Hypoxic Enhancement of Quantal Catecholamine Secretion

Prolonged exposure to hypoxia (10% O2) enhanced quantal catecholamine release evoked from O2-sensing pheochromocytoma (PC12) cells, as monitored using single-cell amperometric recordings. The enhancement of exocytosis was apparent after 12 h of hypoxia and was maximal at 24 h. Elevated levels of secretion were due to the emergence of a Ca2+ influx pathway that persisted during complete blockade of known voltage-gated Ca2+ channels. Secretion triggered by this Ca2+ influx was severely reduced by known inhibitors of Alzheimer's amyloid β-peptides (AβPs), including an N terminus-directed monoclonal antibody. The enhancing effect on secretion of chronic hypoxia was mimicked closely by direct application of AβP to cells under normoxic conditions, although the effects of AβP were more rapid at onset, being maximal after only 6 h. The present results suggest that prolonged hypoxia can induce formation of Ca2+-permeable AβP channels and that such induction can lead directly to excessive neurosecretion. This is a potential contributory factor to AβP pathophysiology following cerebral ischemia.

Dementias in general and Alzheimer's disease in particular are more prevalent following periods of cerebral ischemia caused by cardiovascular dysfunctions such as stroke (1)(2)(3). Ischemia is a complex condition, causing perturbations in several parameters that are essential to neuronal survival such as lack of substrates, accumulation of metabolic products, acidosis, and reduction of oxygen levels. However, the specific contributions of each of these factors to cellular dysfunction and death are presently unknown.
A fundamental feature of Alzheimer's disease is the accumulation of fibrillar deposits consisting of amyloid ␤-peptides (A␤Ps) 1 (reviewed in Ref. 4). A␤Ps are 40 -42-amino acid peptides and are cleavage products derived from amyloid precursor protein (APP) (5,6). APP is one of only a few gene products whose expression is increased following a period of cerebral ischemia (7,8). Since APP is generally perceived to be neuroprotective, this increased expression of APP can be considered a defense mechanism against ischemia. However, increased APP levels seen in ischemia would also permit increased formation of A␤Ps, which cause neuronal damage and death (4); and indeed, A␤P production is increased following both mild and severe ischemia (9,10). Thus, a clear link exists between ischemic insult and elevation of damaging A␤P levels. How-ever, it remains completely unknown how ischemia might increase A␤P levels, and the mechanism(s) by which A␤Ps cause cellular dysfunction and death remain speculative.
Reduction of available O 2 (hypoxia) is a key feature of ischemia. Hypoxia is known to exert a diverse range of responses in cells, each of which serves a specific physiological purpose (11). Acute hypoxia can evoke extremely rapid responses such as selective, membrane-delimited inhibition of ion channels (12,13). Prolonged (chronic) hypoxia can alter gene expression and so, for example, permit increased production of erythropoietin (14) or, in electrically excitable cells, tyrosine hydroxylase, the rate-limiting enzyme in catecholamine synthesis (15,16). In this study, we have utilized the catecholamine-secreting pheochromocytoma (PC12) cell to investigate the effects of both acute and chronic hypoxia on exocytosis, as determined in real time using amperometry (17). PC12 cells represent a well defined, model excitable cell system, which has been used extensively to study the effects of both acute and chronic hypoxia on various cellular processes, including ion channel activity and gene expression (16, 18 -20). PC12 cells have also been extensively characterized as a model system for studying exocytosis (21).
Our recent work has demonstrated that acute hypoxia evokes catecholamine release from PC12 cells by causing membrane depolarization and subsequent Ca 2ϩ influx, primarily through N-type voltage-gated Ca 2ϩ channels (20). Furthermore, chronic mild hypoxia (10% O 2 for 24 h) enhances this secretory response of PC12 cells to acute hypoxia (22). This enhancement is in part due to increased expression of O 2sensitive K ϩ channels that influence membrane potential, but is also due to the emergence of a Ca 2ϩ influx pathway that is resistant to known blockers of voltage-gated Ca 2ϩ channels (22). The present study was aimed at identifying this induced Ca 2ϩ influx pathway, and our results indicate that chronic hypoxia increases evoked exocytosis by increasing the production of A␤Ps that form Ca 2ϩ -permeable membrane channels that are tightly coupled to exocytosis. Given the known increased incidence of Alzheimer's disease following ischemia as described above, these findings are likely to be of widespread importance in understanding causes and effects of increased dementias following ischemic insult. EXPERIMENTAL PROCEDURES PC12 cells were obtained from the American Type Culture Collection (Manassas, VA) and were cultured as described previously (20,22) in RPMI 1640 medium (containing L-glutamine) supplemented with 20% fetal calf serum and 1% penicillin/streptomycin (all from Gibco, Paisley, Strathclyde, United Kingdom). Cells were incubated at 37°C in a humidified atmosphere of 5% CO 2 and 95% air, passaged every 7 days, and used for up to 20 passages. Cells used for experiments were transferred to smaller flasks in 10 ml of medium, to which was added 1 M dexamethasone (Sigma, Poole, UK; from a stock solution of 1 mM in ultrapure water), and were cultured for a further 72-96 h to enrich catecholamine stores (23). Cells exposed to chronic hypoxia were treated identically, except that for 6 -24 h prior to experiments, they were transferred to a humidified incubator equilibrated with 10% O 2 , 5% CO 2 , and 85% N 2 . Following this period in chronic hypoxia, cells were exposed to room air for no longer than 1 h before experimentation.
