Plant Cell Physiol. 38(11): 1232-1241 (1997) JSPP © 1997 Phosphate Uptake in the Cyanobacterium Synechococcus R-2 PCC 7942 Raymond J. Ritchie, Donelle A. Trautman and A.W.D. Larkum Biology A-12, School of Biological Sciences, The University of Sydney, NSW 2006, Australia Key words: Active transport — Cyanobacteria — Electrochemical gradient — Membrane potential — Phosphate nutrition. The cyanobacterium Synechococcus R-2 has been extensively used as a model organism for many membrane transport studies (Ritchie 1991, 1992a, b, c, 1996, Ritchie and Gibson 1987, Ritchie et al. 1996). Phosphorus is an essential element for all cells but relatively little informaAbbreviations: ANOVA, analysis of variance; CAPS, 3-[cyclohexylamino]-l-propane-sulphonic acid; DMO, 5,5-dimethyloxazoline-2,4-dione; PAR, photosynthetically active radiation; c (subscript), refers to the control-treated cells; e (subscript), refers to the experimental treatment cells; i (subscript), refers to the inside of the cells; o (subscript), refers to the outside of the cells or bulk electrolyte. tion is available on the physiological mechanism of uptake in cyanobacteria. Despite the obvious importance of phosphate nutrition to algal growth under eutrophic conditions, little is known about the physiology of uptake of phosphate by cyanobacteria (Grillo and Gibson 1979, Falkner et al. 1980, 1989, Rigby et al. 1980, Budd and Kerson 1987, Wagner et al. 1995). Phosphate is often cited as the cause of cyanobacterial blooms but its removal from waste-water is very expensive, and cause and effect might not be straightforward. Many previous studies have worked with cells in experimental medium likely to lead to spurious results or results that are not applicable to cyanobacteria growing in nature. For example, Mohleji and Verhoff (1980) used a medium (NAAM) with a very low ionic strength, no buffers and the pH was not specified in their experiments. Grillo and Gibson (1979) used triethanolamine/H2SO4 (pH 8) buffer. They tested for K + and Mg2* effects but not for Na + or Ca 2 + . Tris is known to be toxic to many cyanobacteria, but has been used in many previous studies. Falkner et al. (1980) used Tris/H2SO4 (pH 7.5) buffer. Ca 2+ and Mg*+ effects were investigated but not K + or Na + . Rigby et al. (1980) and Budd and Kerson (1987) also used Tris/H 2 SO 4 (pH 8). Both studies tested the effects of many different ions upon phosphate uptake including Na + , K + , Ca 2+ and Mg2"1" but not in combination. The conclusions drawn in these studies might not have been fully justified. More recently, Falkner et al. (1989) and Avendano and Valiente (1994) have used incubation medium containing the same constituents as the culture medium plus HEPES/NaOH or KOH buffer and so their results are more likely to have some relevance. Although there is little doubt that phosphate is actively taken up by cyanobacterial cells (Grillo and Gibson 1979, Falkner et al. 1980), a formal thermodynamic analysis of phosphate uptake has not been attempted. The aim of this study is to perform a biophysical analysis on the mechanism of uptake and the intracellular concentrations of phosphate under eutrophic conditions with attention to pH and Na + , K+ and the divalent cations, Mg2* and Ca 2+ . Much of the previous work on phosphate nutrition has focused upon phosphate deficient rather than eutrophic conditions. Synechococcus was used in the present study because many cyanobacteria have a mucilaginous sheath which makes transport studies very difficult or impractical. Unfortunately, the strains of Anabaena and Microcystis responsible for toxic cyanobacterial blooms are amongst 1232 Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 Phosphate uptake rates in Synechococcus R-2 in BG11 media (a nitrate-based medium, not phosphate limited) were measured using cells grown semi-continuously and in continuous culture. Net uptake of phosphate is proportional to external concentration. Growing cells at pH 0 10 have a net uptake rate of about 600 pmol m~ 2 s" 1 phosphate, but the isotopk flux for 32P phosphate was about 4 nraol m~ 2 s~'. There appears to be a constitutive over-capacity for phosphate uptake. The Km and Vmtx of the saturable component were not significantly different at pH0 7.5 and 10, hence the transport system probably recognizes both H2PO4" and HPO2,". The intracellular inorganic phosphate concentration is about 3 to 10 mol m~3, but there is an intracellular polyphosphate store of about 400 mol m~3. Intracellular inorganic phosphate is 25 to 50 kJ mol' 1 from electrochemical equilibrium in both the light and dark and at pHo 7.5 and 10. Phosphate uptake is very slow in the dark (»100 pmol m~ 2 s~ 1 ) and is light-activated (pHo 7.5«1.3nmolm" 1 s~ 1 , pHo 10«600 pmol m"^" 1 ). Uptake has an irreversible requirement for Mg2+ in the medium. Uptake in the light is strongly Na+-dependent. Phosphate uptake was negatively electrogenic (net negative charge taken up when transporting phosphate) at pHo 7.5, but positively electrogenic at pHo 10. This seems to exclude a sodium motive force driven mechanism. An ATP-driven phosphate uptake mechanism needs to have a stoichiometry of one phosphate taken up per ATP (1 PO 4 , 0 /ATP) to be thermodynamically possible under all the conditions tested in the present study. Phosphate uptake in a cyanobacterium 1233 those that have thick mucilaginous sheaths making them unsuitable for anything but simple growth studies of nutrient uptake which are not likely to make great progress in understanding the mechanism of uptake of phosphate in cyanobacteria. However, the general findings in the present study should be applicable to understanding the mechanism of phosphate uptake in a variety of cyanobacteria including the more notorious Anabaena and Microcystis. Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 chie et al. 1996) using Nuclepore® (Costar Corp., Cambridge. MA, U.S.A.) polycarbonate membrane filters (0.4 or 1 /mi). The usual specific activity was =300 to 10,000 GBq mol" 1 . The cell wall of Synechococcus binds significant amounts of cations (Ritchie and Gibson 1987, Ritchie 1992a). Cells were washed with 20 mol m~3 Ca(NO3)2 (buffered to pH 7.5 with HEPES/ CaJOH)^ for about 10 s to remove most of the extracellular M Rb + . Cells were exposed to ^Rb* radiolabel for a few seconds, were washed and then used as a measure of binding of 86 Rb + to the surface of the cells. Binding of 32PO4 to cells was much slower than for 86 Rb + , taking about 1 min to reach steady-state, and was much more difficult to wash from the cells. Materials and Methods Inorganic phosphate—Total reactive inorganic phosphate Chemicals and radiochemicals—^Rb"1" was from DUPONT was assayed using a semi-micro version of the molybdate/stannous chloride method described in APHA (1992). The intense New England Nuclear, Boston, MA, U.S.A. 32P-H3PO4 was molybdenum blue colour produced by the reduction of the from the Australian Nuclear Science Technology Organization molybdophosphoric acid by stannous chloride was measured at (ANSTO), Lucas Heights, Sydney, Australia. CAPS, DCMU, 700 run. Standard curves were linear to 50 mmol m~3 PO 4 . Fluorinert FC-77, HEPES, methylamine, HC1 and valinomycin Soluble phosphate in experimental cell suspensions was usualwere from Sigma-Aldrich. All reagents used were of analytic standly estimated by centrifuging in Eppendorf tubes in a Beckman Miard. cro fuge (6,000 x g) and sampling the supernatant. Cell densities of General—Synechococcus R-2 (S. leopoliensis, Anacystis about 600 x 1012 cells m" 3 were used. nidulans) (PCC 7942) was grown from axenic stock cultures in Intracellular phosphate—To estimate intracellular phosBG-11 medium (Allen 1968) modified as described previously (Ritphate, a cell suspension was filtered onto a polycarbonate filter chie and Gibson 1987, Ritchie et al. 1996) but with the total phosand washed in buffered calcium nitrate. The cells were then susphate concentration lowered to 50 mmol m~3 and total inorganic pended in an Eppendorf tube in water containing chloroform (1% carbon increased to 1 molm" 3 . The unmodified BG-11 (PO 4 = v/v) for 10 min then centrifuged and the supernatant assayed for 175 mmol m~3) precipitates much of its phosphate upon autoclavreactive soluble phosphate. ing. Phosphate-free BG-11 contained much less than 0.1 mmol m~3 phosphate. The cells were grown on air, either semi-continuUptake ofnP— Both chemical assays and 32P assays of phosously (~ 1/4 dilution each day) or in a continuous culture device, phate showed that significant amounts of phosphate bind to the in continuous light of about 150^mol (quanta) m~2 s~l PAR, ussurfaces of Synechococcus cells (Falkner et al. 1980). Saturation ing cool white fluorescent lights, at 30°C. Similar light conditions kinetics were only observable at concentrations below 5 mmol were used in experiments but the temperature was 25°C. The pH 0 m~3 using very dilute cell suspensions (A 7 5o=0.2=60x 1012 cells of a growing culture was about 10 to 10.5. m~3) and short labelling times. 32P uptake was measured using 0.2 to 5 mmol m~3 phosphate (specific activity = 100 x 109 to 13 x 1012 All glass and plastic laboratory-ware were HC1 acid-washed Bq mol" 1 ) and incubation periods of 2 min and 5 min. The rebefore use. The cells were usually harvested by centrifugation sidual phosphate in the "P-free" incubation media was meas(2,500 x j ) in a preparative centrifuge and washed three times in ured and taken into account in specific activity calculations ( = 0.1 the experimental medium. All experiments were run in media as to 4 mmol m" 3 ). Cells were pre-incubated for one h in buffered close as possible in ionic composition to the BG-11 medium in phosphate-free BG-11 before commencing a labelling experiment. which they had been grown. A routine pre-incubation of 30 min 32 P labelled phosphate was added to the cap of the incubation was run in the light or dark as appropriate, before using cells in vials (Eppendorf tubes) and labelling' started by shaking. At the the experiments. two minute time point, 0.5 ml of the cell suspension was removed Intracellular volumes, cellular surface areas and numbers of from an incubation vial and filtered onto a 0.4 //m filter and washcells were calculated using stereological data from Ritchie and Gibed twice with unlabeled BG-11 to remove as much surface bound son (1987) and Ritchie (1991) using an empirical relationship be32 P-label as possible. The cells on the filter were then resuspended tween light scattering (A75onm), measured using a Varian® Cary in 0.5 ml of distilled water, 3 ml of scintillant was added and the 13E UV-Visible spectrophotometer. samples were counted. Fluxes were calculated from the difference Buffer solutions—The ionic composition of BG-11 medium in uptake by the cells after the 2 and 5 minute incubation times as was changed as little as possible: buffers were adjusted to the apmol m~2 s" 1 . At shorter labelling times it was not possible to dispropriate pH 0 using NaOH or Ca(OH)2 for increasing alkalinity. tinguish intracellular uptake from the extracellular bound compoCa(OH)2 contains some CaCO3: this needs to be removed by filtranent because binding of phosphate to the cells was slower than in tion or sedimentation for preparing media with limited inorganic the case of cations. carbon. The following buffers were used at 5 mol m~3: pH 7.5 HEPES/Ca(OH) 2 or NaOH, or pH 10 CAPS/NaOH or CAPS/ O2 evolution—Photosynthesis was measured using a ClarkCa(OH)2. type electrode (Hansatech, Kings Lynn, Norfork, U.K.) thermostatted to 25°C and used as described by Walker (1990). The Dark techniques—When using Synechococcus, dark experielectrode was calibrated with aerated water for 100% air saturaments must be run in completely dark conditions (Ritchie 1991). A tion and with N2 as zero O2. The oxygen solubility algorithms of dim, green safe-light was only used for taking samples (about 10 s, Carpenter (1966) and Colt (1984) were used to calculate the oxy0.1 /rniol (quanta) m~2 s"1 PAR). gen concentration in air-saturated medium. The chamber was illuMembrane filtration techniques—A Millipore filtration apminated with saturating red light (peak 660 nm), provided by a paratus with a sintered glass surface was used for filtration assays light emitting diode array (175 ^mol (quanta) m" 2 s" 1 PAR). (Ritchie and Gibson 1987, Ritchie 1991). Uptake of M Rb + was measured using filtration techniques as previously described (RitHarvested cells (~300x 1012cells m" 3 ) were filtered onto 1234 Phosphate uptake in a cyanobacterium experiment (Ritchie and Prvan 1996a, b). ±95% confidence limits were then calculated using n-2 degrees of freedom (Zar 1974). Counting methods—32P and 8<Rb+-label were counted using the "P-channel of a Canberra-Packard Tri-CARB 1600 scintillation counter using Emulsifier-Safe scintillant. Statistics—Error-bars are ±9 5 % confidence