HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: 294/3/C726    most recent 00541.2007v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Colvin, R. A. Articles by Holmes, W. R. Search for Related Content PubMed PubMed Citation Articles by Colvin, R. A. Articles by Holmes, W. R.

NERVOUS SYSTEM CELL BIOLOGY

Insights into Zn²⁺ homeostasis in neurons from experimental and modeling studies

¹Department of Biological Sciences, ²Neuroscience Program, and ³Quantitative Biology Institute, Ohio University, Athens, Ohio; ⁴Graduate School of Engineering, Department of Materials and Life Sciences, Osaka University, Osaka, Japan; and ⁵The Mental Health Research Institute of Victoria and ⁶Department of Pathology, The University of Melbourne, Parkville, Victoria, Australia

Submitted 14 November 2007 ; accepted in final form 29 December 2007

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
To understand the mechanisms of neuronal Zn²⁺ homeostasis better, experimental data obtained from cultured cortical neurons were used to inform a series of increasingly complex computational models. Total metals (inductively coupled plasma-mass spectrometry), resting metallothionein, ⁶⁵Zn²⁺ uptake and release, and intracellular free Zn²⁺ levels using ZnAF-2F were determined before and after neurons were exposed to increased Zn²⁺, either with or without the addition of a Zn²⁺ ionophore (pyrithione) or metal chelators [EDTA, clioquinol (CQ), and N,N,N',N'-tetrakis(2-pyridylmethyl)ethylenediamine]. Three models were tested for the ability to match intracellular free Zn²⁺ transients and total Zn²⁺ content observed under these conditions. Only a model that incorporated a muffler with high affinity for Zn²⁺, trafficking Zn²⁺ to intracellular storage sites, was able to reproduce the experimental results, both qualitatively and quantitatively. This "muffler model" estimated the resting intracellular free Zn²⁺ concentration to be 1.07 nM. If metallothionein were to function as the exclusive cytosolic Zn²⁺ muffler, the muffler model predicts that the cellular concentration required to match experimental data is greater than the measured resting concentration of metallothionein. Thus Zn²⁺ buffering in resting cultured neurons requires additional high-affinity cytosolic metal binding moieties. Added CQ, as low as 1 µM, was shown to selectively increase Zn²⁺ influx. Simulations reproduced these data by modeling CQ as an ionophore. We conclude that maintenance of neuronal Zn²⁺ homeostasis, when challenged with Zn²⁺ loads, relies heavily on the function of a high-affinity muffler, the characteristics of which can be effectively studied with computational models.

muffler; zinc buffering; computational model; metal ion transport and homeostasis; metallothionein

Like other cells studied (37, 38), we have shown that neurons maintain a large intracellular buffer capacity for Zn²⁺, such that even micromolar Zn²⁺ loads result in only small changes in free intracellular Zn²⁺ concentration (6, 10). The sources of cellular Zn²⁺ buffering might include cytosolic binding proteins [e.g., the thionein/metallothionein (MT) pair] (37), binding to small molecules such as glutathione (40), sequestration into cytoplasmic organelles (35, 53), and eventually efflux (34, 49). Intracellular free Zn²⁺ is thought to be in equilibrium with many soluble cytosolic Zn²⁺ binding proteins. The metallothionein/thionein (MT/T) pair is a well-studied example of such a protein (37). The apoprotein can bind a total of seven Zn²⁺ ions in two domains, coordinated by several cysteine ligands. Unfortunately, the binding and release kinetics of Zn²⁺ with MT/T are still poorly understood, which, in turn, limits our understanding of its role as a cytosolic Zn²⁺ buffer and/or chaperone. Studies of binding and release kinetics of the MT-2 isoform have suggested that the two domains differ significantly in Zn²⁺ affinity (33, 39).

Two different Zn²⁺ transporter gene families [solute-linked carrier (SLC) 39 and SLC30] have been identified; each family probably has a unique transport mechanism, function, and cellular location. Expression of SLC39 gene family members [protein name: Zrt/Irt-like protein (ZIP)] has been observed in most eukaryotic organisms (for review, see Ref. 24). However, the transport mechanism(s) of mammalian SLC39 proteins has been only partially solved (22, 23). These studies provide convincing evidence that the SLC39A1 protein is normally targeted to the plasma membrane and mediates Zn²⁺ influx, although other family members are targeted to the Golgi (30). The transporter appears to function by simple facilitated transport, dependent on a concentration gradient to provide the free energy for net movement of Zn²⁺ into the cell.

Like the SLC39 gene family, SLC30 gene [protein name: zinc transporter (ZnT)] family members are found at all phylogenetic levels in eukaryotic organisms (24). The function, tissue location, and mechanism(s) of Zn²⁺ transport by SLC30 gene family members are best studied for the mammalian protein ZnT-1 (SLC30A1 gene) and its several homologs (49), where it was shown to be localized to the plasma membrane and presumably transports Zn²⁺ out of the cell (34). Unfortunately, these studies and others have not delineated a clear mechanism of transport. Many SLC30 family member proteins are preferentially localized to intracellular membranes, where they have more limited, tissue-specific expression, probably related to physiological intracellular Zn²⁺ sequestration (for review, see Ref. 15). SLC30A3 is required for the sequestration of Zn²⁺ into synaptic vesicles of glutamatergic neurons (48). Several recent studies demonstrate the localization of SLC30A5-7 proteins to the secretory pathway (in particular, the Golgi and vesicular compartment) and their importance in Zn²⁺ homeostatic mechanisms (16, 55, 56). Recent studies demonstrate also that Zn²⁺ can enter the matrix of brain mitochondria by mitochondrial uniporter-dependent and -independent mechanisms (43).

The first step in understanding the roles that Zn²⁺ plays in various brain pathologies requires a precise description of the neuronal Zn²⁺ metallome and normal mechanisms of Zn²⁺ homeostasis. To accomplish this goal, we have developed a computational model of neuronal Zn²⁺ homeostasis, highly constrained by experimentally derived parameter values. The model faithfully reproduced acute changes in intracellular free Zn²⁺ concentration observed in cultured rat cortical neurons when Zn²⁺ influx was increased. Experimental data have suggested that clioquinol (CQ) (a Zn²⁺-selective metal chelator with therapeutic potential in Alzheimer's disease) has Zn²⁺ ionophoric actions at the plasma membrane (13). By incorporating CQ with ionophoric activity into the model, the model was able to match observed changes in free intracellular Zn²⁺ observed after CQ addition.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Primary culture of embryonic cortical neurons. The care and use of animals in this study adhered to the American Physiological Society's Guiding Principles in the Care and Use of Animals. All experimental procedures were approved by the Ohio University Institutional Animal Care and Use Committee. Primary culture of embryonic (E-17-18) cortical neurons from Sprague-Dawley rats was performed as described previously (10). After 24 h, the plating media was replaced with fresh neurobasal (NB) media (Gibco BRL), supplemented with 0.5 mM glutamine and 2% B-27 (Gibco BRL). Media was subsequently changed as needed.

Fluorescence measurements from cortical neurons attached to glass coverslips. Pluronic acid (20% in DMSO) (Invitrogen) was mixed 1:1 (vol/vol) with 10 mM ZnAF-2F (diacetyl, cell-permeant form) (28), dissolved in DMSO (Sigma). The resulting mixture was diluted to 5 µM in Locke's buffer pH 7.4 (154 mM NaCl₂, 5.6 mM KCl, 2.3 mM CaCl₂, 1 mM MgCl₂, 5 mM HEPES, and 10 mM glucose). Glass coverslips with attached cortical neurons were incubated for 30 min at 37°C in a humid 5% CO₂ incubator with 5 µM ZnAF-2F. The coverslips were inserted into a coverslip holder, rinsed once with low-calcium (loCa) Locke's buffer, pH 8, for 1 min with gentle stirring. After rinsing, the coverslip holder and coverslip were moved to a new cuvette with fresh loCa Locke's buffer. Buffers were bubbled with O₂ (to remove dissolved CO₂) and then adjusted to pH 8, because these conditions give maximum Zn²⁺ uptake (6). To minimize effects of extracellular Ca²⁺ entering the neurons and possibly affecting intracellular ZnAF-2F fluorescence, extracellular Ca²⁺ was reduced to 0.5 mM. This concentration was chosen by adding ionomycin to cultured neurons in the presence of decreasing extracellular Ca²⁺ (without addition of Zn²⁺) until no observable effect on intracellular ZnAF-2F fluorescence was observed (data not shown). Continuous readings of cellular fluorescence intensity (counts per second using the FluoroMax-3 spectrofluorometer, Horiba Jobin Yvon) were obtained using excitation 492 nm and emission of 516 nm. Data presented as rate of fluorescence increase (F/F₀) were calculated as described previously (9). Experimentally obtained fluorescence intensity changes could be expressed as fractional saturation of ZnAF-2F by utilizing data obtained from the addition of pyrithione followed by N,N,N',N'-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) at the conclusion of an experiment. ZnAF-2F fluorescence after pyrithione was used as 100% saturation, and fluorescence after TPEN addition was used as 0% saturation.

Image capture and analysis. Cells were examined using conventional epifluorescence using an inverted microscope (Nikon, Diaphot 300) equipped with a Nikon PlanApo 60/1.40 oil differential interference contrast objective. The coverslips were placed in a sealed perfusion chamber (RC-30, Warner Instruments) that was held at a constant 37°C, and image capture and analysis were performed as described previously (41).

Inductively coupled plasma-mass spectrometry analysis and ⁶⁵Zn²⁺ uptake experiments. Cortical neurons after various treatments were washed three times with loCa Locke's buffer, pH 8, containing 100 µM EDTA and then moved to a new plate on ice, where they were aspirated dry and finally frozen at –70°C at least overnight. Later the plates were brought to room temperature, and, to each well, 225-µl Chelex (Sigma) washed 50 mM HEPES, pH 7.4, was added, and the coverslips were scraped clean to suspend the cellular debris. Total protein was determined by the Bio-Rad protein assay reagent using bovine serum albumin as a standard. The remaining cell suspension was desiccated in a centrifuge under vacuum at 60°C (Vacufuge, Eppendorf) for inductively coupled plasma-mass spectrometry (ICP-MS) analysis. Measurements were made using an UltraMass 700 (Varian), as described previously (61).

