SEARCH

SEARCH BY CITATION

Keywords:

  • hippocampus;
  • inorganic;
  • neocortex;
  • phosphate;
  • precipitation;
  • zinc

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Inorganic phosphate (Pi) is an important polyanion needed for ATP synthesis and bone formation. As it is found at millimolar levels in plasma, it is usually incorporated as a constituent of artificial CSF formulations for maintaining brain slices. In this paper, we show that Pi limits the extracellular zinc concentration by inducing metal precipitation. We present data suggesting that amino acids like histidine may counteract the Pi-induced zinc precipitation by the formation of soluble zinc complexes. We propose that the interplay between Pi and amino acids in the extracellular space may influence the availability of metals for cellular uptake.

Abbreviations used
ACSF

artificial CSF

AM ester

acetoxymethyl ester

EPSP

excitatory postsynaptic potential

ICP–OES

inductively coupled plasma–optical emission spectroscopy

PFS

phosphate-free saline

Pi

inorganic phosphate

An essential step in the process of unraveling the mechanisms of action potential generation, synaptic transmission, and muscular contraction was identifying the nature of the ionic species necessary for the operation of excitable cells (Burton 1975). This began with Ringer’s demonstration (Ringer 1883) that extracellular calcium was necessary for cardiac contraction and set in motion the process of formulating media used to sustain cells in vitro. An important step was Krebs and Henseleit’s (1932) realization that the constituents of normal plasma would be a good starting point for formulating physiological salines. Building on these foundations, McIlwain developed the in vivo brain slice preparation, one that has been enormously influential in the progress of neuroscience (Li and McIlwain 1957).

Inorganic phosphate (Pi, orthophosphate) is an essential ion in living organisms playing indispensable roles in ATP synthesis and bone mineralization, among other processes. Pi exists in two predominant forms at physiological pH; HPO42−and H2PO4 at ∼4 : 1 ratio. Intracellular Pi is sustained at a concentration of about 2 mM in most mammals and is a key determinant of the free energy available from ATP hydrolysis (Erecinska and Silver 1989). Plasma Pi levels vary considerably in different vertebrates (Furman et al. 1997). In human plasma, the normal Pi level is ∼1.1 mM but fluctuates more widely than calcium, and exhibits circadian variations. The concentration of Pi in the CSF of mammals is ∼0.4 mM; however, little is known about the concentration in the interstitial space.

Little information is available on the mechanism of Pi uptake into cells of the CNS. A protein initially identified as a Pi transporter (Ni et al. 1994) was subsequently shown to carry glutamate into synaptic vesicles (Bellocchio et al. 2000; Takamori et al. 2000) and its Pi transporting capabilities are uncertain. Pi transport has been well characterized in the kidneys where it is transported by members of the Slc34 (Murer et al. 2004) and Slc20 (Collins et al. 2004) families for monovalent and divalent species, respectively, in sodium-dependent processes.

There are abundant opportunities for solid minerals to form from the complex mixtures of ions in and around cells, particularly between metals and polyanions like Pi. In a solution with known concentrations of ions it is possible to predict the formation of precipitates from the solubility products (Ksp) for the various ion combinations. As a thermodynamic parameter, the Ksps however give no indication of how fast the precipitate takes to form. One common form hydroxyapatite (Ca10(PO4)6(OH)2) takes a long time, others like hopeite (Zn3(PO4)2·4H2O) form within a few milliseconds.

Zinc is found at a high concentration in certain glutamatergic vesicles within the mammalian forebrain and it has been proposed to be released and act as a neuromodulator (Smart et al. 2004; Paoletti et al. 2009). There is a potential chemical impediment to the free release of zinc ions, namely, that zinc–phosphate has a very low solubility product (9.1 × 10−36 M5), which limits the concentration of free zinc ions in a solution with a high concentration of Pi. In this communication, we show that in brain slices the extracellular free zinc concentration is indeed limited by precipitation. Moreover, we demonstrate that this limitation can be overcome by the provision of amino acids like histidine that increases the solubility of the metal.