Each experimental day, PC12 cells were plated onto poly-L-lysinecoated coverslips and allowed to adhere for ϳ1 h under either normoxic or hypoxic (10% O 2 ) conditions, as required. Fragments of coverslip were then transferred to a recording chamber (volume of 80 l), which was continually perfused under gravity (flow rate of 1-2 ml/min) with a solution of 135 mM NaCl, 5 mM KCl, 1.2 mM MgSO 4 , 2.5 mM CaCl 2 , 5 mM Hepes, and 10 mM glucose (pH 7.4; osmolarity adjusted to 300 mosM with sucrose, 21-24°C). Ca 2ϩ -free solutions contained 1 mM EGTA and no added CaCl 2 . All drugs were applied in the perfusate except Congo red and the 3D6 antibody. Congo red was added to cells in culture for the final 3-4 h before plating. The 3D6 antibody was diluted to a final concentration of 5 g/ml in standard extracellular perfusate, and cells were exposed to this solution for 1 h at 37°C. Exposure of cells to antibody-free perfusate under the same conditions was without effect on the ability of cells to secrete catecholamines in response to depolarization (data not shown). Hypoxic solutions were obtained by continually bubbling one or more of the reservoirs supplying the recording chamber with N 2 as required. Hypoxic reservoirs were pre-equilibrated with N 2 for at least 30 min before experiments.
Carbon fiber microelectrodes (proCFE, Axon Instruments, Inc., Foster City, CA) with a diameter of 5 m were positioned adjacent to individual PC12 cells using a Narishige MX-2 micromanipulator and were polarized to ϩ800 mV to allow oxidation of released catecholamine. Resulting currents were recorded using an Axopatch 200A amplifier (with extended voltage range), filtered at 1 kHz, and digitized at 2 kHz before computer storage. All acquisition was performed using a Digidata 1200 interface and Fetchex software from the pClamp 6.0.3 suite (Axon Instruments, Inc.). The same equipment was also used to monitor PO 2 levels in the recording chamber, except that the polarity of the microelectrode was reversed to Ϫ800 mV (24). The time course of fall in PO 2 in the recording chamber was highly reproducible and was stable after 1 min (between 17 and 23 mm Hg), and analysis of the effects of acute hypoxia was conducted 90 s after exchange to hypoxic perfusate.
Unless otherwise stated, each experiment consisted of current recordings of a control period during which cells were perfused only with normoxic external medium. This was then exchanged for a test solution, and amperometric signals were recorded for a further period of 1-4 min. Catecholamine secretion was apparent as discrete spike-like events, each corresponding to the released contents of a single vesicle of catecholamine (17,25). We did not distinguish between dopamine and noradrenaline, both known to be released from PC12 cells (26); but secretory events were never seen unless the electrode was polarized and adjacent to a cell. Quantification of release was achieved by determining spike number or frequency using the Mini Analysis Program (Synaptosoft Inc., Leonia, NJ). This allowed visual inspection of each event so that artifacts (due, for example, to solution switches) could be rejected from analysis. Results are presented as individual examples or means Ϯ S.E., and statistical comparisons were made using unpaired Student's t test.
To monitor [Ca 2ϩ ] i , cells were incubated for 1 h at room temperature (21-24°C) in control perfusate solution containing 4 M Fura-2/AM. Cells were then placed in a perfusion chamber exactly as described for amperometric recordings, and changes in [Ca 2ϩ ] i were indicated by the fluorescence emitted at 510 nm due to alternate excitation at 340 and 380 nm, using a Joyce-Loebl PhoCal apparatus (Applied Imaging Inc., Rochester, NY). Data are presented as ratio signals.