For ⁶⁵Zn²⁺ uptake, cortical neurons were treated identically, except that ⁶⁵Zn²⁺ (Brookhaven National Laboratory) was added. Depending on the experiment, ⁶⁵Zn²⁺ was mixed with non-radioactive Zn²⁺ to obtain a final specific activity between 0.001 and 0.005 µCi/µl. The specific activity of each reaction buffer was determined by assaying an aliquot for radioactivity. For calculation purposes, specific activity was expressed as counts per minute per nanomole Zn²⁺. ⁶⁵Zn²⁺ was determined in a liquid scintillation counter. Using the specific activity and protein, raw counts per minute were converted to nanomoles per milligram for each well and then could be compared directly to total Zn²⁺ content determined by ICP-MS.

Determination of intracellular ZnAF-2F concentration. Determination of intracellular fluorophore concentration usually involves loading cells, then lysing the cells to release fluorophore and comparing maximal fluorescence of the released fluorophore to a standard curve (e.g., Refs. 12, 38). We used a similar method; however, we found it was not necessary to lyse the cells. Neurons attached to coverslips were preloaded with 5 µM ZnAF-2F, washed, and then exposed to 5 µM pyrithione/30 µM Zn²⁺ [maximum fluorescence (F_max)] followed by 200 µM TPEN [minimum fluorescence (F_min)]. The resulting cellular F value (F_max – F_min) was compared with a calibration curve of ZnAF-2F fluorescence obtained in a cuvette using serial dilutions of ZnAF-2F at saturating Zn²⁺ concentrations with the same instrument settings and sensitivity as used for neurons attached to a coverslip. The intracellular concentration of ZnAF-2F obtained from the calibration curve only needed to be corrected for the actual volume of the neurons attached to the coverslip illuminated by the excitation beam. The area of the coverslip illuminated by the excitation beam was determined by placing a white index card in the beam and marking its size on the card. The volume of the neurons filling this area on a typical coverslip was estimated using 800 µm³ for the volume of a single cortical neuron soma and the average neuron density on a coverslip (from microscopic analysis). The concentration was then corrected for the difference between the volume of the cells and volume illuminated by the excitation beam in a cuvette. The calculated value was slightly less than the loading concentration of 5 µM (see Table 2).

Modeling methods. Three models of Zn²⁺ homeostasis with increasing complexity were used in the analysis. The simplest model, the buffer model, contains only reactions describing Zn²⁺ transport, binding to fluorophore, and Zn²⁺ buffering:

(model 1)

where Zn_o is extracellular Zn²⁺; Zn_i is intracellular Zn²⁺; F and FZn_i are the concentrations of free and bound fluorophore, respectively; and S and SZn_i are the concentrations of free and bound buffer, respectively. The second model, the muffler model, adds reactions to the buffer model for Zn²⁺ "muffling" and sequestration into a "deep store" and Zn²⁺ leak from the deep store:

(model 2)

where DSZn_i is the concentration of Zn²⁺ in the deep store. In the third model, the MT as muffler model, it is assumed that MT is the part of the "muffler" that sequesters Zn²⁺ into the deep store, and reactant, S, is a simple buffer, as in model 1 above.

(model 3)

MT reactions for the seven Zn²⁺ binding sites are modeled in more detail as described below.

In the experiments, fluorophore fluorescence was monitored under a number of conditions, before or after exposure to increased extracellular Zn²⁺, including the addition of pyrithione, EDTA, TPEN, and CQ, and the model was extended to incorporate these conditions when necessary, as described below. The chemical reactions were modeled with standard differential equations that were solved with Gepasi 3 (44), Copasi (29), and MATLAB (The MathWorks, Natick, MA). Rate constant and initial concentration values for the base models are collected in Tables 1 and 2.

Zn²⁺ transport. Zn²⁺ transport, represented above simply as Zn_o Zn_i, was modeled with reversible Michaelis-Menten kinetics with V_max and K_m being the same for influx and efflux. The value for K_m was measured experimentally as reported below. V_max was fixed from the increase in total Zn²⁺ content measured experimentally after 30 min of exposure to 10 µM extracellular Zn²⁺. In most simulations, extracellular Zn²⁺ was 30 µM (or far larger than the intracellular concentration), and results were the same, whether or not we used standard or reversible Michaelis-Menten kinetics to describe transport because efflux was negligible. Efflux came into play only in the few simulations when EDTA was added to the bath, and in these cases precise parameter values for the reverse path were not critical.

Binding to fluorophore. Zn²⁺ binding to the fluorophore ZnAF-2F, represented in the model as Zn_i + F FZn_i, is well characterized. The fluorophore has a K_d of 5.5 nM, and on and off binding rates at 25°C have been measured (28). Because our experiments were done at physiological temperature, we have increased these binding rate values, assuming a temperature coefficient (Q₁₀) value of 3. This proved necessary to match experimental results when EDTA was applied extracellularly.

Zn²⁺ muffling and leak from a deep store. Muffling is a term first used by Thomas et al. (58) to lump together the many additional processes that act to reduce the change in ion concentration. The muffling process is the most difficult to represent in the model, precisely because the actual muffling processes are not known. Several approaches are possible that include one or more "buffer equivalents" (26) and/or one or more stores, and we have examined some of the possibilities here.

In model 1, the muffler is a simple buffer. In model 2, the muffler is a buffer that binds Zn²⁺, and this buffer also transfers Zn²⁺ to a deep store as given by the SZn_i S + DSZn_i reaction. The deep store is an immobile unreactive pool of Zn²⁺ in the cell that may represent Zn²⁺ in mitochondria, the Golgi apparatus, other organelles, or various high-affinity metalloenzymes. Because the capacity of the deep store is finite, there must be a leak from the deep store, which we represent by DSZn_i Zn_i. The capacity of the deep store was determined experimentally as noted below, and this constrained parameter value choices because, in most simulations, capacity is approximately k_d1[S]_total/k_l, where [S]_total is [S] + [SZn_i] (where brackets denote concentration), and the constants are defined in Tables 1 and 2. In model 3, the muffler is composed of a buffer and MT. In this model, the buffer is a simple buffer, and it is MT alone that sequesters Zn²⁺ into the deep store. Deep-store Zn²⁺ capacity again provides constraints on parameter value choices; here it is the ratios of the MT sequestering rate constants, k_r51, k_r61, and k_r71, to the leak rate constant that are constrained by the deep-store Zn²⁺ capacity for a given MT concentration.

Binding to MT. MT can bind seven Zn²⁺ ions, with four sites binding more avidly than the other three. To represent this in the model, we consider different classes of binding sites and model the concentration of binding sites rather than MT concentration itself. In the model, the K_d for the first four binding sites was 1.6 pM. The K_d for the fifth, sixth, and seventh binding sites was 40 pM, 100 pM, and 20 nM, respectively (39). The individual on and off rate constants are not known, but were constrained by the experimental data. Typically, on rates were 80–100 x 10⁶ M^–1·s^–1 with off rates chosen according to the K_d value. MT concentration was 0.15, 0.5, 1, or 2 µM in the models, representing 1.05, 3.5, 7, or 14 µM of Zn²⁺ binding sites. In model 3, the three lowest affinity Zn²⁺ binding sites of MT could transfer Zn²⁺ to and from the deep store, with sequestration efficiency dependent on the K_d of the binding site.

Pyrithione as an ionophore. Pyrithione is known to be a Zn²⁺-selective ionophore, and it was often applied experimentally after the cells were exposed to high levels of extracellular Zn²⁺ for a period of time. We have not found it necessary to model its action in a manner more complicated than a first-order term, where the increase in intracellular Zn²⁺ over time is represented by k_pyr([Zn]_o – [Zn]_i), with the value of pyrithione rate constant k_pyr set to match the experimental data.

EDTA. In some experiments, 100 µM EDTA were added to remove extracellular Zn²⁺. EDTA can bind calcium and magnesium as well as Zn²⁺. Given K_d values for Zn²⁺, calcium and magnesium of 3.2 x 10^–11, 2 x 10^–5, and 1.76 x 10^–3 µM, respectively, and extracellular Zn²⁺, calcium, and magnesium concentrations of 30, 500, and 1,000 µM, respectively, we calculated that the addition of 100 µM EDTA would reduce [Zn]_o to 0.3 nM (in loCa Locke's solution). In the model, we set [Zn]_o to this value when EDTA was added in the experiment.

TPEN. TPEN is added to remove Zn²⁺ from the system. Unlike EDTA, TPEN will cross the cell membrane. In the model, we assume TPEN binding with Zn²⁺ in the extracellular space is at equilibrium, and, with its K_d for Zn²⁺ at 2.58 x 10^–16 M, addition of 200 µM TPEN to the extracellular space will make extracellular Zn²⁺ = 4.55 x 10^–17 M or essentially zero. When the experiment had 30 µM of extracellular Zn²⁺ before the addition of 200 µM TPEN, the model assumed that the concentrations of free and Zn²⁺-bound TPEN, [T]_o and [TZn]_o, were fixed at 170 and 30 µM, respectively. When TPEN was added after EDTA, we did not model the competition between the two for Zn²⁺; results were virtually identical, whether [T]_o and [TZn]_o were 170 and 30 µM or 200 and 0 µM to start. Free and Zn²⁺-bound TPEN cross the membrane according to their concentration gradients with rate constant k_tp, which is constrained by the rate of fluorophore fluorescence decay seen in the experiments. We use 0.1 s^–1 for k_tp and 80 x 10⁶ M^–1·s^–1 and 2.06 x 10^–8 s^–1 for the on and off rates for Zn²⁺ binding to intracellular TPEN, as these values provide the rapid decay time course observed experimentally.

CQ. CQ is a Zn²⁺ chelator that uses two molecules to bind one Zn²⁺ ion (18). When CQ is added extracellularly, we assume that binding with Zn²⁺ is at equilibrium before CQ crosses the cell membrane, and that, because the extracellular space is large compared with the intracellular space, the extracellular concentrations remain constant during the course of the experiment. Consequently, the constant extracellular concentrations of free and Zn²⁺-bound CQ, [CQ]_o and [CQ₂Zn]_o are:

(1)

where [CQ_t]_o is the total concentration of CQ added to the extracellular media, and K_CQ is the dissociation constant set to 100 nM (18). Because CQ is lipophilic, it readily enters the membrane. We assume that free and Zn²⁺-bound CQ cross the membrane, according to their concentration gradients, with a rate constant of 1.5 s^–1. We use 50 M^–1·s^–1 and 5 s^–1 for the on and off binding rate constants for Zn²⁺ binding to intracellular CQ. These values were chosen to fit the experimental data.

