EDTA and Tot-P→TCA (Fig. 2). As expected for the dry sediment of a temporary pond, the poly-P concentration was extremely low in all control sedi- SED+SYN SED SED+CHL SED SED+ACT SED SED+TMP SED % 0 20 40 60 80 100 org-P_EDTA poly-P Tot-P_TCA Figure 2. Percentage of target P-fractions in each test of sediment augmented with Synechocystis PCC 6803 (SED+SYN), Chlorella sp. (SED+CHL), activated sludge (SED+ACT), and trimetaphosphate (SED+TMP) and their corresponding control with sediment alone (SED). Porcentaje de las fracciones de P en cada cultivo de sedimento enriquecido con Synechocystis PCC 6803 (SED+SYN), Chlorella sp. (SED+CHL), fango activado (SED+ACT), y trimetafosfato (SED+TMP), as´ ıcomo sus correspondientes sedimentos control (SED). ment samples, almost approaching the detection limit (0-2 µgg −1d.w.), but its concentration increased in the sediment suspensions augmented with biological samples. In contrast, the sediment augmented with the synthetic trimetaphosphate showed a highly significant increase in the org- P→EDTA fraction, whereas the percentage of poly- PandTot-P →TCA did not increase significantly (p>0.05) compared to the control. Further additions of trimetaphosphate (up to 500 mg) to the control sediment suspension did not yield a significant increase (p>0.05) in the proportion of poly-P and Tot-P→TCA, whereas the org-P→EDTA fraction accounted for nearly the sum of all P- fractions (97.8-98.3 %). The recovery of poly-P in sediment samples augmented with the Anabaena cell culture did not show an exact 1:1 ratio between the expected and observed values. This ratio was 0.79 and 1.12 for the addition of 2 and 5 ml of the culture, respectively. The addition of 2 ml of the cell culture implied an expected poly-P concentration of only 10 µgg −1d.w., probably too low a concentration to be detected without error, whereas the larger volume of 5 ml yielded a more precise and accurate result (Fig. 3). In contrast, the addition of up to 500 mg of Na-trimetaphosphate, a synthetic standard cyclic P-compound, did not produce a significant increase in the observed poly-P concentration because the added P appeared in a different P-fraction (org-P→EDTA) at the beginningof theP-fractionation.Consequently,biologicallyproduced poly-P can be associated with the fraction of poly-P extracted through use of this procedure. 24 Reina et al. poly-P ( gg -1d.w.) 0 5 10 15 20 25 30 2ml 5ml Figure 3. Recovery test of poly-P in sediment samples amended with 2 and 5 ml of Anabaena cultures. Mean concentration and SE of the detected poly-P after the augmentation (black fill) and the calculated mean and SE of the expected concentration (white fill) for each added volume. Recuperaci´ on de poly-P en muestras de sedimento enriquecidas con 2 y 5 ml del cultivo de Anabaena: se muestra la concentraci´ on media y desviaci´ on est´ andar (SE) del poly-P detectado (en negro) y la media y desviaci´ on est´ andar (SE) de la concentraci´ on esperada (en blanco) seg´ un el correspondiente volumen a˜ nadido. DISCUSSION This study found the highest percentage of stored poly-P in the Synechocystis culture (15.4 %). Storage of poly-P in laboratory cultures of autotrophic organisms varies widely according to the literature. The proportion of poly-P has been reported to be higher at faster growth rates (Rhee 1973 in Thompson et al. 1994). In some cyanobacterial cultures, poly-P represented 60 % of the Tot-P in cells of Oscillatoria redekei,10%inSynecoccus, and between 3 and 6 % in Anabaena flos-aquae (Thompson et al. 1994). In Scenedesmus cells, between 5 and 32 % of cellular-P accounted for poly-P (Golterman 1960). This variability results from multiple factors, including the extent of P limitation in the cell cultures, growth rate, N starvation, and specific physiological adaptations (Kuesel et al. 1989, Kulaev et al. 2004. Comparison with other results is difficult because poly-P is quantified by a diversity of methods in the literature: in vivo NMR, NMR after a variety of extractants (NaOH, EDTA, TCA), hot water, cold/hot TCA treatment (Hupfer et al. 2008). The elegance of the method presented here lies in the fact that the same reagent was used for the prepurification of the pellet (TCA at 4◦C) and the extraction of poly-P (TCA at 90 ◦C) from the same supernatant. Only temperature was changed to allow for the extraction of this compound, as poly- P is hydrolysed 105-106times faster at 100◦C than at 0◦C (Greenwood & Earnshaw 1984). After hydrolysis, the TCA extract probably yielded a mixture of compounds (defined here as Tot- P→TCA) with a higher proportion of org-P compounds in the Chlorella culture than the Synechocystis culture. Nucleotide-P and sugar-P are reported to be extracted together with poly-P in TCA-soluble fractions from cell cultures of photoautotrophs such as Scenedesmus quadricauda (Golterman 1960) and the cyanobacterium Anabaena flos-aquae (Thompson et al. 1994). In the sequential procedure presented here, the main inorganic P-fractions had been previously extracted by two sequential treatments with chelating agents (Ca-EDTA and Na2-EDTA) followed by an extraction with diluted H2SO4 (org-P→acid) to prevent contamination of subsequent steps with traces of previous fractions. As