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Brain slices

All animal procedures were in accordance with the NIH Guide for the Care and use of Laboratory Animals and approved by the Institutional Animal Care and Use committee of the University of Iowa. Sprague–Dawley rats (21- to 60-day-old male) were decapitated, the brains were removed, and placed in ice-cold normal saline containing (in mM) 125 NaCl, 2.5 KCl, 2 CaCl2, 1.3 MgSO4, 25 NaHCO3, 1.25 NaH2PO4, and 10 glucose, bubbled with 95% O2–5% CO2. The composition of phosphate-free saline (PFS) was (in mM) 125 NaCl, 2.5 KCl, 2 CaCl2, 1.3 MgSO4, 26 NaHCO3, and 10 glucose bubbled with carbogen. To avoid zinc contamination, high purity reagents were used, while avoiding metals, glass, and plastics that can leach zinc into solutions (Kay 2004). Slices were cut at a thickness of 500 μm on a McIlwain tissue chopper, and held in an interface chamber at 20–24°C for at least 1 h prior to loading with the zinc indicator. FluoZin-3 acetoxymethyl (AM) (5 μM) was applied to slices together with 0.5% DMSO and 0.04% w/v pluronic for an hour.

Imaging brain slices

The slices were stabilized with a U-shaped stainless steel piece cross strung with nylon fibers in a temperature controlled chamber (RC-27L; Warner Instruments). Solutions were provided by a gravity feed system controlled by a VC-6 six channel valve controller (Warner Instruments, Hamden, CT, USA). Images were acquired on an Olympus Optical (Center Valley, PA, USA) BX50WI upright microscope. Illumination was provided by a monochromator set at 480 nm (T.I.L.L. Photonics, Gräfelfing, Germany), passed through a dichroic (Q495lp; Chroma Technology, Rockingham, VT, USA), and then through a filter (HQ530/60; Chroma Technology) onto the faceplate of a Princeton Instruments cooled charge coupled device (CCD) camera (Trenton, NJ, USA). Data were acquired by the MetaFluor program (Universal Imaging Corporation, Downingtown, PA, USA) and images were analyzed using ImageJ (NIH, Bethesda, MD, USA). No black-level adjustment was applied to images. The fluorescence intensity was expressed as %ΔF/F0 = % (F − F0)/F0, where F is the fluorescence intensity and F0 the fluorescence intensity at time zero. The dark current of the charge coupled device (CCD) was subtracted from all intensity measurements. All data are expressed as mean ± SEM.

Synaptic responses

Bipolar tungsten stimulating electrodes (∼4 MΩ; A-M Systems Inc., Carlborg, WA, USA), coupled to a stimulus isolation unit (AMPI, Jerusalem, Israel) were placed in the upper half of the stratum radiatum, and the field potentials were recorded with glass electrodes (2–5 MΩ) coupled to an Axoclamp-2B (Axon Instruments, Union City, CA, USA), connected to a low-noise amplifier (SR 560; Stanford Research Systems, Sunnyvale, CA, USA) to a Digidata 1322A A/D (Axon Instruments) using Clampex 9.2 software (Molecular Devices, Union City, CA, USA). Field EPSPs were stimulated with a pulse of 60 μs duration and the current was adjusted to give a response approximately 60% of the maximal amplitude. Slices that did not produce a maximum field excitatory postsynaptic potentials (EPSPs) amplitude > 1 mV were discarded.

Light scattering

The 90° light scattering (excitation 480 nm and emission 484 nm) was determined in a Hitachi F-4500 spectrofluorimeter (Naperville, IL, USA) in a rapidly stirred methacrylate cuvette whose temperature was controlled by a circulating water bath (32°C). The solution was continuously bubbled with 95% O2–5% CO2.

Inductively coupled plasma–optical emission spectroscopy

To measure the amount of zinc precipitated, solutions were prepared with different concentrations of zinc, centrifuged at 15 500 g for 30 min at 32°C, and the zinc concentration of the supernatant was then determined on Varian 720 (Palo Alto, CA, USA) inductively coupled plasma–optical emission spectroscopy (ICP–OES). Yttrium was used as an internal standard. Standard curves were constructed using zinc solutions of 10, 30, and 50 μM. The following lines were used to establish calibration curves and calculate the zinc concentration: 202.548, 206.200, and 213.857 nm; internal standard: Y line 371.021 nm.

Particle size distributions were measured on a Malvern Instruments Zetasizer (Malvern, UK), Nano series; following parameters are applied to our experiments: reflective index for Zn3(PO4)2·H2O4 is 1.594 and absorption is 0.1.