limits with the number of replicates in brackets 0- Where two numbers appear in brackets (a, b), the first is the number of separate experiments conducted and the second is the total number of observations. Students t-tests, linear regressions including the error-limits of the slopes (m) and y-intercepts, and other statistics were calculated as described by Zar (1974). Results Uptake of phosphate—Figure 1A shows the uptake of phosphate by Synechococcus cells versus time in the light. After the 30 min pre-incubation period in nominally phosphate-free medium the cells maintained an extracellular phosphate of about 1 mmol m~3. Cell suspensions did not pull extracellular phosphate down to zero. Following the addition of 50 mmol m~3 phosphate, the cells rapidly depleted the phosphate to near zero over 4 h. It was expected that such a curve could be used as a method for estimating the Km of phosphate uptake (Duggleby 1985). Uptake follows an exponential decay curve which implies that the net flux is proportional to concentration. Figure IB shows a plot of the rate of uptake of phosphate by SynechoEquation 1 coccus cells versus time, calculated as the rate of uptake between each sampling time, using the data from Figure 1A. where, F/RT are the Faraday, Gas Constant and absolute temperaThe net uptake rate was a linear function of concentration ture respectively, A !Pe is the membrane potential of cells given the experimental treatment at t = 0 , A *FC is the membrane potential of and does not exhibit saturation kinetics (Fig. 1 A, B). The ythe control cells (assumed to be constant), <t>c is the uptake flux of intercept is not significantly different to zero (—0.208 ± the permeant cation in the experimentally treated cells, & is the 0.225) and so the rate of uptake is directly proportional to flux of the permeant cation in the control cells. Equation 1 can concentration (0PO =k«[PO ] ; r=0.990, where k=41.6 4 4 o only be solved iteratively. The error of A ^ can be estimated by a partial differentiation with respect to A Vz, <j>c and 0C (Young 1962). (±5.8)x 10~'ms"'). The saturable component of phosphate uptake was only apparent if very low phosphate conEquilibrium accumulation of these permeant cations was centrations were used (below 5 mmol m~3; see Methods: routinely measured for both control and experimental treatments 32 and so both the initial effects of an offered substrate and the memUptake of P). brane potential after the cells had been exposed to the changed Preliminary experiments showed that phosphate upconditions for a considerable time could also be measured. take was strongly inhibited at acid pH 0 , had a broad maxIntracellular pH—Intracellular pH (pHJ was determined using silicone gradient centrifugation techniques as described in Rit- imum at pH o 7 to 9 then fell off at more alkaline conditions. Uptake of phosphate was significantly higher chie (1991) and Ritchie et al. (1996). The weak acid, I4C-DMO (ANOVA p<0.001) in the light than in the dark, and the (5,5-dimethyloxazoline-2,4-dione), was used as the pH-probe for the experiments at pH o 7.5 and the weak base, 14C-methylamine, rate of uptake of phosphate was much greater at pH 7.5 for pH : determinations at pH o 10. Both probes were used at a final than at pH 10 in the light (ANOVA p<0.001). Under the concentration of 100 mmol m~3 with an incubation time of about alkaline conditions under which the cells were grown, 14 3 5 min. In the C-methylamine experiments, 100 mmol m~ NH«" p H 0 ~ 10, the net uptake rate could be described by the was present to discourage metabolism of methylamine. Label carequation (0PO4=k-rPO 4 ] o , where k=9.0 ( ±1 . 8 ) x l 0 ~ 9 ried through the silicone oil with the cells in the extracellular water space and label bound to the surfaces of the cells were allowed for. ms" 1 ). pH did not significantly alter the rate of phosphate Kinetics—Uptake of phosphate was followed routinely for up uptake in the dark (Table 1). The photosystem II inhibitor, to 6 h. Net uptake vs time was fitted to an exponential model usDCMU (5 mmol m~ 3 ), decreased the rate of phosphate uping a non-linear least squares fit (Johnson and Faunt 1992) giving take by the cells in the light from 1,770±240 (6) pmol m~2 estimates of the exchange constant (k) and initial concentration s~' to l,270±370 (6)pmolm~ 2 s""' (p=0.02) and so did (E0,t=.o). and the initial uptake flux calculated as k-E o l = 0 . TheK m 32 not decrease net phosphate uptake to the very low rates and V^a of the isotopic uptake flux of P were estimated using non-linear least squares fitting with at least ten data points in each found in the dark. Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 polycarbonate membrane niters, then washed three times in experimental media. Oxygen electrode experiments were run on cells incubated for about 1 h in phosphate-free BG-11 at pH o 7.5 and pre-incubated in the light. Samples were purged with water saturated N2 gas before 1 ml of cell suspension was placed in the oxygen electrode chamber. The effect of 50 mmol m~3 phosphate was tested using a paired t-test (control vs. added phosphate). Membrane potential—The membrane potential (A !Pi,0) was measured using the valinomycin expedited M Rb + equilibration technique described and discussed in Ritchie (1991) and Ritchie et al. (1996) but using filtration and washing to measure Valinomycin-mediated uptake of s *Rb + . The membrane potential was calculated from the accumulation ratio of the permeant cation (Ritchie 1991, 1992a) using the Nernst equation (Nobel 1983) after cell wall binding was corrected for as described above. Activities were calculated from concentrations using the Bjerrum version of the Debye-Huckel equation (Hamer 1968). For consistency with another study, all A fu0 estimates using the 86 Rb + /valinomycin probe quoted in the present paper are based on 2 h, rather than 1/2 h incubations. Transitory polarization of the membrane potential—Provided a cation is taken up passively (86Rb+/valinomycin) and its permeability across the cell membrane is constant, the uptake rate of the cation under a control and experimental condition can be used to detect polarizations and recoveries of A ¥, o (Ritchie 1992a, Ritchie et al. 1996). The A Yx_0 of the experimentally treated cells can be calculated as; Phosphate uptake in a cyanobacterium 1235 [PO 4 ] O = 55.12*10"0.326t = 0.047[PO4] - 0.208 r = 0.999 r = 0.990 • • Extracellular phosphate