expected, this diluted-acid washing accounted for a very small percentage of all P-fractions (0.1-0.6 %), similar to that of the final residual pellet (org-Pres). The use of EDTA allowed for the quantification of a P-fraction soluble in EDTA but not detectable until hydrolysed, and thus defined here as an org-P fraction (org-P→EDTA). However, in biological samples, it may correspond to the inorganic phosphate pool of cells (Lawrence et al., 1998). In vivo 31P-NMR spectroscopy of Synechocystis sp. strain PCC 6308 has indicated that inorganic phosphate, but not poly-P with a molecular weight >1000 Da, leaks from the cell as the result of treatment with EDTA (Lawrence et al. 1998). The reason that this cellular inorganic phosphate requires a hydrolysis for its detection is unclear. The lowmolecular-weight trimetaphosphate also required hydrolysis for its detection. Trimetaphosphate is a synthetic cyclic poly-P compound and appears to be more stable in aqueous solution than linear polyphosphates. The assignment of peaks for poly-P in 31P-NMR spectroscopy of soil and sediment extractions has been traditionally based on a standard of commercial trimetaphosphate (Adams & Byrne, 1989; Cade-Menum & Preston, 1996). In our study, however, we used sev- Quantification of polyphosphate in sediment samples 25 eral standards (a commercial trimetaphosphate, cell cultures grown in a P-rich medium and activated sludge) to allow an analytical comparison for biologically stored poly-P. These biological suspensions provided a positive control in order to test the efficiency of this procedure for the detection of cell poly-P. The trimetaphosphate provided a negative control, as it was not produced biologically. Although this procedure does not determine poly-P molecular size or separate poly-P from different intracellular locations, it can be a useful tool to assess the relevance of biologically produced poly-P. This poly-P of biological origin is eventually deposited in the sediment under natural conditions. The deviation from a 1:1 ratio between the expected and detected poly-P concentrations in sediment samples augmented with Anabaena was slightly higher when 2 ml rather than 5 ml of cell culture were added. The former addition implied a lower targeted poly-P concentration (10 and 26 µgg −1d.w., respectively) and was thus more likely to accumulate errors. Nevertheless, only 1 µgg −1d.w. of difference was found to occur between the replicates for the poly-P concentration detected in the duplicated suspension amended with 2 ml, whereas no such error in precision occurred with the 5 ml addition. The calculation of recovery ratios for augmented preparations made using other P sources was hampered by several factors: a) the low biomass of the biological sources implied very low targeted poly-P concentrations; b) the cell cultures produced inhomogeneous suspensions that caused low precision among replicates: c) the number of replicates was severely limited by the low biomass of the cell cultures. The recovery test was conducted for the augmentations involving Anabaena because these cultures produced the least inhomogeneous suspensions. Indeed, dry weight was calculated with a precision 5-10 times higher than the precision obtained with the other P sources. CONCLUSIONS The application of this methodology allows the quantification of the sediment P-fraction that corresponds to the P stored by aquatic microorganisms. The method is feasible for this purpose because it has been tested with biologically produced poly-P (positive control) and a synthetic trimetaphosphate (negative control). The advantage of this method is that it follows the EDTA method of sediment P-fractionation long used in natural aquatic environments. Moreover, it ensures that the main inorganic P-fractions (i.e., Fe- and Ca-bound phosphate) have been extracted prior to the determination of org-P fractions. Additionally, the quantification of biological poly-P is carried out sequentially in the same supernatant after a change in temperature combined with extraction in cold TCA as a specific blank for each sample. Therefore, this procedure is suitable for ecological studies that aim at the quantification of biologically produced poly-P in aquatic systems. ACKNOWLEDGEMENTS We are very grateful to the following colleagues for providing several types of biological material for our analyses: Maribel Muro, Eva Rodr´ ıguez, M.aCarmen Portillo, Juan Gonz´ alez and Mercedes Gonz´ alez. REFERENCES ADAMS, M. A., & L. T. BYRNE. 1989. 31-NMR analysis of phosphorus compounds in extracts of surface soils from selected Karri (Eucalyptus diversicolor f. Muell.) forests. Soil Biol. Biochem., 21: 523–528. AHLGREN, J., K. REITZEL, L. TRANVIK, A. GOGOLL & E. RYDIN. 2006. Degradation of organic phosphorus compounds in anoxic Baltic Sea sediments: A 31P nuclear magnetic resonance study. Limnol. Oceanogr., 51: 2341–2348. CADE-MENUM, B. J. & C. M. PRESTON. 1996. A comparison of soil extraction procedures for 31P NMR Spectroscopy. Soil Sci., 161: 770–785. DE GROOT, K. J. & H. L. GOLTERMAN. 1990. Sequential fractionation of sediment phosphate. Hydrobiologia, 192: 143–149. EIXLER, S., U. SELIG & U. KARSTEN. 2005. Extraction and detection methods for polyphosphate