Immunofluorescent labeling with the 3D6 antibody was performed with cells plated onto coverslips and subjected to normoxic or hypoxic conditions as described above. Cells were fixed by immersion in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 15 min and then rinsed thoroughly in several changes of 0.1 M phosphate-buffered saline (PBS) over a period of 2 h. The coverslips were incubated with the 3D6 antibody (diluted to 0.5 or 2.0 g/ml in PBS) in 24-well microtitration plates on a shaker for 18 h at 4°C. After two 10-min rinses in PBS, the cells were incubated for 2 h in a 1:200 dilution of Cy2-conjugated anti-mouse IgG (Jackson ImmunoResearch Laboratories, Inc.). After two additional 10-min rinses in PBS, the coverslips were mounted onto glass microscope slides with glycerol/PBS, and the edges of the coverslips were fixed with clear nail polish. The cells were examined using a Zeiss Axioskop epifluorescence microscope using a No. 10 (fluorescein) filter set. Photographs were taken using an Eastman Kodak MDS120 digital camera system.

RESULTS
Exposure of PC12 cells to elevated extracellular [K ϩ ] (50 mM) evoked quantal catecholamine secretion, which was apparent as transient oxidative currents detected with a carbon fiber microelectrode (Fig. 1A). Release evoked by 50 mM K ϩ has previously been shown to be due to membrane depolarization and Ca 2ϩ influx through voltage-gated Ca 2ϩ channels (20,27,28) and so could be completely inhibited by removal of extracellular Ca 2ϩ or by application of the nonselective inhibitor of voltage-gated Ca 2ϩ channels, Cd 2ϩ (Fig. 1A). When cells were cultured in an environment of reduced O 2 (10% instead of the normal 21%) for 24 h, secretion evoked by the same stimulus (50 mM K ϩ ) was significantly enhanced (Fig. 1B), from an average exocytotic frequency of 0.83 Ϯ 0.12 Hz (n ϭ 26 cells) observed in control cells to 1.93 Ϯ 0.16 Hz (n ϭ 19; p Ͻ 0.001, unpaired t test). In both control cells (n ϭ 70) and cells maintained in chronic hypoxia (n ϭ 75), no secretion was detected until the cells were exposed to 50 mM K ϩ (data not shown). These findings for the effects of chronic hypoxia to enhance K ϩ -evoked secretion are extremely similar to our previously reported effects of chronic hypoxia on the secretory responses to acute hypoxia (22).
Enhancement of evoked secretion from cells cultured under chronically hypoxic conditions remained entirely dependent on the presence of extracellular Ca 2ϩ since removal of Ca 2ϩ from the perfusate (replaced with 1 mM EGTA) caused complete cessation of exocytosis (Fig. 1B, upper trace, representative of recordings from nine cells). However, Cd 2ϩ (200 M) was unable to prevent fully the secretory response to 50 mM K ϩ (Fig.  1B, lower trace, representative of 19 cells). In the presence of 200 M Cd 2ϩ , secretion was significantly reduced (p Ͻ 0.002) to 0.72 Ϯ 0.11 Hz. Thus, a period of chronic hypoxia induced an enhanced secretory response from PC12 cells that was entirely Ca 2ϩ -dependent, but ϳ37% of which was resistant to Cd 2ϩ (i.e. was not mediated by Ca 2ϩ influx through known voltage-gated Ca 2ϩ channels).
A Cd 2ϩ -resistant Ca 2ϩ entry pathway induced by 24 h of hypoxia was also indicated by microfluorometric recordings, as shown in Fig. 2. For control cells ( Fig. 2A), bath application of solution containing 50 mM K ϩ caused a rapid and reversible rise of [Ca 2ϩ ] i that could be attributed to Ca 2ϩ influx through voltage-gated Ca 2ϩ channels since it was completely abolished by 200 M Cd 2ϩ (co-applied as indicated by the horizontal bars in Fig. 2). By contrast, rises of [Ca 2ϩ ] i measured in chronically hypoxic PC12 cells could only be reduced by ϳ60% in the presence of 200 M Cd 2ϩ (Fig. 2B). Mean changes in [Ca 2ϩ ] i evoked by 50 mM K ϩ in control (open bars) and chronically hypoxic (closed bars) PC12 cells are summarized in Fig. 2C.