Statistical analysis. For image analysis, representative cells were selected, and cell bodies defined the limits of individual regions of interest. Average pixel intensity for all regions was averaged to obtain mean ± SE. Where appropriate, data comparisons were analyzed by one-way ANOVA, followed by Tukey's multiple-comparison test. Differences in the means were judged significant when P 0.05.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
Experimental analysis of Zn²⁺ influx and increases in intracellular free Zn²⁺ in cultured neurons. Figure 1, A–D, shows representative images of cortical neurons exposed to various experimental conditions, and Fig. 1E shows the results of quantitative analysis of Fig. 1, A–D. Neurons loaded with ZnAF-2F in nominally zero extracellular Zn²⁺ (5-min exposure to 100 µM EDTA/loCa Locke's buffer; without Zn²⁺ addition) show a diffuse fluorescence in the cell body and nucleus with no evidence of compartmentalization (Fig. 1A). ZnAF-2F fluorescence appears also in cell processes, but this is not as readily apparent as cell body fluorescence in the images shown in Fig. 1. The same pattern of diffuse intracellular fluorescence is seen even after maximum Zn²⁺ uptake in the presence of 5 µM pyrithione/30 µM Zn²⁺ (Fig. 1B). In Fig. 1C, neurons were incubated with 30 µM Zn²⁺ only for 5 min, and an increase in intracellular fluorescence intensity was clearly visible as well. Finally, neurons treated with Zn²⁺ and pyrithione were subsequently treated with 200 µM TPEN, showing that the increase in intracellular fluorescence was completely reversed (Fig. 1D). These data verify the basic assumptions of the computational model of Zn²⁺ homeostasis; i.e., that intracellular ZnAF-2F and free Zn²⁺ are free to equilibrate, ZnAF-2F is not compartmentalized, ZnAF-2F does not show rapid photobleaching or leakage, and that newly entering Zn²⁺ is free to equilibrate with the resident pool of free Zn²⁺.

View larger version (55K): [in this window] [in a new window]   Fig. 1. Cortical neurons attached to glass coverslips were loaded with 5 µM ZnAF-2F, then placed in a cuvette containing low-calcium (loCa) Locke's buffer containing 100 µM EDTA (A), 30 µM Zn2+ plus 5 µM pyrithione (B), 30 µM Zn2+ (C), or same as in B, followed by 200 µM N,N,N',N'-tetrakis(2-pyridylmethyl)ethylenediamine (TPEN) for 5 min (D), then coverslips were moved to a perfusion chamber with the same buffer for live-cell imaging. Cell bodies were analyzed for average pixel intensity, and those data are shown in E. Each bar represents mean ± SE (n = 7 individual cells on one coverslip as pictured in A–D). A Tukey multiple-comparison test (GraphPad Prism, version 4.03) showed that, for any column pair, P < 0.001, except for EDTA vs. TPEN, where P > 0.05. Size bar = 10 µm.

View larger version (14K): [in this window] [in a new window]   Fig. 2. Cortical neurons, attached to coverslips, were treated with various conditions, as described below, for 30 min, then rinsed 3x with loCa Locke's buffer containing 100 µM EDTA, lysed by freezing at –80°C, scraped, and resuspended for Zn2+ analysis by inductively coupled plasma-mass spectrometry (ICP-MS; A) or treated as in A with the addition of 65Zn2+ (B), and lysate radioactivity was determined by liquid scintillation counting. Lysate protein was determined by Bio-Rad protein assay using BSA as a standard. Each bar represents mean ± SE; nos. in parentheses are the no. of replicates. An unpaired t-test showed that, for either A or B, Zn2+ uptake after pyrithione treatment was significantly increased (P < 0.01). C: cortical neurons attached to glass coverslips were removed from neurobasal (NB) media to loCa Locke's buffer for 1 min at 37°C. Coverslips were then treated for 30 min with 10 µM Zn2+ containing 65Zn2+. Coverslips were moved to loCa Locke's buffer containing 100 µM EDTA for various times, then briefly rinsed 3x in fresh loCa Locke's buffer/100 µM EDTA, and treated as described for B. Data are presented as means ± SE; nos. in parentheses are the no. of replicates.

View larger version (15K): [in this window] [in a new window]   Fig. 3. A: cortical neurons loaded with 5 µM ZnAF-2F were treated with either 100 µM EDTA or 30 µM Zn2+; 5 µM pyrithione at 3 min and 200 µM TPEN at 6 min were added to loCa Locke's buffer. B: raw data from A with 30 µM Zn2+ added, expressed as percent saturation (see MATERIALS AND METHODS). F/F0, rate of fluorescence increase.

View larger version (34K): [in this window] [in a new window]   Fig. 4. Comparison of predictions using models 1, 2, and 3. In all cases, extracellular Zn2+ is 30 µM, pyrithione is applied at 3 min, and TPEN at 6 min (see Fig. 3B). Parameter values are given in Tables 1 and 2, except as noted below. A: model 1 results, with and without the constraint that total resting Zn2+ is 250 µM. With the constraint, buffer Kd is 5.1 nM, and starting S and SZni values are 1,216 and 249.28 µM, respectively. Results are shown when buffer rate constants (kb1, kb2) are either fast (dashed line; 10, 0.05122) or slow (dotted line; 0.08, 0.00041). Without the constraint, buffer Kd is 78 nM (solid line); parameter values are those for model 1 in Tables 1 and 2. B: model 2 predictions with and without metallothionein (MT) added (curves overlap) compared with model 1 results (buffer Kd = 78 nM) with kpyr = 0.052 or 0.15 [fast pyrithione rate (fp)]. C: MT as the exclusive muffler (model 3). Predictions are shown for MT concentrations of 0.15, 0.5, 1.0, and 2.0 µM. D: comparison of model 2 predictions with model 3 results for MT = 1.0 and 2.0 µM. E: a blow-up of the first 180 s showing the lack of linearity with various parameter value choices when MT = 0.5 or 1.0 µM compared with the MT = 2.0 µM case. Parameter values are as in Tables 1 and 2 for MT = 0.5, 1.0, and 2.0 µM cases. For cases labeled MT = 0.5a, MT = 1.0a, and MT = 1.0b, parameter values (kb2, kr52, kr62, kr72, kl) were 0.012, 0.005, 0.002, 1e–6, 0.0037; 0.009, 0.025, 0.01, 5e–6, 0.015; and 0.008, 0.005, 0.002, 1e–6, 0.015, respectively. In addition, kr51, kr61, and kr71 were 3, 3, and 3, respectively, for the MT = 0.5a case. Starting concentration values, similar to those in Table 2, were recomputed with these rate constant changes. F: net flux to the deep store from MT Zn2+ binding sites 5 (Kd = 40 pM), 6 (Kd = 100 pM), and 7 (Kd = 20 nM). Model 3 with MT = 2 µM.

View larger version (27K): [in this window] [in a new window]   Fig. 5. A: starting ZnAF-2F saturation is varied from 4 to 20%. Starting values for Zn2+, free muffler, bound muffler, deep store Zn2+, and total Zn2+ are 0.23 nM, 27 µM, 74 µM, 3.4 µM, 78 µM; 0.48 nM, 24.3 µM, 139 µM, 7 µM, 146 µM; 0.75 nM, 21.8 µM, 194 µM, 11 µM, 205 µM; 1.05 nM, 19 µM, 234 µM, 15.2 µM, 250 µM; and 1.4 nM, 16.1 µM, 256 µM, 19.7 µM, 277 µM for 4, 8, 12, 16, and 20% starting saturation levels, respectively. B and C: using model 2, data were derived for changes in intracellular free Zn2+ concentration as a function of time, with and without ZnAF-2F present when 30 µM extracellular Zn2+ was added. B: effects of Zn2+ addition followed by pyrithione for 3 min and then TPEN. C: changes for the first 3 min only. D: model 2-derived predictions for changes in intracellular free Zn2+ concentration after addition of 30 µM Zn2+ are plotted on the same graph as experimentally derived estimates using Eq. 2. ZnAF-2F Kd = 5.5 nM, maximum fluorescence (Fmax) = fluorescence measured after the addition 30 µM Zn2+ and pyrithione, and minimum fluorescence (Fmin) = fluorescence measured after addition of TPEN.

View larger version (15K): [in this window] [in a new window]   Fig. 6. Experimental data (solid line) and model 2-derived predictions (dotted line) plotted together on the same graphs. A: experimental data. Cortical neurons were loaded with 5 µM ZnAF-2F and treated with 30 µM Zn2+ for 3 min, and then 100 µM EDTA for 3 min, and finally with 200 µM TPEN. Data are expressed as F/F0 (left axis). B: experimental data. Cortical neurons were treated with 5 µM pyrithione and 30 µM Zn2+ for 3 min, and then as in A. Raw data were converted to %saturation (see MATERIALS AND METHODS). Model 2 predictions (A and B) are expressed as %saturation and scaled to match the experimental data (right axis in A). Each experimental plot is the mean of data obtained from 3 individual coverslips.

We observed a steady increase in ZnAF-2F fluorescence when neurons were placed in buffer with added Zn²⁺ (i.e., Zn²⁺ influx, Fig. 3A). We found that the F/F₀ was dependent on extracellular Zn²⁺ concentration, as would be expected for transporter mediated Zn²⁺ uptake. The F/F₀ (F/F₀ per second) was computed from linear regression analysis of the first minute of data collected. These rates were then plotted as a function of extracellular Zn²⁺ concentration and fit to a rectangular hyperbola to obtain K_m for transporter-mediated Zn²⁺ influx (K_m = 4.65 µM). This K_m value is consistent with previous kinetic measurements of Zn²⁺ transport mediated by SLC39A1 (23) in cultured cells. With the addition of pyrithione, ZnAF-2F fluorescence rapidly reached a maximum (additional increases in Zn²⁺ concentration did not increase fluorescence, data not shown). After TPEN addition, a new steady state in intracellular free Zn²⁺ was obtained, yielding a fluorescence slightly less than at the start of the reaction, identical to that obtained without addition of Zn²⁺ (Fig. 3A). In Fig. 3B, the experimental data are reported as percent saturation to allow direct comparison with model results. It can be seen in Fig. 3B that, after 3-min exposure to 30 µM Zn²⁺, ZnAF-2F was 40% saturated.