Q Preincubated zero PO4 Net PhotphaM Flux (nmol m-2 »"1) ato 5 | a. a •5 X 3 1 .. E s 2 s •2 1 1 S 2 0 2 5 3 0 1 5 4 0 4 S 5 0 5 5 M 1.5 External PO4 (mmol m"3) Time (h) Fig. 1A Fig. IB Fig. 1A Uptake of phosphate by Synechococcus cells versus time in the light. Cells were pre-incubated in the experimental conditions for 30 min before the addition of phosphate. The cells had reduced the extracellular phosphate to about 1 mmol m~3 during the pre-incubation period in nominally phosphate-free medium. The pH 0 was 7.5 using 5 mol m~3 HEPES/Ca(OH)2 buffer. Care was taken to add enough Na2CO3 so that cells did not run out of inorganic carbon before the incubation was complete. Uptake follows an exponential decay curve. Fig. IB Rate of uptake of phosphate by Synechococcus cells versus time in the light using the data from Fig. 1A calculated as the rate of uptake between each sampling time shown in Fig. 1 A. The net uptake rate is a linear function of concentration and does not exhibit Phosphate uptake rates of cells growing in continuous culture could be calculated from the upstream and downstream concentrations of phosphate, the washout rate and the cell numbers in the effluent stream (Tempest 1970). Table 1 shows that the phosphate consumption rate of the exponentially growing cells (=600 pmol m~2 s"1) was close- ly comparable to that measured on harvested cells under similar conditions (pH 0 = 10). The effluent stream contained 18±3 (12,44) mmol m" 3 from a feedstock of =50 mmol m~3 and so were not phosphate limited cells. Table 2A shows the intracellular phosphate concentrations of the alga after 1 h incubation with either nominally Table 1 Phosphate consumption by Synechococcus Condition Net uptake rate (pmolrn~ 2 s"') Dark Light Harvested cells pH o 7.5 l,303±146 (11,118) 136±91 (5,41) pH 0 10 449± 90(11,83) 146±53 (5,38) 586± 65 (12,48) NA Continuous culture (~pH o 10) Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 OJ 1236 Phosphate uptake in a cyanobacterium Table 2A Thermodynamic data for analysis of phosphate uptake [P04]0 (mraol ra"3) Phosphate in media and cells POJ, (mol m ) Light Dark Membrane potential (mV) ila or proton motive force (kJmor 1 ) Intracellular pH Light Dark Light Dark Light -13211.2 (2,16) -11411.6 (2,8) 7.3810.12 (3,12) 7.0010.18 (3,12) -13.410.7 10.311.5 8.911.2 -12912.8 -13113.1 (3,12) (3,14) (7,44)' (5,20)° " Combined data from this study and Ritchie et al. (1996). * Taken from Ritchie et al. (1996). 7.2210.07 (6,36)* 7.3610.11 (4,24)* +3.410.5 pH07.5 1.7310.64 (9) 5.9+1.9 (5,20) 3.010.8 (2,8) pH07.5 50 pHolO 3.8112.08 (5) 9.9+2.4 (5,20) 7.711.5 (3,12) 9.713.3 (3,14) 4.712.2 (2,8) pHo10 50 Table 2B Species H 2 PO 4 " HPO 4 ~ PO 3 ,Total H2PO4" HPO 4 ~ PO3," Total -13.911 +2.410.11 Intracellular pH and membrane potential—Synecho^ coccus cells are able to maintain their intracellular pH within narrow limits, even when the cells are maintained at a pH o which is several pH units different to that of the internal concentration (Ritchie 1991, Ritchie et al. 1996). Cells incubated at pH o 10 have similar membrane potentials and intracellular pH in the light and dark. Cells incubated in the dark at pH 0 7.5 have a different membrane potential and intracellular pH to those of cells incubated at pH 0 7.5 in the light. The membrane potential and intracellular pH of cells in the light at pH 0 7.5 are similar to those found in both the light and dark at pH 0 10. Using the A !PiiO and pHj data, the electrochemical potential for protons or proton motive force (A/xH^0 or pmfi>0) was about —13 kJ mol" 1 at pH 0 7.5 (active extrusion of H + ) but only about + 3 kJ mol" 1 at pH 0 10 (active uptake of H + ) . Thermodynamic analysis—Tables 2B and 2C present the results of thermodynamic analyses, using the Nernst criterion, and show the driving force required to active- Thermodynamics of phosphate uptake at pH 0 7.5 6 500 ±185 xlO" 1.23± 0.46 xlO" 3 20 ± 7.4 xlO" 9 1.73+ 0.64X10" 3 14.4 35.6 577 50 xlO" 3 xlO" 3 x10" 9 x 10"3 Dark Light [PO4]O mol m 3 [POJli molm" 3 AuPOli kJ mol"^ [PO4]i molm 1.95 ±0.65 3.95±1.31 48±16xlO"« 5.9 ±1.9 + 33 + 1.2 +45 ±0.63 + 57±0.43 1.63 ±0.45 1.38+0.38 6.9±1.9xlO" 6 3.0 ±0.8 3.27±0.84 6.63 ±1.69 8 0 ±2 1 x l 0 " 6 9.9 ±2.4 +26±0.65 + 38±0.34 + 50±0.24 5.26±1.84 4.45 ±1.56 22±7.9xlO" 6 9.7 ±3.3 AfiPOzti kJ mol + 31±1.15 + 39±0.60 +47±0.41 + 25±0.88 + 33±0.46 + 41 ±0.33 Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 zero (about 1 to 4 mmol m 3) or 50 mmol m 3 added phosphate. Under low-phosphate conditions the intracellular phosphate concentration is about halved in the dark at both pH o 7.5 and 10. For cells in low phosphate the apparent intracellular phosphate was 6.6± 1.3 mol m~ 3 (8,32) or 787 ±173 n m o l m " 2 on a cell surface area basis in the light. Cells in 50 mmol m~ 3 phosphate had a higher intracel, lular phosphate concentration of 1 0 ±1 . 6 m o l m ~ 3 (8,32) or 1.21 ±0.22//mol m~ 2 . There was no apparent difference in intracellular phosphate in the light or dark or for cells in pH o 7.5 or 10 when high external phosphate was present. Chemically detectable free phosphate was only a small proportion of the total phosphate accumulated by the cells. The chemostat data could be used to calculate the total phosphorus held by the cells from the depletion of phosphate from the medium and the volume of the cells in the effluent stream. The total intracellular phosphorus was 424±47 mol m " 3 (12,48). Thus, only about 2A±0A% of total intracellular phosphorus is present as phosphate. Dark Phosphate uptake in a cyanobacterium 1237 Table 2C Thermodynamics of phosphate uptake at pH 0 10 Dark Light Species H2PO4" HPO 4 ~ POJTotal H2PO4" HPO 2 " PO 4 " Total mol m" 3 [PO4]j [PO4L molm 3 molm" 3 kJ m o r 5 4.9 ± 2.7 x l O - 6 3.79± 2.07 xlO" 3 19 ±11 xlO" 6 3.8 ± 2.1 xlO" 3 3.20±0.64 3.50±0.90 3 8 ±7 . 