Elemental analysis of precipitates

The energy dispersive X-ray spectroscopy analysis was performed on a Hitachi S-3400N variable pressure SEM equipped with a Bruker AXS Quantax X-ray microanalysis system (Madison, WI, USA). Quantitation was determined with standard-less analysis using the peak to background method.

Calculation of precipitate formation

Precipitate formation was calculated using the program minteqa2 (Allison 2003). The National Institute of Standards and Technology database of formation constants was used and all calculations were performed under 5% CO2 at 32°C.

Reagents

FluoZin-3 AM (Molecular Probes, Invitrogen, Eugene, OR, USA), N,N,N′,N′-tetrakis-(2-pyridylmethyl)-ethylenediamine (Fluka, St Louis, MO, USA), and other reagents were from Sigma (St Louis, MO, USA).

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Zinc precipitation is induced by Pi in normal artificial CSF

The provision of zinc to brain slices above a few micromolar poses something of a problem in most artificial CSF (ACSF) formulations, because they typically include Pi that will precipitate zinc and thus limit the availability of free zinc ions. Slices can be sustained in phosphate free solutions and appear to exhibit normal synaptic responses and cellular activity, however, the effects of withholding phosphate are as yet unknown. To increase the solubility of zinc while preserving Pi we added histidine to normal saline, which forms both mono- and bis-histidine complexes with zinc (Martell and Hancock 1996) but does not act as a neurotransmitter (Godfraind et al. 1973).

To detect the formation of precipitate in physiological solutions we used 90° light scattering in a cuvette within a spectrofluorimeter. In PFS, the addition of zinc up to a concentration of 330 μM led to no discernable precipitate formation, which is consistent with the high solubility of the zinc–bicarbonate complex. In normal saline, however, additions of zinc above about ∼10 μM led to the formation of a precipitate (Fig. 1a and b). Adding histidine increased the solubility of zinc in normal saline, prevented the formation of precipitate and led to the dissolution of existing precipitate (Fig. 1a and b).

image

Figure 1.  Precipitation of zinc by phosphate. (a) Formation of precipitates measured by light scattering in a spectrofluorimeter. Black line in the absence of histidine, red in its presence (200 μM). (b) Addition of histidine to ACSF after the addition of zinc. (c) The concentration of soluble zinc species in ACSF as a function of zinc added calculated by minteqa2 (solid colored circles) at different concentrations of histidine. Values measured with ICP–OES are shown by the open symbols; squares in ACSF and triangles in PFS. (d) Influence of Pi concentration on zinc solubility calculated with minteqa2.

Download figure to PowerPoint

The geochemical modeling program minteqa2 was used to calculate the concentration of soluble zinc species in different saline formulations (Allison 2003). The program takes into account all possible species that might form in a mixture and uses empirical thermodynamic parameters to calculate the equilibrium concentration of species, including any that might precipitate out of solution. The soluble zinc species considered are Zn2+, ZnOH+, Zn(OH)2, Zn(OH)3, Zn(OH)4−2, ZnCl+, ZnCl2, ZnCl3, ZnCl4−2, ZnOHCl, ZnSO4, Zn(SO4)2−2, ZnCO3, and ZnHCO3+. Fifty insoluble species were considered in the calculation and the only one that formed was Zn3(PO4)2·4H2O. Figure 1c shows estimates of the soluble zinc concentration in normal saline as a function of the added zinc concentration. In histidine-free saline the soluble zinc concentration could not exceed approximately 4 μM. The addition of progressively higher concentrations of histidine led to an increase in the concentration of soluble zinc species. For example with 100 μM zinc and 200 μM histidine, there was 24.4 μM of Zn.His, 17.8 μM of Zn.His2 and 2.84 μM of all the other soluble zinc species. As there is some uncertainty about the extracellular Pi concentration we calculated the solubility of zinc at different Pi concentrations and this is plotted in Fig. 1d.

To test the predications of the modeling program we assessed the solubility of zinc in ACSF using ICP–OES (inductively coupled plasma optical emission spectroscopy) with different zinc concentrations. The solutions were centrifuged at 12 000 g to remove any precipitate and the supernatant was analyzed by ICP–OES. The results are shown in Fig. 1c and there is an excellent correspondence between the solubility measured by ICP–OES (open symbols) and that estimated by minteqa2 (closed symbols).