Constraints used for model development. Parameter values for fluorophore binding are known. Transport K_m and V_max were fixed from the experiments described above. Other parameter values were constrained by experimental results. Key experimental constrains reported above are as follows: 1) fluorescence at rest is 16% of maximum, on average; 2) total neuronal Zn²⁺ content is 250 µM at rest; 3) total Zn²⁺ content rises by an additional 320 µM when the cell is exposed to 10 µM extracellular Zn²⁺ for 30 min and reaches a maximum capacity of 1.5 mM when pyrithione is also present; 4) the percent saturation of the fluorophore rises from 16% at rest to 39% after 3-min exposure to 30 µM extracellular Zn²⁺; 5) the F/F₀ upon exposure to 30 µM extracellular Zn²⁺ is linear for 3 min, increases steeply, and saturates following pyrithione application and decays upon exposure to TPEN, as shown in Fig. 3. These constraints determine the resting bound and unbound concentrations of ZnAF-2F and constrain parameter value choices for the size and kinetics of the muffler in our three models.

The buffer model (model 1) cannot explain the experimental results. In the buffer model, there are only three unknown parameter values (total buffer concentration, on and off rates for Zn²⁺ binding to buffer), and the experimental constraints should be sufficient to allow all parameter values to be determined uniquely. However, no set of parameter values satisfied all of the constraints. Resting total Zn²⁺ level and Zn²⁺ capacity required a buffer with a K_d of 5.1 nM. Slow buffer on and off binding rate constants could match the 3-min fluorescence value, but the increase in fluorescence was very hyperbolic; fast buffer rate constants made the increase in fluorescence linear, but the 3-min fluorescence value was much too small (Fig. 4A). No set of rate constants could match both the linearity of the fluorescence increase and the value at 3 min found experimentally.

We then dropped the constraint that resting total Zn²⁺ was 250 µM, assuming that perhaps the model 1 reactions occur independently of a large unreactive pool of Zn²⁺ [e.g., the nucleosome (1)]. In this case, buffer K_d was 78 nM to match the total Zn²⁺ capacity and the 3-min fluorescence value (solid line in Fig. 4A; parameter values given in Tables 1 and 2). However, the model still did not match the experimental data very well. The fluorescence increase predicted by the model was never linear, typically having an initial delay and a later hyperbolic bend. Even more noticeable were the poor fits to the experimental data when pyrithione and TPEN were added. The fluorescence increase predicted by the model when pyrithione was added was far too small, unless the value for k_pyr was increased, but when k_pyr was increased, the fit after TPEN addition was poor (Fig. 4B). Finally, we note that there are a number of substances that bind Zn²⁺ with micromolar affinity, or an affinity much weaker than the 78 nM found here [e.g., glutathione (40)]. While a buffer with a 10-fold lower affinity can match the 3-min fluorescence value, this happens in the model only for a total Zn²⁺ capacity nearly 10-fold larger, in contradiction to the experimental constraint.

Insights and predictions from the muffler model (model 2). The muffler model (model 2) has seven free parameters: buffer (muffler) concentration, deep-store Zn²⁺ concentration, on and off muffler Zn²⁺ binding rates, on and off rates to the deep store, and leak from the deep store. While the experimental data could not fix all of the parameter values uniquely for this model, the data severely constrained parameter value choices, and sets of parameter values that satisfied all of the experimental constraints shared important features in common. First, the muffler had to have a high affinity for Zn²⁺ with a fast on-rate constant for Zn²⁺ binding. Good solutions were found with K_d values of 60–80 pM with on-rate constants of 80–100 x 10^–6 M^–1·s^–1. Lower affinity and slower on-rate constants introduced an early bump in the fluorescence increase, and higher affinity values produced a bowed curve, unlike the experimental data, which showed a linear increase in fluorescence in the initial 3 min of exposure to 30 µM of extracellular Zn²⁺ (Fig. 3B). Because of its high affinity, 210–240 µM of the 250 µM total Zn²⁺ in the cell at rest was bound to this muffler. Second, free muffler concentration was consistently between 15–25 µM for good solutions. This is a consequence of the choice of the K_d value for Zn²⁺ binding to the muffler. If the muffler K_d was greater than 120 pM, the amount of free muffler in the model had to be very high; ZnAF-2F fluorescence upon exposure to 30 µM extracellular Zn²⁺ then rose much more slowly than in the experiments, because Zn²⁺ would bind to the free muffler rather than to ZnAF-2F. Small muffler K_d values produced the opposite problem; the amount of free muffler was small and it became quickly overwhelmed when Zn²⁺ influx was increased. The rate constants for muffler interaction with the deep store could be tuned to compensate somewhat for these problems, but this introduced other problems with the solutions, such as bowing or an early bump in the fluorescence curve mentioned earlier. Third, the resting amount of Zn²⁺ in the deep store was 12–30 µM with parameter value choices that produced results matching the data. This forms a prediction of the amount of Zn²⁺ in mitochondria, Golgi, and other organelles at rest.

The muffler model results given so far assume that the 250 µM total Zn²⁺ in the neuron is all available for the reactions modeled, but, as mentioned above, it is possible that a significant portion of cellular Zn²⁺ may be so tightly bound or otherwise shielded, even from a muffler with a K_d of 60–80 pM, to make it unavailable to the model reactions (1). Thus simulations were done assuming that only 100 µM of the total Zn²⁺ were available to the model reactions. The major difference from the previous results was that the K_d of Zn²⁺ binding to the muffler that provided the best fits was larger (80–300 pM). For good solutions, the on-rate for Zn²⁺ binding still had to be large, and free muffler concentration and resting deep-store Zn²⁺ concentration were in the same ranges as before (12–27 and 14–30 µM, respectively).

Incorporating MT into the model (models 2 and 3). Because of the significant role MT is thought to play in Zn²⁺ homeostasis (37), we added MT as a Zn²⁺ buffer to the muffler model (model 2) to see what effect this would have on the results. We chose a concentration of 1 µM, yielding 7 µM of Zn²⁺ binding sites. This concentration is in excess of the 0.34 µM actually determined, to allow an effect to be seen more clearly, if there is one. Total muffler concentration had to be reduced by 7 µM to continue to satisfy the experimental constraints, but the overall effect on the results was negligible (curves entirely overlap in Fig. 4B). At rest, MT Zn²⁺ binding sites 1-6 were nearly saturated, and site 7 was largely free, and there was little change during exposure to 30 µM extracellular Zn²⁺ for 3 min, aside from a low level of Zn²⁺ binding to site 7.

Given this result, we turned to a third model, the MT as muffler model (model 3). In this model, the muffler of model 2 became a simple buffer (additional buffer capacity must be included to bind the 250 µM total resting Zn²⁺ in the neuron), and MT was postulated as the exclusive route for trafficking Zn²⁺ to the deep store. Results computed with different values for MT concentration are shown in Fig. 4C. There were a number of interesting features uncovered with these models. First, the results with model 3 approached those of model 2 (and also the experimental results) only for the largest values of MT concentration tested (Fig. 4D). Models with 0.15, 0.5, and even 1.0 µM of MT had trouble producing a linear increase in fluorescence over the first 3 min, as seen experimentally. Curves would bow, bend, or assume a sigmoid appearance, depending on the particular parameter values chosen (Fig. 4E). Such variations were particularly sensitive to buffer affinity. The reason for these deviations from linearity was that low MT concentrations were not sufficient to transfer Zn²⁺ to the deep store fast enough, forcing the buffer (and the fluorophore) to handle an increasing amount of the Zn²⁺ load. Making the transfer reaction arbitrarily fast in the model did not help, because the leak rate constant had to be increased as well because of the Zn²⁺ capacity constraint (see MATERIALS AND METHODS). In contrast, if MT concentration was 2.0 µM, the amount of Zn²⁺ that could be shuttled to the deep store in 3 min approached that in the muffler model (model 2), and linearity in the fluorescence increase was more easily obtained. Second, as can be seen with the parameter values in Table 2, Zn²⁺ binding sites 5 and 6 on MT are not saturated at rest in this model as they were when MT was a simple Zn²⁺ buffer. In fact, these sites range from being half saturated to being largely free of Zn²⁺ in these models (see Table 2; Fig. 4E legend); this is a result of the muffling reaction between these sites and the deep store. Third, even though site 7 has a lower affinity for Zn²⁺ than sites 5 and 6 and is always largely free of Zn²⁺ at rest, it turns out that the net Zn²⁺ flux to the deep store is greater from site 7 than from site 6 after the first minute, and both sites 6 and 7 had greater Zn²⁺ flux to the deep store than site 5 (Fig. 4F). This may be a consequence of our choice of rate constants (Table 1) that were based on the assumption that rates for the interaction of a Zn²⁺-bound MT Zn²⁺ binding site with the deep store should be related to the affinity of the individual binding sites for Zn²⁺. Also, as might be expected, net fluxes were much smaller when MT concentration was 0.15, 0.5, or 1.0 µM. Finally, we note that results were sensitive to the value of the on-binding rate of Zn²⁺ to MT relative to the Zn²⁺ on-binding rate to the buffer, perhaps reflecting competition for Zn²⁺, because the K_d of the buffer was similar to that of MT Zn²⁺ binding sites 5 and 6 (data not shown). Because the muffler model (model 2) is the best constrained model and it matches the experimental data well, we use this model for all further model predictions and comparisons to experimental data.

The intracellular free Zn²⁺ concentration. While the models assumed that the fluorophore was 16% saturated at rest, there was variation in this percentage among individual experiments. The muffler model was used to explore the consequences of these differences. An important experimental constraint for these simulations was that the slope of the fluorescence increase during the first 3 min was independent of the starting saturation percentage. Parameter values were found that satisfied the constraint that the total resting Zn²⁺ content is 250 µM for all starting saturation percentages, but the use of these values caused the slopes of the fluorescence increases to be different, in contradiction to the experimental data. To match the slopes found in the data, we let total Zn²⁺ content vary, and results are shown in Fig. 5A. The starting saturation percentage not only determined the intracellular free Zn²⁺ concentration as expected, but it also determined the total Zn²⁺ in the neuron, the total muffler concentration, and the total Zn²⁺ capacity. For example, if we compare starting saturation percentages of 8 and 20%, the resting intracellular free Zn²⁺ concentration is 0.5 and 1.4 nM, total resting Zn²⁺ content is 146 and 277 µM, total muffler concentration is 163 and 272 µM, free muffler concentration is 24.3 and 16.1 µM, and total Zn²⁺ capacity is 1 and 1.6 mM, respectively. The intracellular free Zn²⁺ concentration is maintained in a narrow range by comparatively large changes in total muffler concentration (which determines resting Zn²⁺ content and capacity), whereas free muffler concentration remains in a much more narrow range.