5 x l O - 6 7.7 ±1.5 +45±1.5 +42±0.77 + 38±0.55 1.60±0.76 3.10±1.47 36±17xlO" 6 4.7 ±2.2 64 x 10- 6 49.7 xlO" 3 255 x 10- 6 50 x 1Q-3 4.29±0.65 6.02±0.91 5O±7.6xlO- 6 10.3 ±1.5 +40±0.46 + 36±0.33 + 33±0.3 3.03±0.46 5.87±0.90 6 8 ±1 0 x l 0 " 6 8.9 ±1.2 + 39±0.46 + 37±0.33 + 34±0.32 previously for Synechococcus (Ritchie et al. 1996). Effects of ion deficiencies—The effects of different ions upon phosphate uptake rates are shown in Table 3. Lack of K + decreased phosphate uptake by about 50%. Phosphate uptake in nominally Na+-free media was consistently inhibited by approximately 80%. Nominally Na+-free media contained about 6 mmol m~ 3 Na + due to contamination of chemicals (Ritchie et al. 1996). Lack of Ca 2+ had no effect on the uptake of phosphate by Synechococcus. Mg24" not only decreased the net uptake of phosphate, but there was a net loss by the cells. The effects of lack of Mg2"1" appeared to be irreversible as adding this ion back into a cell suspension in Mg2+-free BG-11 did not lead to a recovery of phosphate uptake by the cells. Electrogenicity of phosphate uptake—Table 4 shows the polarizing effects of added phosphate (50 mmol m~3) measured using the Valinomycin-mediated uptake of 86 Rb + as the membrane potential probe (Equation 1). The effects of phosphate uptake upon the membrane potential appear to be different at pH o 7.5 and 10. This is consistent with the great differences in the abundances of HzPO^ and HPO 2 " at pH 0 7.5 and 10 (see above). At pH 0 7.5 uptake of phosphate in the light led to a hyperpolarization of the apparent membrane potential. This indicates a net uptake of negaTable 3 Effect of mineral ion deficiencies on the uptake of PO 4 in the light at pH 0 10 Ion Uptake flux (pmol m 2 s ') With added cation Cation omitted K+ + 546±70 (2,12) + + 871 ±276 (5,39) +280 ±47 (2,12) + 174±40 (5,37) 2+ + 308±71 (4,27) + 361 ±159 (2,13) +426±121 (1,7) - 1 4 5 ±47 (2,13) Na Ca Mg 2+ Fluxes are positive for net uptake and negative for net loss by the cells. Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 ly transport each species of soluble phosphate into the cell (Nobel 1983). The relative amounts of each form of phosphate was calculated using the equations and pKa values published by Lindsay (1979). No form of phosphate (HjPOr, HPO 4 ~ or PO4~) could be accumulated passively by the cells. All forms are at least +25 kJ mol" 1 from electrochemical equilibrium at both pH 0 7.5 and 10. Uptake of 32P—At pH 0 7.5 the Km was 1.65±1.04 (4,44) mmol m~3 and 2.92± 1.46 (4,47) mmol m" 3 at pH o 10. These values were not significantly different and so an overall mean of 2.31 ±0.90 (8,91) mmol m~3 could be calculated. Vmax was also similar at pH 0 7.5 (K mM =3.52±0.52 (4,44) nmol m ^ s " 1 ) and pH o 10 (Vmlx=4.25±1.04 (4,47) nmol m~2 s~l) giving an overall mean of 3.90±0.59 (8,91) nmolm~ 2 s~'. Isotopic fluxes were considerably higher than the chemically measured net fluxes and showed different kinetics. Taking the intracellular phosphate content on a cell surface area basis calculated above (=1.2/«nol m~2) and the observed isotopic flux, the exchange constant is 3.22 (±0.761) x l 0 ~ 3 s " ' o r a t I/2 of 215±51 s.It can be concluded that the isotopic flux is a measure of phosphate transport across the plasmalemma, whereas the chemical assay, because of the time course used, is a measure of metabolic uptake of phosphate. O2 evolution—Synechococcus has a very large intracellular store of phosphorus and so short-term incubations in phosphate-free media was not expected to affect photosynthesis or respiration. There was no significant effect of added phosphate in the dark using a paired t-test or nonparametric sign test (R= 12.6± 1.55 nmol m~ 2 s~' (3,12)). There was, however, a significantly inhibitory effect upon photosynthesis of added phosphate when it was added to the phosphate-free cells (paired t-test, p=0.2%; Nonparametric Sign test p=0.02%). The gross photosynthetic rate (P o ) of the control cells was 56.2±4.7 nmol m~ 2 s~' (3,12) vs. 50.9±3.6 nmol m" 2 s~' (3,12) after adding phosphate. These R and P G rates are comparable to those found kJ mor° +44±1.82 +41 ±0.95 + 39±0.67 1238 Phosphate uptake in a cyanobacterium Table 4 Polarization effects of phosphate using 86Rb+/valinomycin as the membrane potential probe Membrane potential (mV) Conditions BG-11, BG-11, BG-11, BG-11, BG-11, BG-11, A'F, pH o 7.5, Light pH o 7.5, Dark pH 0 7.5, Light+DCMU pH o 10, Light pH o 10, Dark pH o 10, Light+DCMU control 120 min exposure -132±2 -112±1.8 -122±1.8 -128±4.6 -126±1.8 -137±1.9 -132±2 -115±1.8 -128±1.9 -128±5.1 -129±1.9 -137±2 AVcc polarization - 5 9 ±23 -17±18 -26± 8 + 39± 7.7 -4±18 + 50± 9.7 (2,64) (32) (32) (3,96) (32) (32) Cells were pre-incubated in BG-11 medium in capped Eppendorf tubes. The valinomycin-mediated uptake flux of !6 86 + was measured at 0, 2, 4 and 6 min. Control and experimental cells were incubated as matched pairs. The ^Rb"1" fluxes were measured by shaking the tubes and removing a cell sample for filtration after the appropriate labelling time. Membrane potentials were calculated from the equilibrium accumulation ratio of ^Rb* after labelling for 120 min of both control and experimentally treated cells. Discussion Distinguishing transport of phosphate across theplasmalemma from assimilation of phosphate—At concentrations of phosphate above about 5 mmol m~3, net uptake was directly proportional to concentration with no saturable component (Fig. 1A, B) and if uptake flux was plotted against concentration, the apparent intercept was through zero. Wagner et al. (1995) found that when phosphatestarved cells were subjected to pulses of phosphate, the uptake behaviour of the cells was linearised, but failed to recognize that the y-intercept passed through zero. Since the cells used in the present study were not phosphate-limited, then the linearized behaviour found in the present study and by Wagner et al. (1995) is not restricted to phosphate limited cells. Isotopic fluxes were considerably higher than the chemically measured net fluxes and showed saturable kinetics if labelling times comparable to the filling time of the intracellular inorganic phosphate pool (t, / 2 ~4min) and very low phosphate concentrations were used. The Km for phosphate was only about 2 mmol m~3 and the saturating flux was about 4 nmol m~2 s" 1 . The Km value is comparable to previous estimates in Synechococcus (Budd and Kerson 1987, Avendano and Valiente 1994, Wagner et al. 1995), other cyanobacteria (Istvanovics et al. 1993, Valiente and Avendano 1993) and the bacterium Streptococcus (Poolman et al. 1987). Taking the intracellular phosphate pool on a surface area basis, it was possible to calculate a turnover time for the cytoplasmic phosphate. Turnover of the intracellular chemically detectable phosphorus was very rapid (ti / 2 ~215±51 s) which is comparable to an estimate of = 10 min by Suttle et al. (1991). Thus net uptake experiments, such as Figure 1, conducted over a period of hours, with samples taken every 15 or 30 min, were actually measuring the rate of