In human CSF, the total amino acid concentration is ∼700 μM (∼500 μM glutamine) and the histidine concentration ∼12 μM (Davson et al. 1993; Wishart et al. 2008). Glutamine leads to little if any increase in the solubility of zinc compared with that in normal saline alone. In normal saline with 100 μM zinc and 500 μM glutamine, the free zinc concentration as determined by ICP–OES was 4.4 ± 0.2 (= 5) (5.9 μM from minteqa2) and was uninfluenced by the addition of more zinc. Citrate can chelate zinc and is found at a concentration of 200–400 μM in CSF (Wishart et al. 2008). With 400 μM citrate in normal saline the addition of 100 μM zinc only leads to a soluble zinc concentration of 13.9 μM (minteqa2 calculation).

The composition of the material that precipitated out after the addition of zinc to ACSF was determined by energy dispersive X-ray spectroscopy (Fig. 2b). Atomic percentages were 60.9 ± 4 O, 15.9 ± 0.7 P, 4.6 ± 0.3 Ca, and 18.7 ± 0.5 Zn (= 4, from two different samples), with trace amounts of K+ and Mg2+. The analysis of the precipitate is consistent with an amorphous compound composed largely of Zn3(PO4)2·4H2O (Hopeite) and CaZn2(PO4)2·2(H2O) (Scholzite).

image

Figure 2.  Analysis of the precipitates formed on the addition of zinc to normal saline. (a) Scanning electron microscopy of particles precipitated by the addition of zinc to ACSF (scale bar, 100 nm). (b) Energy dispersive X-ray spectroscopy of the precipitate.

Download figure to PowerPoint

The size distribution of the precipitate in normal saline was assessed by light scattering and found to be polydisperse. The mean particle size of the precipitate formed 1 min after zinc addition was 688 nm and the minimum particle size ∼200 nm. This is consistent with scanning electron microscopic images of the precipitate (Fig. 2a).

Pi reduces zinc influx into brain slices

To assess the availability of zinc in neuronal tissue we used zinc transport into rat brain slices loaded with the AM ester form of the zinc-sensitive fluorescent indicator FluoZin-3 (Gee et al. 2002). In such slices, the fluorescence was elevated slightly above the tissue autofluorescence, suggesting that there is some chelatable zinc within cells. Application of the membrane impermeant chelator Ca–EDTA did not lead to a substantial reduction in the signal, whereas the membrane permeable chelators diethyldithiocarbamate or N,N,N′,N′-tetrakis-(2-pyridylmethyl)-ethylenediamine induced a decrease in fluorescence, consistent with the intracellular location of the zinc-indicator complex (data not shown). FluoZin-3 exhibits little response to calcium or magnesium, even in the millimolar range, and any elevations of fluorescence intensity are only likely to arise from the formation of a zinc–FluoZin-3 complex (Zhao et al. 2008).

Numerous scattered puncta were visible when viewing the parenchyma of FluoZin-3 loaded neocortical slices with little evidence of perikaryal or nuclear staining. Staining was most evident on the outermost aspect of the slices extending ∼20 μm into the slice. From these images, it seems that FluoZin-3 enters synaptic vesicles rendering the endogenous zinc visible, which is consistent with the known ability of AM derivatives to load intracellular membranes (Thomas et al. 2000). To test this, hippocampal slices were labeled with FluoZin-3 AM, which led to a faint but clear demarcation of areas that are highlighted by the Timm’s stain, namely the hilus and stratum lucidum, showing that FluoZin-3 enters zinc-rich synaptic vesicles (Fig. 3).

image

Figure 3.  The hilus of a hippocampal slice loaded with FluoZin-3 AM. str. gr. –stratum granulosum (scale bar, 10 μm).

Download figure to PowerPoint

To measure the zinc influx we imaged neocortical slices around layer 4, which includes all cellular elements, viz. dendrites, somata, synaptic boutons, glia, and axons. In hippocampal slice experiments, no differences were found in recording from the pyramidal cell layer as opposed to recordings from areas outside of this. Although we cannot be certain about the compartment that zinc translocates into, in this communication we have simply employed it as an assay to monitor zinc passage into cells.