Next, the muffler model was used to predict how much the presence of ZnAF-2F (i.e., the Zn²⁺ buffering effect of ZnAF-2F itself) distorted the intracellular free Zn²⁺ concentration. The distortion was small, as shown in Fig. 5, B and C. Using model 2 (16% ZnAF-2F saturation at rest), the concentration of ZnAF-2F bound with Zn²⁺ was 758 nM. When ZnAF-2F was removed from the model and this 758 nM was added to the free Zn²⁺ concentration, the system reached a new equilibrium in which 47% of the 758 nM went to the deep store and 53% became bound to free muffler. The result was a 26 pM increase in the resting free Zn²⁺ concentration. Upon exposure to 30 µM extracellular Zn²⁺ for 3 min, the difference in free Zn²⁺ concentration with and without ZnAF-2F was 183 pM (Fig. 5C). Subsequent application of pyrithione increased free Zn²⁺ levels, but the difference with and without ZnAF-2F was only 4.5 nM at the peak (Fig. 5B). Thus the model predicts that cultured neurons have an intrinsic Zn²⁺ buffer capacity far in excess of that potentially added by 5 µM ZnAF-2F. The cytosolic buffer (working in concert with a deep store) is able to "absorb" nearly all of the Zn²⁺ that enters the neuron when Zn²⁺ influx is increased.

The Zn²⁺ buffer capacity of cultured cortical neurons, therefore, might be expected to be greater than that of HT-29 cells, which have been studied by Krezel and Maret (38). The presumed large buffer capacity of neurons should allow for the application of the equation (25):

(2)

to calculate the intracellular free Zn²⁺ concentration with reasonable accuracy (see Fig. 5, B and C). It can be seen that model predictions of intracellular free Zn²⁺ concentration changes and those calculated using the above equation agree quite well when ZnAF-2F is only partially saturated. As ZnAF-2F nears 100% saturation, the equation tends to significantly overestimate free Zn²⁺ concentration (Fig. 5D). This is largely because F_max estimated as addition of Zn²⁺ and pyrithione probably slightly underestimates the true F_max. This inaccuracy in estimating F_max becomes exaggerated as F approaches F_max, making the denominator in the above equation quite small. In other cell types with a much smaller intrinsic Zn²⁺ buffer capacity, intracellular fluorophores may act as significant competing cytosolic buffers and will "clamp" changes in intracellular free Zn²⁺ concentration (12, 45), such that the intracellular free Zn²⁺ concentration will largely depend on the fluorophore concentration, invalidating the above equation.

A kinetic analysis of Zn²⁺ efflux and model simulations of efflux. Neurons loaded with 5 µM ZnAF-2F were exposed to either 30 µM extracellular Zn²⁺ (Fig. 6A), or 30 µM extracellular Zn²⁺ and 5 µM pyrithione (Fig. 6B). Addition of 30 µM extracellular Zn²⁺ resulted in a roughly linear rise in ZnAF-2F fluorescence (Fig. 6A). Addition of pyrithione at the start of a reaction resulted in a much faster rate of Zn²⁺ influx, as shown by the rapid increase in ZnAF-2F fluorescence that quickly reached near saturation levels (Fig. 6B). After 3 min, 100 µM EDTA were added to block Zn²⁺ influx, thus allowing for unopposed Zn²⁺ efflux. The time course of the decrease in ZnAF-2F fluorescence after EDTA addition, with or without addition of pyrithione, appeared to follow a slow but steady decline. The observed changes in ZnAF-2F fluorescence were completely and rapidly reversed by the addition of TPEN. These findings are entirely consistent with our contention that the vast majority of newly transported Zn²⁺ (even in the presence of pyrithione) is buffered and intracellular free Zn²⁺ remains in the nanomolar range. The match between model 2 predictions and the experimental data was quite good (see Fig. 6). The model predicted an initial small but rapid decline followed by a much slower but steady decline after addition of EDTA. The small, rapid decline was caused by the fast redistribution of Zn²⁺ between intracellular ZnAF-2F and the muffler once influx was stopped. This small decline could be converted into a large decline (or even a large increase) in the model by scaling the ZnAF-2F rate constants. The subsequent longer, slower decline reflects Zn²⁺ concentration gradient-dependent Zn²⁺ efflux, modeled as a simple reversal of the influx transporter. It appeared that the experimental data were unable to resolve the small, rapid initial decline in fluorescence from the slow and steady Zn²⁺ efflux. Regardless, both the model and experimental data agree that the intracellular free Zn²⁺ concentration must remain low, even in the face of much larger net Zn²⁺ uptakes and that Zn²⁺ efflux remain small, most likely as a result of the low intracellular free Zn²⁺ levels (i.e., large intrinsic buffering capacity). Thus the plateau in ZnAF-2F fluorescence observed after addition of pyrithione reflects mostly ZnAF-2F saturation.

Addition of CQ and Zn²⁺ resulted in a large increase in Zn²⁺ influx and net uptake of Zn²⁺. A previous study has suggested that CQ might act as a Zn²⁺ ionophore, but did not directly measure an increase in cellular Zn²⁺ content after CQ addition (13). We made direct measurements of Zn²⁺ uptake in the presence of low micromolar concentrations of CQ to confirm this finding, and we felt we could test the validity of the model by including CQ actions. Figure 7 shows images illustrating the increase in fluorescence observed after neurons were incubated with 30 µM Zn²⁺ or 30 µM Zn²⁺ and 1 µM CQ for 5 min (compare Fig. 7, A and B). Next, neurons treated with Zn²⁺ and CQ were subsequently treated with 200 µM TPEN, showing that the increase in intracellular fluorescence was completely reversed (Fig. 7C). Results of quantifying the changes in cellular fluorescence intensity observed in Fig. 7, A–C, are shown in Fig. 7D. These findings were confirmed by determination of total cellular Zn²⁺ uptake using ICP-MS analysis and ⁶⁵Zn²⁺ uptake as shown in Fig. 8. No effect of 100 nM CQ upon Zn²⁺ uptake after incubation with 10 µM extracellular Zn²⁺ was observed (Fig. 8, A and B). When Zn²⁺ and 3 µM CQ were added, both ⁶⁵Zn²⁺ uptake and total cellular Zn²⁺ determined by ICP-MS analysis were increased to levels similar to that seen with the addition of 5 µM pyrithione and Zn²⁺ (compare Figs. 2 and 8). The effects of CQ on the cellular Zn²⁺ content were specific, as no change in total Mn, Fe, or Cu levels were observed by ICP-MS analysis. As shown previously in Fig. 2C, exposure of cortical neurons to increased extracellular Zn²⁺ resulted in increased Zn²⁺ surface binding. It was determined also whether addition of CQ resulted in an additional increase in surface binding of Zn²⁺. As shown in Fig. 8C, a small increase in Zn²⁺ surface binding was observed in the presence of CQ, as might be expected for Zn²⁺/CQ complexes associated with the plasma membrane. However, data presented in Figs. 7 and 8 clearly show that exposure to CQ in the presence of extracellular Zn²⁺ results in an increased intracellular free Zn²⁺ concentration and total Zn²⁺ content.

View larger version (60K): [in this window] [in a new window]   Fig. 7. Cortical neurons attached to glass coverslips were treated as described in Fig. 1 and exposed to 30 µM Zn2+ (A), or 30 µM Zn2+ plus 1 µM CQ (B), or same as in B followed by 200 µM TPEN (C). Cell bodies were analyzed for average pixel intensity, and those data are shown in D. Each bar represents mean ± SE (n = 7 individual cells on one coverslip as pictured in A–C). A Tukey multiple-comparison test (GraphPad Prism, version 4.03) showed that, for any column pair, P < 0.001. Size bar = 10 µm.

View larger version (17K): [in this window] [in a new window]   Fig. 8. Cortical neurons were exposed to 10 µM Zn2+ with 0, 100 nM, or 3 µM CQ for 30 min, as described in Fig. 2. Results are of ICP-MS analysis (A) or treated as in A with the addition of 65Zn2+ and lysate radioactivity determined by liquid scintillation counting (B). C: cortical neurons treated as in B, with or without addition of 100 µM EDTA for 5 min after the 30-min incubation with 65Zn2+. Lysate protein was determined by Bio-Rad protein assay using BSA as a standard. Each bar represents mean ± SE; nos. in parentheses are the no. of replicates. A Tukey multiple-comparison test (GraphPad Prism, version 4.03) was performed on each data set (A, B, or C) and showed that only the differences between Zn2+ and Zn2+ with 100 nM CQ (A or B) and Zn2+ and Zn2+ with 5-min EDTA treatment (C) were not significant; otherwise, all other comparisons were significant (P 0.01).

When increasing concentrations of CQ (0.1–3 µM) were added in the presence of 30 µM extracellular Zn²⁺, the F/F₀ (i.e., Zn²⁺ influx) increased in a concentration-dependent manner (Fig. 9). Addition of pyrithione 3 min after the start of each reaction resulted in a rapid increase in ZnAF-2F fluorescence that quickly reached saturation (see also Fig. 3). At 3 µM CQ, ZnAF-2F was already saturated before pyrithione addition (Fig. 9D). Addition of TPEN after pyrithione addition resulted in a complete reversal of the fluorescence signal (Fig. 9). In Fig. 9, results are shown also from the model predictions with CQ added. CQ was modeled as an ionophore, and the responses to 30 µM extracellular Zn²⁺ with the addition of various concentrations of CQ generally followed the experimental data. The primary difference was that the model did not match exactly the dose-dependent increases in Zn²⁺ influx seen experimentally as a result of CQ addition.