incorporation of phosphate into polyphosphate storage material. Kinetic analyses of phosphate uptake in Synechococcus have shown that both the Km and Vmax are not significantly different at pH 0 7.5 and pH 0 10. This has the implication that the phosphate uptake mechanism operating in neutral media is the same as that operating at alkaline pH 0 . From the abundance of H2POi" and HPOj" at pH o 7.5 and pH 0 10, both the mono- and divalent species of phosphate are accepted by the transport system. At pH 0 7.5 about 29% of total phosphate is in the form of H2PO4~ and 7 1 % in the form of HPOl~ whereas at pH o 10 only 0.13% is present as H2PO4~, 99.4% as H?Ol~ and 0.5% as PO3," (Lindsay 1979). Active uptake of phosphate—The process of metabolic assimilation of phosphate by Synechococcus is energy dependent is indicated by the substantial decrease in the rate of uptake when the experiments were performed in the dark. This result agrees with the results for some species of cyanobacteria (Avendano and Valiente 1994), but does not hold true for all species (Istvanovics et al. 1993). Phosphate uptake is not significantly decreased in the presence of the photosynthetic inhibitor DCMU, indicating that ATP is the likely energy source for active uptake of phosphate. Similar conclusions have previously been drawn for Cl~ (Ritchie 1992b, c), HCO3" (Ritchie et al. 1996) and S O ^ Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 tive charge when the cells were supplied with phosphate. A similar, but smaller effect, was found in the presence of DCMU but there was no apparent polarization effect in the dark. This result is consistent with findings on the effects of DCMU and darkness on phosphate uptake. At pH o 10, in the light with and without DCMU there was a depolarization, not hyperpolarization of the membrane potential. As at pH 0 7.5, there was no apparent effect in the dark. In all cases, any electrical effect of phosphate is transitory: after a 120 minute incubation there is only a small or no difference in the membrane potential of cells with or without phosphate. Phosphate uptake in a cyanobacterium tion effects of added phosphate were detected in the light but not in the dark (Table 4) despite the inherent imprecision of using valinomycin expedited uptake of 86 Rb + to measure changes in membrane potential (Ritchie 1992a, Ritchie et al. 1996). DCMU did not abolish the polarization effects of added phosphate in the light. At pH o 10, where virtually all phosphate is present as HPO 2 ~ (Table 2C), added phosphate depolarizes the membrane potential indicating a net uptake of positive, not negative charge. At pH o 7.5, where substantial H2PO4~ is present (Table 2B), added phosphate hyperpolarizes the membrane potential. Complex charge balancing operations are involved in active uptake of phosphate. Mechanism of uptake of phosphate—Poolman et al. (1987) suggested that uptake of phosphate in Streptococcus was electroneutral, with a variable number of protons being cotransported with the phosphate depending on the valency of the phosphate being transported. Table 4 does not point to an electroneutral uptake mechanism. Tables 2A, 2B and 2C show that a proton-motiveforce-driven mechanism is very unlikely in Synechococcus. The pmfi0 at pH 0 7.5 is only about - 1 3 kJ m o P 1 (Table 2B) but at pH 0 10 the pmfio is positive, about 2 to 3 kJ m o P ' Table 2C). At pH 0 10, protons are actively taken up by the cells and the pmfio is far short of the —30 kJ m o P 1 required for active uptake of phosphate (Table 2C). A pmfio-driven pump would not be thermodynamically possible under alkaline conditions. Table 3 shows that phosphate uptake was clearly Na + dependent as found previously for C P and HCOf and other ions but this does not necessarily point to the electrochemical potential for Na + (4^Na£,) being the driving force for active uptake of phosphate (Ritchie 1992b, c, Ritchie et al. 1996). Avendano and Valiente (1994) found no effect of Na + upon the Km of phosphate uptake but they found a large effect on the rate of uptake. The AfiNa£0 for Synechococcus cells in the light is about —13 to —16 kJ moP 1 at pH o 7.5 and 10 (Ritchie 1992a, Ritchie et al. 1996). At least 4Na + would be needed to drive the uptake of one phosphate (Table 2B, C). Such an uptake mechanism would be strongly electrogenic leading to uptake of net positive charge for each phosphate taken up at both pH 0 7.5 and 10, contrary to the complex polarization effects shown in Table 4. Falkner et al. (1980), Budd and Kerson (1987) and others have proposed that phosphate uptake by Synechococcus is an ATP-driven primary transport system. The hydrolysis of an ATP molecule produces a driving force of approximately - 5 0 kJ mol" 1 (Reid and Walker 1983, Muchl and Peschek 1984) and so the phosphate uptake mechanism would need to have a stoichiometry of one phosphate taken up per ATP to be thermodynamically possible under all the conditions tested in the present study. For thermodynamic reasons a 1 PO 4 /ATP pump could not reduce exter- Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 (Ritchie 1996). Thermodynamic analysis clearly demonstrates that phosphate is actively taken up by the cells regardless of which form of phosphate is considered (Table 2B, C). All forms of phosphate are at least +25 kJ moP 1 from equilibrium and so a driving force of at least —25 kJ moP 1 is required to drive active uptake of phosphate. Tables 2A and 2B however, show that a proton-motive-force (pmfio) driven mechanism is very unlikely in Synechococcus. The pmfi0 at pH 0 7.5 is only about - 1 3 kJ moP 1 but at pH 0 10 the pmfio is positive, about 2 to 3 kJ moP 1 . At pH 0 10 protons are actively taken up by the cells and is far short of the + 3 0 k J m o P ' required for active uptake of phosphate (Table 2B). A pmfio-driven pump is conceivable at pH 0 7.5 but would need to have a stoichiometry of at least 4 H + / PO 4 (4 x —13 kJ moP 1 ) but such a mechanism would not be thermodynamically possible under alkaline conditions. Effect of ion deficiencies—The rate of phosphate uptake was significantly reduced when the cells were incubated in Na+-free media, as found previously by Avendano and Valiente (1994) and for C P transport (Ritchie 1992b, c) and HCO3~ transport (Ritchie et al. 1996) but contrasting with the lack of a Na + effect in the case of SOJ~ transport (Ritchie 1996). Avendano and Valiente (1994) also found a K+-effect comparable to that