In normal saline, the application of 100 μM zinc led to a small elevation of fluorescence in FluoZin-3 AM loaded neocortical slices, consistent with the precipitation of zinc by phosphate (Fig. 4a). The co-application of histidine (200 μM) with 100 μM zinc boosted the metal influx. Histidine by itself in normal saline did not increase the fluorescence intensity (data not shown), which suggests that it does not appropriate enough zinc in the saline or slice to induce any metal influx. In contrast, the addition of 100 μM zinc in PFS induced a substantial metal influx that was not augmented by histidine (Fig. 4b).

image

Figure 4.  The influence of Pi and histidine on the influx of zinc into neocortical slices. In all experiments, the fluorescence was measured in slices loaded with FluoZin-3 AM in an area of interest roughly midway through the neocortex. (a) Influences of amino acids (200 μM) on zinc uptake in ACSF. The period of zinc (100 μM) and amino acid application is indicated by a horizontal bar. Control (con) no amino acid was added (= 4), d-Ala (= 4), l-His (= 41), and d-His (= 6). (b) Zinc uptake in PFS. Control (con, = 5) and l-His (= 5).

Download figure to PowerPoint

There is some evidence from experiments on intestine that zinc and histidine are cotransported into cells. To test this hypothesis we co-applied either l- or d-histidine with zinc. There was little difference between the augmentation of zinc transport by l- or d-histidine (Fig. 4a). This suggests that the amino acid is probably not interacting with a stereospecific transporter. Both l- and d-alanine, which form weaker complexes with zinc than histidine also augmented the zinc transport (Fig. 4a). In addition, transport was not inhibited by the presence of gylcyl-sarcosine which acts as a substrate for the dipeptide/tripeptide transporters, which have been suggested to carry metal–amino acid complexes (Conrad and Ahearn 2005). All these findings suggest that histidine is simply increasing the solubility of zinc in solution, rather than acting as a specific carrier.

To determine if the outcome of our experiments resulted from an edge artifact, because the probe only labeled the outer most aspect of the slice, we performed experiments where FluoZin-3 AM was injected into the depths of the slice. Under these conditions similar results were obtained to those shown in Fig. 4a (data not shown).

Although we did not systematically study the temperature dependence of zinc transport, no transport was evident at 24°C, while at 32°C it was. This suggests that the process has a high Q10, consistent with an active transport mechanism.

We also investigated the solubility and transport of cadmium because it has similar properties to zinc and appears to be able to translocate into some cells as a ‘rogue hitchhiker’ on zinc transporters (He et al. 2006). Cadmium also increases the fluorescence of FluoZin-3, though only to ∼30% of the level of zinc (Zhao et al. 2009). Unlike zinc, cadmium is not precipitated by Pi in normal saline; we found this both empirically and in theoretical calculations. Application of cadmium to slices first led to a decline and then to an increase in fluorescence. Histidine, however, did not increase the rate of cadmium transport, consistent with the solubility of cadmium in normal saline. The latency of the fluorescence increase induced by cadmium application was longer than that in zinc applications (∼2 vs. ∼1 min) (Fig. 5). This suggests that either the rate of cadmium transport is considerably slower than zinc or that cadmium displaces zinc from endogenous chelators like metallothionein and so induces an increase in fluorescence.

image

Figure 5.  Cadmium uptake in neocortical slices as measured by FluoZin-3. The horizontal gray bar represent the time period when Cd (100 μM) and l-His (200 μM) were applied (= 3 in both cases).

Download figure to PowerPoint

To determine the effect of removing Pi from ACSF on synaptic transmission we measured the field potentials evoked by stimulation of the Schaffer collaterals in the stratum radiatum of region CA1 of rat hippocampal slices. Over a period of 5.5 h, there was no decrease in the slope of the field EPSP (Fig. 6).

image

Figure 6.  Effect of Pi removal on fESPS evoked by stimulation of the Schaffer collaterals. The pathway was stimulated every 5 s with a 60 μs pulse. The graph plots the slope of the fESPS averaged over 1 min, normalized to the initial value, for three slices. Error propagation was carried through all the calculations. The red line indicates the time PFS flow commenced.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We have shown here that in brain slices, the presence of phosphate in physiological saline limits the concentration that soluble zinc can reach after metal application by the formation of zinc phosphate precipitates. In addition, we have shown that amino acids like histidine can increase zinc solubility in the presence of Pi, but they do not play a direct role in the transport of zinc. These results have clear implications for experiments where exogenous zinc is applied to brain slices. For example, it accounts for Molnar and Nadler’s (2001) observation that the presence of phosphate inhibited the action of exogenous zinc on GABA receptors. Our results also have perhaps less obvious implications for zinc release and uptake in vivo that we will discuss below. Moreover, our experiments have cast a spotlight on a rather underappreciated yet ubiquitious anion, Pi.