View larger version (35K): [in this window] [in a new window]   Fig. 9. Experimental data (solid line) and model 2 predictions (dotted line) plotted together on the same graphs. Cortical neurons were loaded with 5 µM ZnAF-2F and treated with 30 µM Zn2+ for 3 min (A) with the addition of 100 nM (A), 300 nM (B), 1 µM (C), or 3 µM (D) CQ, followed by 5 µM pyrithione for 3 min, and finally 200 µM TPEN. Each plot is the mean of data obtained from 3 individual coverslips.

View larger version (17K): [in this window] [in a new window]   Fig. 10. A: cortical neurons were loaded with 5 µM ZnAF-2F and treated with 30 µM Zn2+ for 3 min, then treated with 1 µM CQ, with (shaded line) or without (solid line) 100 µM EDTA, and finally treated with 200 µM TPEN. Each plot is the mean of data obtained from 3 individual coverslips. B: model 2 predictions for the experimental conditions as described in A (see legend). Model predictions without CQ but with EDTA are plotted also to allow comparison of efflux rates.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

While the free parameters of the muffler model (model 2) could not be determined uniquely, sets of parameter values that satisfied the experimental constraints shared important features. The muffler had to have high affinity with K_d of 60–100 pM, and the free muffler concentration had to be near 20 µM, characteristics similar to those noted by Krezel and Maret (38). These characteristics, plus the ability to deposit Zn²⁺ into the deep store, allow the muffler to exert a fine control over intracellular free Zn²⁺ concentration over a wide range of Zn²⁺ loads. The model also predicts that 12–30 µM Zn²⁺ resides in the deep store at rest. When neurons are exposed to Zn²⁺ and pyrithione, the total neuronal Zn²⁺ content can be increased nearly 50-fold over extracellular concentrations, with Zn²⁺ being shuttled to the deep store by the muffler. Recent experimental evidence suggests that the large-capacity deep store could be represented, in part or combination, by mitochondria (53), the endoplasmic reticulum/Golgi (17), and other sites of Zn²⁺ compartmentalization within the soma of cultured neurons (10).

The model provides an estimate of the resting intracellular free Zn²⁺ concentration in the soma of cultured cortical neurons. Given the constraint that, on average, 16% of ZnAF-2F is bound at rest, the resting intracellular free Zn²⁺ concentration in the presence of ZnAF-2F is 1.05 nM. When ZnAF-2F is removed from the model, this value increases slightly to 1.07 nM. It is interesting to note that this concentration of free Zn²⁺ represents 500 free Zn²⁺ ions in the soma of a typically sized cortical neuron. This estimate is somewhat higher than the estimate obtained for HT-29 cells (38), and 100-fold higher than that obtained with PC-12 cells. Each study used a different method of estimating intracellular free Zn²⁺, which could account for some of the differences; however, it is quite likely that different cell types and culture conditions result in different buffer capacity and set points for intracellular free Zn²⁺ levels.

The variability in resting ZnAF-2F saturation levels (i.e., differences in resting intracellular free Zn²⁺) found in the experiments was attributed by the model to different levels of total Zn²⁺ in the cultured neurons. We did not systematically measure total Zn²⁺ and resting intracellular free Zn²⁺ in a series of different neuron cultures. However, it seems reasonable to expect that different cultures might vary in total Zn²⁺ content. What factors might cause such a variation? We would expect that the free and total Zn²⁺ concentrations of NB media remain within tight limits. Thus it is most likely that one or more intrinsic factors, beyond the control of the experimenter, determine the average Zn²⁺ content of particular cultures. For example, small changes in the culture density might lead to significant, albeit subtle, differences in expression of various genes influencing Zn²⁺ homeostasis. Even though the total Zn²⁺ content of resting cultured neurons seemed to vary, the neurons maintain resting intracellular free Zn²⁺ concentration within narrow limits. The model predicts that neurons accomplish this feat by increasing or decreasing the cellular muffler concentration to match the cellular Zn²⁺ load.

Neuronal free Zn²⁺ concentration is quite low and strongly buffered. The Zn²⁺ buffer capacity of neurons can be quantified using the method developed for estimating proton buffering capacity of cells (51). These authors defined buffer capacity as all of the cellular processes (including sequestration in cytoplasmic organelles, but excluding plasma membrane transport) that act to mitigate changes in cytosolic pH when cells are challenged with an acid or base load. A similar definition can be used to quantify Zn²⁺ buffer capacity and will allow the comparison of intrinsic buffer capacity in different cell types and provide a quantitative measure of changes in buffer capacity after physiological or pathological perturbations. Thus

(3)

where β_Zn is the intrinsic buffer capacity for Zn²⁺, and d[Zn] and dpZn represent corresponding changes in the total cellular Zn²⁺ content (µM) and the –log of the intracellular free Zn²⁺ concentration, respectively. Using ⁶⁵Zn²⁺ uptake and model estimates of changes in intracellular free Zn²⁺ when cultured cortical neurons were exposed to 30 µM Zn²⁺, we found that neuronal Zn²⁺ content changed 88 µM/min, whereas dpZn only changed 0.22 unit/min. Therefore, β_Zn for cultured cortical neurons is 396 µM/pZn unit. As eukaryotic cells evolved, Zn²⁺ and Ca²⁺ became important intracellular metals utilized in a host of biologically significant interactions because of availability and lack of redox activity. However, for both metals, even small increases in intracellular concentrations can be cytotoxic. Thus multiple protective mechanisms evolved to maintain free metal concentrations within narrow limits, one of these being cytosolic buffering through the use of soluble cytosolic metal binding proteins. In the case of Ca²⁺ buffering, a large family of EF-hand containing proteins shares responsibility for the control of resting cytosolic Ca²⁺ concentration and functions as Ca²⁺ sensors (4). The same may be true of cytosolic Zn²⁺ buffering. The large Zn²⁺ buffering capacity of neurons can be a "double-edged" sword. On the one hand, Zn²⁺ buffering capacity protects neurons from large swings in intracellular free Zn²⁺ concentrations, but on the other hand poses significant risk to the neuron when exposed to large extracellular Zn²⁺ concentrations for prolonged periods. Under these conditions, large Zn²⁺ loads can be accumulated, and, since intracellular free Zn²⁺ remains low, efflux mechanisms at the plasma membrane are slow to eliminate such loads.

Role of MT in cytosolic Zn²⁺ buffering. We first explored whether MT could play a role in Zn²⁺ homeostasis as a simple Zn²⁺ buffer. Even when added at a concentration of 1 µM to model 2, which is greater than the 0.34 µM measured experimentally, MT could not buffer the Zn²⁺ loads described here. At rest, MT Zn²⁺ binding sites 1-6 were nearly saturated, and site 7 was largely free, and there was little change during exposure to 30 µM extracellular Zn²⁺. For MT to be an effective cytosolic buffer in cultured neurons under the conditions of model 2, the concentration chosen for MT would have to be very much larger. However, MT certainly could play a more prominent role in simple cytosolic Zn²⁺ buffering in cell types other than cultured neurons, as has been shown in mouse lung fibroblasts (54).

In the MT as muffler model (model 3), we assumed that MT was exclusively responsible for sequestration of Zn²⁺ to the deep store. However, for MT to be an effective muffler of the Zn²⁺ loads applied in these studies, its concentration could not be small. Concentrations of 0.15, 0.5, or 1.0 µM could not shuttle enough Zn²⁺ to the deep store fast enough to match experimental constraints, but 2.0 µM could. If MT concentration is indeed low, then a more parsimonious explanation for MT's role in resting neurons is that MT is part of the high-affinity muffler pool represented in the muffler model (model 2). Interestingly, the low-affinity MT Zn²⁺ binding site, site 7 (K_d of 20 nM), was able to shuttle Zn²⁺ to the deep store, at least as effectively as site 6 (K_d of 100 pM) and more effectively than site 5 (K_d of 40 pM) in these simulations, despite having much less Zn²⁺ bound at rest. Thus it may be more important to consider the relative Zn²⁺ binding site occupancy, particularly at sites 6 and 7, than the ratio of MT/T when considering the potential actions of MT as a muffler. On the other hand, the model predictions would be consistent with MT playing an important role in a neuron's long-term response to increased Zn²⁺ loads. In this situation, increases in MT protein expression (resulting in increased cellular concentrations) could allow MT to become the dominant cytosolic Zn²⁺ muffler. There are numerous examples and many treatments that have been shown to increase cellular MT levels severalfold (36). In addition, direct evidence of increased cytosolic Zn²⁺ buffering capacity with increased MT expression has been reported (38, 42).

MT may actually play a far more important role as a Zn²⁺ chaperone involved in pathways of Zn²⁺ trafficking, where MT transfers Zn²⁺ to highly specific cellular moieties (e.g., glutathione) under the control of cellular redox state (31, 32) or to mitochondrial transporters (11). For example, glutathione could facilitate the net movement of Zn²⁺ from MT to a deep store organelle or apoenzyme. We have included a transfer agent in models, but we have not reported the results here, because results were similar to those of the MT as muffler model (model 3) for parameter values chosen; also, there was no satisfying way to constrain the additional free parameters in the model for the transfer agent, especially since there are likely to be several different types of transfer agents with different concentrations and interaction rates with MT Zn²⁺ binding sites. Very little useful experimental data exist right now to inform such a model.

Routes for Zn²⁺ influx and efflux. The model uses a straightforward Michaelis-Menten description of Zn²⁺ influx. It is recognized that numerous studies have indicated that multiple pathways may exist for both the influx and efflux of Zn²⁺ in neurons (8). However, modeling influx in this way with efflux as a simple reversal of the influx process was adequate to match the experimental data. Concerning Zn²⁺ efflux, both experimental data and the model showed that efflux was small, even after neurons were maximally loaded with Zn²⁺ by exposure to ionophores, presumably a result of the large Zn²⁺ buffering capacity of neurons. In other words, even after a large Zn²⁺ load, intracellular free Zn²⁺ levels remain low and apparently are not enough to drive substantial amounts of Zn²⁺ efflux. No evidence for the active extrusion of Zn²⁺ (like that seen for Ca²⁺) has been described to date. These data might appear to be in conflict with a previous report (20), but these experiments incubated neurons with Ca-EDTA for 1 h. Our experiments show a slow and steady Zn²⁺ efflux, which, after 1 h, would be expected to result in significant net Zn²⁺ release. Experimental evidence suggests that, when neurons are exposed to ⁶⁵Zn²⁺, a large portion (nearly 40% of the original load) of the accumulated tracer is surface bound. This may seem like a large amount, but there are many sites, including the glass surface on which the neurons are grown, that could provide ample Zn²⁺ binding sites. The model has yet to be expanded to include a description of channel-mediated Zn²⁺ influx and efflux (52), but this is a logical next step in the model development.