found in the present study. Comparison of Tables 1 and 3 show that the net phosphate uptake in Na+-free media in the light fell to the rates typically found in the dark. In contrast to the present study, Budd and Kerson (1987) reported that Na + was not required for phosphate uptake in Synechococcus; a result quite different to that found in the present study and by Avendano and Valiente (1994). Rigby et al. (1980) and Budd and Kerson (1987) also tested for the effects of K + , Ca 2+ and Mg 2+ upon phosphate uptake by cyanobacteria. Rigby et al. (1980) and Budd and Kerson (1987) both reported a stimulatory effect of Ca 2+ but not Mg 2+ . However, their results were probably not reliable, as inappropriate experimental media were used for their experiments. They first isolated their cells in Tris/H 2 SO 4 buffer (pH 8) then tested for the effects of added cations. Not only is Tris known to be toxic to cyanobacteria but the Tris/H 2 SO 4 buffer probably caused irreversible damage to the cells because of the lack of magnesium ions. We found that the damaging effect of Mg2+-free media was not readily reversible. Washing the cells in magnesium-free media was enough to disable the mechanism for phosphate uptake. Table 3 shows Ca 2+ had no apparent effect on phosphate uptake. Since the Mg2+-free BG-11 used in the present study contained Ca 2+ , K+ and Na + it can be concluded that none of these cations can substitute for Mg 2+ . Lack of Na + or K + inhibited phosphate uptake but since K+-free BG-11 contained Na + , apparently Na + will not substitute for K + . Polarization studies—Significant transient polariza- 1239 1240 Phosphate uptake in a cyanobacterium quence for one cyanobacterium, Synechocystis PCC 6803 is on the INTERNET as a genetic database called CYANOBASE (bibliography; http://www.pka3.kazusa.or.ip/ cyano/biblio.html, similarity search; http://www.kazusa. or.ip/cyanobase/kwd.html). It has a very large listing of putative ion transporters including phosphate transporters. The mechanism of action of some of these transporters has been identified from sequence homologues (Kaneko et al. 1996a, b). For example, one phosphate transporter has been identified as a primary ABC-type ATP-driven pump (pstB: sllO683; sll0684; slrl250). Two other phosphate transporters are known (pstA: sllO682; slrl249 and pstC: sll0681; sir 1248) whose mechanism of action are unknown. Periplasmic phosphate binding proteins have also been identified, one of which is associated with phosphate starvation (slrl247). The present physiological findings on Synechococcus are consistent with current genetic information on cyanobacteria. Dr. R.J. Ritchie holds an ARC Postdoctoral Research Fellowship. 32P-H3PO4 was purchased from the Australian Institute of Nuclear Science and Engineering (AINSE). Dr. Donelle Trautman was supported in part from an ARC Institutional Grant awarded to Dr R.J. Ritchie and Prof A.W.D. Larkum. References Allen, M.M. (1968) Simple conditions for growth of unicellular blue-green algae. / . Phycol. 4: 1-3. APHA (American Public Health Association) (1992) Standard Methods for the Examination of Water and Waste-water. American Public Health Association, Washington DC, U.S.A. Avendano, M.C. and Valiente, E.F. (1994) Effect of sodium on phosphate uptake in unicellular and filamentous cyanobacteria. Plant Cell Physiol. 35: 1097-1101. Budd, K. and Kerson, G.W. (1987) Uptake of phosphate by two cyanophytes: cation effects and energetics. Can. J. Bot. 65: 1901-1907. Carpenter, J.H. (1966) New measurements of oxygen solubility in pure and natural waters. Limnol. Oceanogr. 11: 264-277. Colt, J. (1984) Computation of Dissolved Gas Concentrations in Water as Functions of Temperature, Salinity and Pressure. Bethesda Publ., American, Fisheries Society Special Publication 14, Maryland, MY, U.S.A. Ouggleby, R.G. (1985) Estimation of the initial velocity of enzyme-catalysed reactions by non-linear regression analysis of progress curves. Biochem. J. 228: 55-60. Falkner, C , Falkner, R. and Schwab, A.J. (1989) Bioenergetic characterization of transient state phosphate uptake by the cyanobacterium Anacystis nidulans. Archiv. Microbiol. 152: 353-361. Falkner, C , Homer, W.F. and Simonis, W. (1980) The regulation of the energy-dependent phosphate uptake by the blue-green alga Anacystis nidulans. Planta (Berlin) 149: 138-143. Grillo, J.F. and Gibson, J. (1979) Regulation of phosphate accumulation in the unicellular cyanobacterium Synechococcus. J. Bacteriol. 140: 508517. Hamer, W.J. (1968) Theoretical Mean Activity Coefficients of Strong Electrolytes in Aqueous Solutions from 0 to 100 C. US Government Printing Office, National Standard Reference Data Series, National Bureau of Standards 24, United States Department of Commerce, Washington. Istvanovics, V., Pettersson, K., Rodrigo, M.A., Pierson, D., Padisak, J. and Colom, W. (1993) Cloeotrichia echinulata, a colonial cyanobacterium with a unique phosphorus uptake and life strategy. J. Plankton Res. 5: 531-552. Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 nal phosphate concentrations to much below the 1 mmol m~ 3 found for cells incubated in nominally P-free media. Freshly made up P-free media had an inorganic phosphate concentration well below 0.1 mmol m " ' and so cells lost phosphate until reaching a steady-state between uptake and loss. Table 2A suggests that the cells are unable to maintain their intracellular phosphate in the dark under low phosphate conditions because phosphate uptake is so strongly light-dependent. The stalling potential of the phosphate uptake mechanism is about —40 kJ mol" 1 or a thermodynamic efficiency of about 80%. Compared to the magnitude of the fluxes of other ions such as sodium (Ritchie 1991, 1992a) and bicarbonate fluxes (Ritchie et al. 1996) both the isotopic and net fluxes of phosphate are very low and so not likely to impose a large metabolic cost to the cells. If it is assumed that one ATP is used to take up one phosphate, then the power consumption of a normal phosphate uptake flux (Table 1) would be 5 8 6 ± 6 5 p m o l m ~ 2 s " ' x 5 0 k J m o r ' = 0 . 