It is widely believed that synaptic zinc acts as a neuromodulator, being released during exocytosis and then diffusing into the synaptic cleft (Frederickson et al. 2005). In contrast, our laboratory has provided evidence which conflicts with this idea and instead we have suggested that rather than being released, zinc is presented to the extracellular space while bound to exocytosed vesicular proteins (Kay 2003; Kay and Tóth 2006). We have termed this scenario ‘externalization’ in contrast to simple release (Kay and Toth 2008). The presence of extracellular Pi poses a problem for the zinc release hypothesis, as precipitates could form if large amounts of zinc are released. Whether they do form will depend on the concentration of zinc chelators, small ligands, macromolecules, Pi, zinc, and pH. We cannot, however, exclude the existence of zinc chelators more powerful than those so far identified.

The concentration of glutamate within vesicles is estimated to be ∼300 mM, but after exocytosis it declines very rapidly as the molecule diffuses within the synaptic cleft (Barbour and Hausser 1997). Glutamate has a low affinity for zinc; nevertheless at high concentrations, it can solubilize zinc. For example, at a zinc concentration of 100 μM in normal saline if the glutamate concentration is above ∼30 mM no precipitate forms, however below this concentration progressive amounts of precipitate form with half the zinc being precipitated with ∼14 mM glutamate (minteqa2 calculations).

Histidine has been proposed to augment the transport of zinc by forming a 2 : 1 complex (Sivarama Sastry et al. 1960) that is either transported as a unit or hands zinc off to a transporter. Histidine has been found to facilitate zinc transport in erythrocytes (Aiken et al. 1992) and in the intestines of rats (Wapnir et al. 1983), trout (Glover and Hogstrand 2002), and lobsters (Conrad and Ahearn 2007). However, our results suggest that in brain slices histidine simply increases the solubility of zinc and does not serve to facilitate the transport of zinc.

Our finding that histidine does not augment zinc transport in PFS suggests that under these conditions there is little phosphate in the extracellular space of brain slices. Similarly, because histidine increases zinc transport in normal saline, this indicates that the amino acid concentration is rather low in slices. In human CSF, the total amino acid concentration is ∼700 μM (∼500 μM glutamine) and the histidine concentration ∼12 μM (Davson et al. 1993; Wishart et al. 2008). This concentration of histidine is too low to increase the solubility of zinc in the presence of Pi.

However, the fact that in normal saline histidine augments zinc transport does not imply that there are no amino acids in the extracellular space. It could be that the zinc–phosphate particles do not penetrate through the extracellular space, and too little amino acid is likely to leach out to solubilize it. It is likely that the amino acid levels in the extracellular space of brain slices are diminished as they diffuse into the bathing solution. However, there is evidence that extracellular glutamate and perhaps glycine levels remain elevated in slices (Sah et al. 1989).

It does not seem to have been widely appreciated that Pi is an essential component of the extracellular medium that goes beyond its role in pH regulation. If Pi is removed from saline bathing neocortical cells, the intracellular ATP and Pi levels remain stable for 30 min but both decline to ∼60% of control levels after an hour (Glinn et al. 1997). Furthermore, synaptosomes derived from chronically phosphate-deprived rats show an increase in cytosolic calcium and a decrease in ATP (Massry et al. 1991).

Controlled precipitation plays an important role in skeleton formation and other biomineralization processes (Dorozhkin and Epple 2002). On the other hand, the uncontrolled formation of insoluble aggregates plays a prominent role in the pathogenesis of atherosclerosis (Giachelli et al. 2001) and may do so in a number of neuropathologies, including Alzheimer’s disease, Huntington’s disease, and amyotrophic lateral sclerosis. We would like to suggest that the formation of inorganic precipitates could serve as a nucleus for the accretion of molecules and ions. However, if one is to implicate precipitation in a neuropathology it is important to exclude aggregates that arise during the analysis. In this regard, it is instructive to consider the case of entities termed ‘nanobacteria’ that on closer examination turned out to be calcium carbonate particles (Martel and Young 2008).