Insights into the cellular mechanisms of action of CQ. The model results indicate that the actions of CQ seen in the experiments are consistent with CQ acting primarily as a Zn²⁺ ionophore when added to cultured cortical neurons. At the concentrations of CQ added, CQ certainly did not significantly affect Zn²⁺ buffering capacity, as has been suggested by some to explain its mechanism of action in Alzheimer's disease. In the model, CQ binds to extracellular Zn²⁺, crosses the plasma membrane, releases Zn²⁺ inside the neuron, and then returns to the extracellular space to rebind to Zn²⁺. Both Zn²⁺ bound and free CQ cross the neuron membrane, according to their concentration gradients. Modeling the actions of CQ in this way was sufficient to reproduce the experimental results.

The addition of micromolar concentrations of CQ to cultured cells has been shown to have several important biological actions related to its presumed metal chelating activity when administered in vivo. For example, incubation with CQ and Zn²⁺ or Cu²⁺ for extended periods (6 h) resulted in large increases in cellular levels of Zn²⁺ or Cu²⁺ (61) and was responsible for an upregulation of metalloproteinase activity that caused a reduction in the production and secretion of Aβ 1-40 and 1-42. Other studies have shown that micromolar CQ can increase functional levels of hypoxia-inducible factor-1, (protective in ischemic diseases) (5), and micromolar CQ downregulates mutant Huntington protein expression (46). Each of these effects could be due, at least in part, to the ionophoric actions of CQ shown here. Earlier studies have shown that much higher concentrations of CQ (1 mM) mediate cellular Cu²⁺ uptake (60). In the present report, CQ levels as low as 300 nM had discernible effects on the intracellular free Zn²⁺ concentration. Thus micromolar concentrations of CQ should be effective at increasing the total and intracellular free Zn²⁺ concentration in neurons, but require micromolar Zn²⁺ concentrations to be present extracellularly. It should be noted that resting extracellular free Zn²⁺ in the brain is estimated to be <25 nM (19). This concentration is much too low to activate the ionophoric actions of CQ. Thus CQ would not be expected to be a generalized Zn²⁺ ionophore in the central nervous system. However, since extracellular Zn²⁺ levels during normal activity at Zn²⁺ containing glutamatergic synapses should be much greater (19), this mechanism provides for a selective synaptic effect in those neurons. The effective treatment dosage in humans is 250–750 mg daily, giving blood levels of 10 µM (50). CQ shows a rapid brain uptake upon administration (47), but it is not known whether brain concentrations reach micromolar levels. Further studies will be required to determine the role that the ionophoric actions of CQ play in its various biological and therapeutic actions.

	GRANTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
This work was supported by the Interdisciplinary Graduate Program in Molecular and Cellular Biology, Ohio University, and R. A. Colvin was supported by National Institute on Aging Grant AG20536.

	FOOTNOTES

 

Address for reprint requests and other correspondence: R. A. Colvin, Dept. of Biological Sciences, Ohio Univ., Athens, OH, 45701 (e-mail: colvin{at}ohio.edu)

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

	REFERENCES

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION GRANTS REFERENCES

 
1. Adamczyk M, Poznanski J, Kopera E, Bal W. A zinc-finger like metal binding site in the nucleosome. FEBS Lett 581: 1409–1416, 2007.[CrossRef][Web of Science][Medline]

2. Bozym RA, Thompson RB, Stoddard AK, Fierke CA. Measuring picomolar intracellular exchangeable zinc in PC-12 cells using a ratiometric fluorescence biosensor. ACS Chem Biol 1: 103–111, 2006.[CrossRef][Medline]

3. Canzoniero LMT, Turetsky DM, Choi DW. Measurement of intracellular free zinc concentrations accompanying zinc-induced neuronal death. J Neurosci 19: 1–6, 1999.[Abstract/Free Full Text]

4. Case RM, Eisner D, Gurney A, Jones O, Muallem S, Verkhratsky A. Evolution of calcium homeostasis: from birth of the first cell to an omnipresent signaling system: calcium channels and transporters. Cell Calcium 42: 345–350, 2007.[CrossRef][Web of Science][Medline]

5. Choi SM, Choi KO, Park YK, Cho H, Yang EG, Park H. Clioquinol, a Cu(II)/Zn(II) chelator, inhibits both ubiquitination and asparagine hydroxylation of hypoxia-inducible factor-1 alpha, leading to expression of vascular endothelial growth factor and erythropoietin in normoxic cells. J Biol Chem 281: 34056–34063, 2006.[Abstract/Free Full Text]

6. Colvin RA. pH dependence and compartmentalization of zinc transported across plasma membrane of rat cortical neurons. Am J Physiol Cell Physiol 282: C317–C329, 2002.[Abstract/Free Full Text]

7. Colvin RA, Davis N, Nipper RW, Carter PA. Zinc transport in the brain: routes of zinc influx and efflux in neurons. J Nutr 130: 1484S–1487S, 2000.[Abstract/Free Full Text]

8. Colvin RA, Fontaine CP, Laskowski M, Thomas D. Zn²⁺ transporters and Zn²⁺ homeostasis in neurons. Eur J Pharmacol 479: 171–185, 2003.[CrossRef][Web of Science][Medline]

9. Colvin RA, Fontaine CP, Thomas D, Hirano T, Nagano T, Kikuchi K. Evidence for pH dependent Zn²⁺ influx in K562 erythroleukemia cells: studies using ZnAF-2F fluorescence and ⁶⁵Zn²⁺ uptake. Arch Biochem Biophys 442: 222–228, 2005.[CrossRef][Web of Science][Medline]

10. Colvin RA, Laskowski M, Fontaine CP. Zinquin identifies subcellular compartmentalization of zinc in cortical neurons. Relation to the trafficking of zinc and the mitochondrial compartment. Brain Res 1085: 1–10, 2006.[CrossRef][Web of Science][Medline]

11. Costello LC, Guan Z, Franklin RB, Feng P. Metallothionein can function as a chaperone for zinc uptake transport into prostate and liver mitochondria. J Inorg Biochem 98: 664–666, 2004.[CrossRef][Web of Science][Medline]

12. Dineley KE, Malaiyandi LM, Reynolds IJ. A reevaluation of neuronal zinc measurements: artifacts associated with high intracellular dye concentration. Mol Pharmacol 62: 618–627, 2002.[Abstract/Free Full Text]

13. Ding WQ, Liu B, Vaught JL, Yamauchi H, Lind SE. Anticancer activity of the antibiotic clioquinol. Cancer Res 65: 3389–3395, 2005.[Abstract/Free Full Text]

14. Eaton DL, Cherian MG. Determination of metallothionein in tissues by cadmium-hemoglobin affinity assay. Methods Enzymol 205: 83–88, 1991.[CrossRef][Web of Science][Medline]

15. Eide DJ. Zinc transporters and the cellular trafficking of zinc. Biochim Biophys Acta 1763: 711–722, 2006.[Medline]

16. Ellis CD, Wang F, MacDiarmid CW, Clark S, Lyons T, Eide DJ. Zinc and the Msc2 zinc transporter protein are required for endoplasmic reticulum function. J Cell Biol 166: 325–335, 2004.[Abstract/Free Full Text]

17. Ellis CD, MacDiarmid CW, Eide DJ. Heteromeric protein complexes mediate zinc transport into the secretory pathway of eukaryotic cells. J Biol Chem 280: 28811–28818, 2005.[Abstract/Free Full Text]

18. Ferrada E, Arancibia V, Loeb B, Norambuena E, Olea-Azar C, Huidobro-Toro JP. Stoichiometry and conditional stability constants of Cu(II) or Zn(II) clioquinol complexes; implications for Alzheimer's and Huntington's disease therapy. Neurotoxicology 28: 445–449, 2007.[CrossRef][Web of Science][Medline]

19. Frederickson CJ, Giblin LJ, Krezel A, McAdoo DJ, Muelle RN, Zeng Y, Balaji RV, Masalha R, Thompson RB, Fierke CA, Sarvey JM, de Valdenebro M, Prough DS, Zornow MH. Concentrations of extracellular free zinc (pZn)e in the central nervous system during simple anesthetization, ischemia and reperfusion. Exp Neurol 198: 285–293, 2006.[CrossRef][Web of Science][Medline]

20. Frederickson CJ, Suh SW, Koh JY, Cha YK, Thompson RB, LaBuda CJ, Balaji RV, Cuajungco MP. Depletion of intracellular zinc from neurons by use of an extracellular chelator in vivo and in vitro. J Histochem Cytochem 50: 1659–1662, 2002.[Abstract/Free Full Text]

21. Frederickson CJ, Koh JY, Bush AI. The neurobiology of zinc in health and disease. Nat Rev Neurosci 6: 449–462, 2005.[CrossRef][Web of Science][Medline]

22. Gaither LA, Eide DJ. Functional expression of the human hZIP2 zinc transporter. J Biol Chem 275: 5560–5564, 2000.[Abstract/Free Full Text]

23. Gaither LA, Eide DJ. The human ZIP1 transporter mediates zinc uptake in human K562 erythroleukemia cells. J Biol Chem 276: 22258–22264, 2001.[Abstract/Free Full Text]

24. Gaither LA, Eide DJ. Eukaryotic zinc transporters and their regulation. Biometals 14: 251–270, 2001.[CrossRef][Web of Science][Medline]

25. Grynkiewicz G, Poenie M, Tsien RY. A new generation of Ca²⁺ indicators with greatly improved fluorescence properties. J Biol Chem 260: 3440–3450, 1985.[Abstract/Free Full Text]

26. Gunzel D, Zimmermann F, Durry S, Schlue WR. Apparent intracellular Mg²⁺ buffering in neurons of the leech Hirudo medicinalis. Biophys J 80: 1298–1310, 2001.[Web of Science][Medline]

27. Hidalgo J, Garcia A, Oliva AM, Giralt M, Gasull T, Gonzalez B, Milnerowicz H, Wood A, Bremner I. Effect of zinc, copper and glucocorticoids on metallothionein levels of cultured neurons and astrocytes from rat brain. Chem Biol Interact 93: 197–219, 1994.[CrossRef][Web of Science][Medline]