0 2 9 ± 0.003 mW m~2 (Ritchie et al. 1996). This power consumption is very small compared to the requirements of Na + and HCO^" transport and the power available from photosynthesis (Ritchie 1992a, Ritchie et al. 1996). Not unexpectedly, oxygen electrode experiments have failed to find large effects on respiration or photosynthesis of added phosphate to cells prepared in P-free media in the short term. Phosphate requirements of exponentially growing Synechococcus—The progressively changing nutrient environment of a batch culture is not a realistic model for an algal-bloom being continuously fed new nutrients from the stream-flow. Chemostats can be used as laboratory models of cyanobacterial blooms in streams receiving treated effluent from sewage treatment plants (Tempest 1970). Table 1 shows that Synechococcus consumes about 600pmolm~ 2 s~' of phosphate to sustain optimum exponential growth in a medium with a phosphate concentration comparable to the effluent of a typical secondary sewage plant (50 mmol m~ 3 ). In practical situations, an exponentially growing Synechococcus culture is capable of assimilating nearly all phosphate offered to it. However, the cells are not capable of pulling dissolved phosphate down to undetectable levels because at about 1 mmol m~3 external phosphate an ATP-driven pump is close to its thermodynamic limits in maintaining intracellular phosphate levels (Table 2B, C). In continuous culture, Synechococcus has an intracellular pool of total phosphate equivalent to 400 mol m~ 3 : this store of phosphorus would be able to supply cells with adequate phosphate for about 6 generations or at least 3 days. Cyanobacteria are notorious for scavenging phosphate and storing it as polyphosphate which takes several cell generations to deplete after environmental phosphate has been exhausted (Falkner et al. 1989). Genetic database of Synechocystis—The complete se- Phosphate uptake in a cyanobacterium PCC 7942. J. Plant Physiol. 139: 320-330. Ritchie, R.J. (1992b) The cyanobacterium Synechococcus R-2 (Anacystis nidulans, S. leopoliensis) PCC 7942 has a sodium-dependent chloride transporter. Plant Cell Environ. 15: 163-177. Ritchie, R.J. (1992c) Kinetics of chloride transport in the cyanobacterium Synechococcus R-2 (Anacystis nidulans) PCC 7942. Plant Cell Environ. 15: 179-184. Ritchie, R.J. (1996) Sulphate transport in the cyanobacterium Synechococcus R-2 (Anacystis nidulans, S. leopoliensis) PCC 7942. Plant Cell Environ. 19: 1307-1316. Ritchie, R.J. and Gibson, J. (1987) Permeability of ammonia, methylamine and ethylamine in the cyanobacterium, Synechococcus R-2 (Anacystis nidulans) PCC 7942. J. Membr. Biol. 95: 131-142. Ritchie, R.J., Nadolny, C. and Larkum, A.W.D. (1996) Driving Forces for bicarbonate transport in the cyanobacterium Synechococcus R-2 (Anacystis nidulans, S. leopoliensis) PCC 7942. Plant Physiol. 112: 1573-1584. Ritchie, R.J. and Prvan, T. (1996a) A simulation study on designing experiments to measure the Km and Vmix of Michaelis-Menten kinetics curves. /. Theor. Biol. 178: 239-254. Ritchie, R.J. and Prvan, T. (1996b) Current statistical methods for estimating the Km and Kmu of Michaelis-Menten kinetics. Biochem. Ed. 24: 196-206. Suttle, C.A., Cochlan, W.P. and Stockner, J.G. (1991) Size-dependent ammonium and phosphate uptake, and nitrogen:phosphate supply ratios in an oligotrophic lake. Can. J. Fish. Aquat. Sci. 48: 1226-1234. Tempest, D.W. (1970) The continuous cultivation of microorganisms: I. Theory of the chemostat. Methods Microbiol. 2: 259-276. Valiente, E.F. and Avendano, M.C. (1993) Sodium-stimulation of phosphate uptake in the cyanobacterium Anabaena PCC 7119. Plant Cell Physiol. 34: 201-207. Wagner, F., Falkner, R. and Falkner, G. (1995) Information about previous phosphatefluctuationsis stored via an adaptive response of the highaffinity phosphate uptake system of the cyanobacterium Anacystis nidulans. Planta (Heidelberg) 197: 147-155. Walker.D. (1990) The Use of the Oxygen Electrode and Fluorescence Probes in Simple Measurements of Photosynthesis. University of Sheffield, Robert Hill Institute, Sheffield, UK. Young, H.D. (1962) Statistical Treatment of Experimental Data. McCrawHill Publ., New York. Zar, J.H. (1974) Biostatistical Analysis. Prentice-Hall Publ., Englewood CUffs, NJ, U.S.A. (Received June 16, 1997; Accepted September 4, 1997) Downloaded from pcp.oxfordjournals.org by guest on September 13, 2011 Johnson, M.L. and Faunt, L.M. (1992) Parameter estimation by leastsquares methods. Methods Enzymol. 210: 1-37. Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., Nakamura, Y., Miyajima, N., Hirosawa, M., Sugiura, M., Sasamoto, S., Kimura, T., Hosouchi, T., Matsuno, A., Muraki, A., Nakazaki, N., Namo, K., Okumura, S., Shimpo, S., Takeuchi, C , Wada, T., Watanabe, A., Yamada, M., Yasuda, M. and Tabata, S. (1996a) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions. DNA Res. 3: 109-136. Kaneko, T., Sato, S., Kotani, H., Tanaka, A., Asamizu, E., Nakamura, Y., Miyajima, N., Hirosawa, M., Sugiura, M., Sasamoto, S., Kimura, T., Hosouchi, T., Matsuno, A., Muraki, A., Nakazaki, N., Naruo, K., Okumura, S., Shimpo, S., Takeuchi, C , Wada, T., Watanabe, A., Yamada, M., Yasuda, M. and Tabata, S. (1996b) Sequence analysis of the genome of the unicellular cyanobacterium Synechocystis sp. strain PCC6803. II. Sequence determination of the entire genome and assignment of potential protein-coding regions (supplement). DNA Res. 3: 185-209. Lindsay, W.L. (1979) Chemical equilibria in soils. Wiley-Interscience Publ., New York. Mohleji, S.C. and Verhoff, F.H. (1980) Sodium and potassium ions effects on phosphorus transport in algal cells. /. Water Pollut. Control Fed. 52: 110-125. Muchl, R. and Peschek, G.A. (1984) Valinomycin pulse-induced phosphorylation of AOP in dark anaerobic cells of the cyanobacterium Anacystis nidulans. Curr. Microbiol. 11: 179-182. Nobel, P.S. (1983) Introduction to Biophysical Plant Physiology. Freeman Publ., San Francisco. Poolman, B., Nijssen, R.M.J. and Konings, W.N. (1987) Dependence of Streptococcus lactis phosphate transport on internal phosphate concentration and internal pH. J. Bacteriol. 169: 5373-5378. Reid, R.J. and Walker, N.A. (1983) Adenylate concentrations in Chara: variability, effects of inhibitors and relationship to cytoplasmic streaming. Aust. J. Plant Physiol. 10: 373-383. Rigby, C , Craig, S.R. and Budd, K. (1980) Phosphate uptake by Synechococcus leopoliensis (cyanophyceae): enrichment by calcium ion. J. Phycol. 16: 389-393. Ritchie, R.J. (1991) Membrane potential and pH control in the cyanobacterium Synechococcus R-2 (Anacystis nidulans) PCC 7942. /. Plant Physiol. 137: 409-418. Ritchie, R.J. (1992a) Sodium transport and the origin of the membrane potential in the cyanobacterium Synechococcus R-2 (Anacystis nidulans) 1241