Aggregates of material with a predominantly inorganic basis have been described in a number of neuropathologies that have for the most part been ascribed to the precipitation of calcium. The development of calcifications within the brain has been noted in the case of ischemia and exitotoxicity (Mahy et al. 1999); no zinc was found in this case (Mahy N., personal communication). Calcifications have also been described in the case of Fahr’s syndrome in the basal ganglia (Bouras et al. 1996), spasmodic dysphonia (Simonyan et al. 2008), and in Urbach–Wiethe disease in the amygdala (Thornton et al. 2008); some zinc is found in the case of Fahr’s syndrome. It is also worth noting that excess intracellular Pi may lead to the precipitation of calcium as it does in the case of skeletal muscle sarcoplasmic reticulum, limiting calcium mobilization (Dutka et al. 2005).

Although Pi is usually incorporated in solutions used for sustaining brain slices, there are cases where its omission is seemingly without effect (Miles 1990), but this has not been studied systematically. In cultured neurons, many investigators leave out phosphate from the minimal solutions used to perform physiological experiments but preserve it in the media used to culture the cells.

The saline formulations used currently for sustaining brain slices are based on the concentrations of ions in plasma. The plasma concentration of Pi is around 1.1 mM in humans and tends to be higher in other animals. In rats, the plasma Pi concentration is 3.2 ± 0.1 mM and in CSF 0.47 ± 0.01 mM (Mulroney et al. 2004). The latter is close to the concentration in humans and it is likely but not certain that the extracellular concentration is similar. Most mammalian slice ACSF formulations have Pi at a concentration of around 1 mM; to mimic CSF more closely it may be worthwhile shifting to a concentration of ∼0.5 mM.

It is worth considering whether Pi might play roles other than that of the substrate for ATP synthesis. For example might it act as an allosteric regulator of ion channels and transporters? Moreover, the outward directed Pi gradient could be employed in carrier-mediated mechanisms to transport ions, although none have thus far been identified.

There is a pressing need for experiments to determine the range of concentrations of Pi in the extracellular space and to determine the effect of changes in Pi on neuronal and synaptic activity. NMR measurements in the intact brain can distinguish between extra and intracellular Pi and may provide a means for assessing the extracellular Pi concentration (Gilboe et al. 1998).

To the best of our knowledge, there have been no systematic studies of the effect of phosphate-free ACSF on synaptic transmission. We found that phosphate removal for up to 5.5 h had no effect on transmission at the Schaffer collateral CA1 synapse. This suggests that intact neuronal tissue either has considerable reserves of Pi or that it is endowed with an exceptional capacity to reclaim Pi that passes into the extracellular space. Until it becomes feasible to measure extracellular Pi, it is not possible to say whether or not Pi levels are sustained in slices held in PFS.

It is clear that the role of Pi extends beyond that of a pH buffering agent. There is little information on the mechanisms to control the intra and extracellular Pi levels, which could have a profound impact on the precipitation of metals. Moreover, the levels of amino acids, like histidine could play an important role in formation of metal–phosphate precipitates.

Perhaps because phosphate has been identified as part of the pH buffer system of physiological salines, it has not received much attention as a necessary component of media. Pi seems to have become part of the scenery when in fact it is a significant supporting player in cellular biology.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

We would like to thank Dr Sarah Larsen for the use of her ICP–OES and particle size analyzer, Daphne Lison, Drs Richard Miles, and Bill Shuttleworth for helpful comments on earlier versions of this manuscript; and also Drs Wei Chen, Pierre Paoletti, and Kristina Simonyan for their useful suggestions. This work was supported by grants from NINDS (NS47508 to ARK) and NIEHS through the University of Iowa Environmental Health Sciences Research Center, NIEHS/NIH P30 ES0560. IN was supported in part by a fellowship from the Center for Biocatalysis & Bioprocessing. We thank Jean Ross and Dr Jonas Baltrusaitis of the Central Microscopy Research Facility (funded by NSF grant CHE 0320387) for performing the elemental analysis on precipitates.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References