To characterize further the Cd 2ϩ -resistant Ca 2ϩ influx pathway coupled to secretion that was apparent after chronic hypoxia, we tested the effects of other known blockers of Ca 2ϩ influx pathways in cells previously maintained in hypoxia for 24 h. We have previously found that Cd 2ϩ -resistant secretion evoked in these cells by acute hypoxia could be significantly reduced by inorganic cations (22). These cations (Zn 2ϩ and La 3ϩ ) also reduced Cd 2ϩ -resistant secretion evoked by 50 mM K ϩ in chronically hypoxic PC12 cells (Fig. 3). Furthermore, Cd 2ϩ -resistant release evoked by 50 mM K ϩ was also inhibited by Congo red (Fig. 3). Both Congo red and Zn 2ϩ have been shown previously to inhibit A␤P-mediated Ca 2ϩ fluxes (29,30), suggesting the surprising possibility that a 24-h period of hypoxia induced formation of Ca 2ϩ -permeable A␤P channels that coupled closely to exocytosis. In further support of this, we found that exposure of cells to a monoclonal antibody (3D6) raised against the extracellularly located NЈ terminus of A␤P, anti-A␤P-(1-5) (30, 31), for 1 h at 37°C also significantly suppressed the Cd 2ϩ -resistant component of the secretory response (Fig. 3). Similar inhibition of Cd 2ϩ -resistant secretion evoked in response to acute hypoxia (PO 2 ϳ 20 mm Hg) was also found with Congo red and the 3D6 antibody (data not shown).
Immunocytochemical labeling of cells with the 3D6 antibody demonstrated specific binding of this antibody to cells previously exposed to chronic hypoxia for 24 h (Fig. 4A). This was seen as diffuse fluorescence over most of the surface area of each cell. This was a consistent pattern of labeling in all cells examined (Ͼ100) at both concentrations of the antibody tested (0.5 and 2.0 g/ml), and this is exemplified by comparison of the phase-contrast image of the same cells (Fig. 4B). Fluorescent labeling was detectable only at very low levels in the normoxic cells processed under identical conditions ( Fig. 4C; phase-contrast view of the same image shown in Fig. 4D).
If chronic hypoxia did indeed induce formation of A␤P channels that coupled to exocytosis, we anticipated that direct application of A␤Ps to normoxically cultured cells would mimic the effects of chronic hypoxia. Therefore, in a separate series of experiments, we exposed cells to 100 nM A␤P-(1-40) for 24 h under normoxic conditions (21% O 2 ). These A␤P-treated cells secreted catecholamines in response to both acute hypoxia and raised K ϩ levels (Fig. 5A). Like both control and chronically hypoxic cells, they showed no spontaneous exocytosis under normoxic, non-depolarizing conditions (n ϭ 47 cells; data not shown). Secretion evoked by either 50 mM K ϩ or acute hypoxia was significantly greater (p Ͻ 0.001) than in control cells (Table  I) and was wholly dependent on the presence of extracellular Ca 2ϩ , and a substantial fraction was resistant to Cd 2ϩ (Fig. 5A; see also Table I); such effects of A␤P treatment were extremely similar, both qualitatively and quantitatively, to the effects of chronic hypoxia (see above). Furthermore, the Cd 2ϩ -resistant component of the secretory response displayed the same sensitivity to other blockers as was observed in cells kept chronically hypoxic for 24 h, including the 3D6 monoclonal antibody (Fig.  5B). These findings strongly support the idea that chronic hypoxia appeared to induce formation of A␤Ps tightly coupled to exocytosis. resistant, K ϩ -evoked secretion in cells either kept chronically hypoxic (open circles) or exposed to 100 nM A␤P-(1-40) (closed circles). It is clear from this figure that exposure to hypoxia for up to 6 h was without effect on the secretory responses to 50 mM K ϩ (closed circles). Thereafter, the Cd 2ϩ -resistant component of secretion increased steeply for cells exposed to chronic hypoxia for between 12 and 24 h. This component of the secretory response then declined gradually following 30 -48 h of exposure to hypoxia. These findings indicate that enhancement of secretion caused by the ability of hypoxia to induce Cd 2ϩresistant, Ca 2ϩ -dependent secretion was transient in nature and was maximal after 24 h of chronic hypoxia. By contrast, in cells treated with A␤P-(1-40), Cd 2ϩ -resistant secretion was apparent even after 1 h of treatment, was maximal at 6 h, and remained approximately constant for up to 24 h, after which time point there was a gradual decline in this component of secretion. Thus, there was a clear lag in the emergence of Cd 2ϩ -resistant, K ϩ -evoked secretion induced by hypoxia as compared with A␤P-(1-40)-treated cells.