28. Hirano T, Kikuchi K, Urano Y, Nagano T. Improvement and biological applications of fluorescent probes for zinc, ZnAFs. J Am Chem Soc 124: 6555–6562, 2002.[CrossRef][Web of Science][Medline]

29. Hoops S, Sahle S, Gauges R, Lee C, Pahle J, Simus N, Singhal M, Xu L, Mendes P, Kummer U. COPASI–a COmplex PAthway SImulator. Bioinformatics 22: 3067–3074, 2006.[Abstract/Free Full Text]

30. Huang L, Kirschke CP, Zhang Y, Yu YY. The ZIP7 gene (Slc39a7) encodes a zinc transporter involved in zinc homeostasis of the golgi apparatus. J Biol Chem 280: 15456–15463, 2005.[Abstract/Free Full Text]

31. Jacob C, Maret W, Vallee BL. Control of zinc transfer between thionein, metallothionein, and zinc proteins. Proc Natl Acad Sci USA 95: 3489–3494, 1998.[Abstract/Free Full Text]

32. Jiang LJ, Maret W, Vallee BL. The glutathione redox couple modulates zinc transfer from metallothionein to zinc-depleted sorbitol dehydrogenase. Proc Natl Acad Sci USA 95: 3483–3488, 1998.[Abstract/Free Full Text]

33. Jiang LJ, Vasak M, Vallee BL, Maret W. Zinc transfer potentials of the alpha- and beta-clusters of metallothionein are affected by domain interactions in the whole molecule. Proc Natl Acad Sci USA 97: 2503–2508, 2000.[Abstract/Free Full Text]

34. Kim AH, Sheline CT, Tian M, Higashi T, McMahon RJ, Cousins RJ, Choi DW. L-type Ca²⁺ channel-mediated Zn²⁺ toxicity and modulation by ZnT-1 in PC12 cells. Brain Res 886: 99–107, 2000.[CrossRef][Web of Science][Medline]

35. Kirschke CP, Huang L. ZnT7, a novel mammalian transporter, accumulates zinc in the golgi apparatus. J Biol Chem 278: 4096–4102, 2003.[Abstract/Free Full Text]

36. Klaassen CD, Liu J. Induction of metallothionein in primary rat hepatocyte cultures. Methods Enzymol 205: 567–574, 1991.[Web of Science][Medline]

37. Krezel A, Hao Q, Maret W. The zinc/thiolate redox biochemistry of metallothionein and the control of zinc ion fluctuations in cell signaling. Arch Biochem Biophys 463: 188–200, 2007.[CrossRef][Web of Science][Medline]

38. Krezel A, Maret W. Zinc-buffering capacity of a eukaryotic cell at physiological pZn. J Biol Inorg Chem 11: 1049–1062, 2006.[CrossRef][Web of Science][Medline]

39. Krezel A, Maret W. Dual nanomolar and picomolar Zn(II) binding properties of metallothionein. J Am Chem Soc 129: 10911–10921, 2007.[CrossRef][Web of Science][Medline]

40. Krezel A, Wojcik J, Maciejczyk M, Bal W. May GSH and L-His contribute to intracellular binding of zinc? Thermodynamic and solution structural study of a ternary complex. Chem Commun (Camb) 6: 704–705, 2003.[Medline]

41. Lin W, Mohandas B, Fontaine CP, Colvin RA. Release of intracellular Zn(2+) in cultured neurons after brief exposure to low concentrations of exogenous nitric oxide. Biometals 20: 891–901, 2007.[CrossRef][Web of Science][Medline]

42. Malaiyandi LM, Dineley KE, Reynolds IJ. Divergent consequences arise from metallothionein overexpression in astrocytes: zinc buffering and oxidant-induced zinc release. Glia 45: 346–353, 2004.[CrossRef][Web of Science][Medline]

43. Malaiyandi LM, Vergun O, Dineley KE, Reynolds IJ. Direct visualization of mitochondrial zinc accumulation reveals uniporter-dependent and -independent transport mechanisms. J Neurochem 93: 1242–1250, 2005.[CrossRef][Web of Science][Medline]

44. Mendes P. Biochemistry by numbers: simulation of biochemical pathways with Gepasi 3. Trends Biochem Sci 22: 361–363, 1997.[CrossRef][Web of Science][Medline]

45. Neher E. The use of fura-2 for estimating Ca buffers and Ca fluxes. Neuropharmacology 34: 1423–1442, 1995.[CrossRef][Web of Science][Medline]

46. Nguyen T, Hamby A, Massa SM. Clioquinol down-regulates mutant Huntington expression in vitro and mitigates pathology in a Huntington's disease mouse model. Proc Natl Acad Sci USA 102: 11840–11845, 2005.[Abstract/Free Full Text]

47. Opazo C, Luza S, Villemagne VL, Volitakis I, Rowe C, Barnham KJ, Strozyk D, Masters CL, Cherny RA, Bush AI. Radioiodinated clioquinol as a biomarker for beta-amyloid: Zn²⁺ complexes in Alzheimer's disease. Aging Cell 5: 69–79, 2006.[CrossRef][Web of Science][Medline]

48. Palmiter RD, Cole TB, Quaife CJ, Findley SD. ZnT-3, a putative transporter of zinc into synaptic vesicles. Proc Natl Acad Sci USA 93: 14934–14939, 1996.[Abstract/Free Full Text]

49. Palmiter RD, Findley SD. Cloning and functional characterization of a mammalian zinc transporter that confers resistance to zinc. EMBO J 14: 639–649, 1995.[Web of Science][Medline]

50. Ritchie CW, Bush AI, Mackinnon A, Macfarlane S, Mastwyk M, MacGregor L, Kiers L, Cherny R, Li QX, Tammer A, Carrington D, Mavros C, Volitakis I, Xilinas M, Ames D, Davis S, Beyreuther K, Tanzi RE, Masters CL. Metal-protein attenuation with iodochlorhydroxyquin (clioquinol) targeting Abeta amyloid deposition and toxicity in Alzheimer disease: a pilot phase 2 clinical trial. Arch Neurol 60: 1685–1691, 2003.[Abstract/Free Full Text]

51. Roos A, Boron WF. Intracellular pH. Physiol Rev 61: 296–434, 1981.[Free Full Text]

52. Sensi SL, Canzoniero LMT, Yu SP, Ying HS, Koh JY, Kerchner GA, Choi DW. Measurement of intracellular free zinc in living cortical neurons: routes of entry. J Neurosci 17: 9554–9564, 1997.[Abstract/Free Full Text]

53. Sensi SL, Ton-That D, Sullivan PG, Jonas EA, Gee KR, Kaczmarek LK, Weiss JH. Modulation of mitochondrial function by endogenous Zn²⁺ pools. Proc Natl Acad Sci USA 100: 6157–6162, 2003.[Abstract/Free Full Text]

54. St. Croix CM, Wasserloos KJ, Dineley KE, Reynolds IJ, Levitan ES, Pitt BR. Nitric oxide-induced changes in intracellular zinc homeostasis are mediated by metallothionein/thionein. Am J Physiol Lung Cell Mol Physiol 282: L185–L192, 2002.[Abstract/Free Full Text]

55. Suzuki T, Ishihara K, Migaki H, Ishihara K, Nagao M, Yamaguchi-Iwai Y, Kambe T. Two different zinc transport complexes of cation diffusion facilitator proteins localized in the secretory pathway operate to activate alkaline phosphatases in vertebrate cells. J Biol Chem 280: 30956–30962, 2005.[Abstract/Free Full Text]

56. Suzuki T, Ishihara K, Migaki H, Matsuura W, Kohda A, Okumura K, Nagao M, Yamaguchi-Iwai Y, Kambe T. Zinc transporters, ZnT5 and ZnT7, are required for the activation of alkaline phosphatases, zinc-requiring enzymes that are glycosylphosphatidylinositol-anchored to the cytoplasmic membrane. J Biol Chem 280: 637–643, 2005.[Abstract/Free Full Text]

57. Takeda A. Movement of zinc and its functional significance in the brain. Brain Res Rev 34: 137–148, 2000.[CrossRef][Medline]

58. Thomas RC, Coles JA, Deitmer JW. Homeostatic muffling. Nature 350: 564, 1991.[Medline]

59. Thompson RB, Peterson D, Mahoney W, Cramer M, Maliwal BP, Suh SW, Frederickson C, Fierke C, Herman P. Fluorescent zinc indicators for neurobiology. J Neurosci Methods 118: 63–75, 2002.[CrossRef][Web of Science][Medline]

60. Treiber C, Simons A, Strauss M, Hafner M, Cappai R, Bayer TA, Multhaup G. Clioquinol mediates copper uptake and counteracts copper efflux activities of the amyloid precursor protein of Alzheimer's disease. J Biol Chem 279: 51958–51964, 2004.[Abstract/Free Full Text]

61. White AR, Du T, Laughton KM, Volitakis I, Sharples RA, Xilinas ME, Hoke DE, Holsinger RM, Evin G, Cherny RA, Hill AF, Barnham KJ, Li QX, Bush AI, Masters CL. Degradation of the Alzheimer disease amyloid beta-peptide by metal-dependent up-regulation of metalloprotease activity. J Biol Chem 281: 17670–17680, 2006.[Abstract/Free Full Text]

This article has been cited by other articles:

				D. A. Andersson, C. Gentry, S. Moss, and S. Bevan Clioquinol and pyrithione activate TRPA1 by increasing intracellular Zn2+ PNAS, May 19, 2009; 106(20): 8374 - 8379. [Abstract] [Full Text] [PDF]

				M. Hollenstein, C. J. Hipolito, C. H. Lam, and D. M. Perrin A self-cleaving DNA enzyme modified with amines, guanidines and imidazoles operates independently of divalent metal cations (M2+) Nucleic Acids Res., April 1, 2009; 37(5): 1638 - 1649. [Abstract] [Full Text] [PDF]

				P. J. Smith, M. Wiltshire, E. Furon, J. H. Beattie, and R. J. Errington Impact of overexpression of metallothionein-1 on cell cycle progression and zinc toxicity Am J Physiol Cell Physiol, November 1, 2008; 295(5): C1399 - C1408. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 294/3/C726    most recent 00541.2007v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Colvin, R. A. Articles by Holmes, W. R. Search for Related Content PubMed PubMed Citation Articles by Colvin, R. A. Articles by Holmes, W. R.