DISCUSSION
Prolonged periods of hypoxia are well known to alter gene expression and hence functional expression of numerous proteins (11). Several groups have documented altered levels of ion channel expression (particularly K ϩ channel expression) following periods of chronic hypoxia in tissues such as pulmonary vascular smooth muscle (32) and the carotid body (33). In this study, we investigated the effects of prolonged hypoxia on the secretory responses of PC12 cells to both acute hypoxia and raised extracellular [K ϩ ] since these cells have been used extensively as a model system for studying the effects of chronic hypoxia on gene expression (16,18). As anticipated from these previous studies, we found that PC12 cells released more vesicles of catecholamine when stimulated by either acute hypoxia or exposure to tetraethylammonium (via blockade of O 2 -sensitive K ϩ channels) and that each vesicle contained greater amounts of catecholamine (22) due to increased levels of tyrosine hydroxylase (16). What was completely unexpected, however, was the emergence of a Ca 2ϩ influx pathway coupled to secretion that was not attributable to any of the known voltagegated Ca 2ϩ channels (L-, N-, or P/Q-type (34)) present in PC12 cells (28). This study provides compelling evidence that this pathway is due at least in part to Ca 2ϩ -permeable channels formed by A␤Ps. Thus, in the presence of a supramaximal concentration of Cd 2ϩ to block voltage-gated Ca 2ϩ channels, residual secretion (constituting ϳ37% of total secretion in response to 50 mM K ϩ ), arising from Ca 2ϩ influx as indicated by microfluorometric recordings (Fig. 2), was further reduced by known blockers of A␤Ps, including Zn 2ϩ , Congo red, and the 3D6 monoclonal antibody (Fig. 3).
A␤Ps are neurotoxic products associated with Alzheimer's disease (4). The mechanisms underlying neuronal toxicity caused by A␤Ps are complex and remain to be fully resolved, although most evidence indicates that toxicity involves A␤P disruption of intracellular Ca 2ϩ homeostasis (4,35). There is also strong evidence that toxicity is oxidative and involves free radical damage (36 -38). An alternative (or perhaps additional) hypothesis indicates that A␤Ps disrupt intracellular Ca 2ϩ homeostasis by forming Zn 2ϩ -sensitive, Ca 2ϩ -permeable channels (39,40). Such an effect may account for increased central synaptic activity. Indeed, enhancement of long-term potentiation and elevated glutamate release have been demonstrated in hippocampal neurons exposed to A␤Ps in vitro (41,42).
It is possible that hypoxic induction of A␤Ps enhanced secre- FIG. 4. Immunofluorescent labeling of chronically hypoxic cells using an A␤P-directed monoclonal antibody. A, fluorescence image of individual cells (arrows) that had previously been exposed to hypoxia (10% O 2 ) for 24 h and exposed to the 3D6 antibody; B, phasecontrast image of the same field; C, fluorescence image of individual cells (arrows) that had previously been cultured under normoxic conditions; D, phase-contrast image of the same field. Scale bars ϭ 50 m. Processing of cells led to flattening, leading to a larger apparent diameter than when they were perfused with solutions used for electrochemical recordings.

FIG. 5. Exposure of cells to A␤P-(1-40) mimics the effects of chronic hypoxia.
A, amperometric detection of secretion evoked in cells exposed previously to 100 nM A␤P-  in response to either acute hypoxia (upper traces; PO 2 ϳ 20 mm Hg, application began 1 min before beginning of traces) or 50 mM K ϩ (lower traces; stimulus application began 10 s before beginning of traces). For the periods indicated by the horizontal bars, cells were exposed either to Ca 2ϩ -free perfusate tion by altering the activity of native voltage-gated Ca 2ϩ channels, acting in a manner comparable to an auxiliary subunit (43). However, we consider this unlikely; Cd 2ϩ is a Ca 2ϩ channel pore blocker (i.e. it binds to similar regions within the pore-forming ␣ subunit of Ca 2ϩ channels), and it is highly unlikely that interaction of A␤Ps with native Ca 2ϩ channels could cause conformational changes that would render the pore Cd 2ϩ -insensitive, yet more permeable to Ca 2ϩ (to allow greater influx to account for greater secretory responses). We have not addressed the possibility that A␤Ps could also act intracellularly to modulate secretory responses, and such studies are worthy of future investigation. The simplest interpretation of our present findings, particularly in light of the knowledge that A␤Ps have been demonstrated to form Ca 2ϩ permeable channels in membranes (see above), is that chronic hypoxia causes formation, in the plasma membrane, of Ca 2ϩ -permeable A␤P channels, which permit Ca 2ϩ entry sufficient to trigger exocytosis. This interpretation also raises the further question of gating: following chronic hypoxia or direct application of A␤Ps to PC12 cells, there was no detectable basal secretion, and cells required stimulation (by either acute hypoxia or 50 mM K ϩ ) before enhanced secretion was observed, indicating that this Ca 2ϩ influx pathway was voltage-dependent. This finding contrasts with studies in which A␤Ps were inserted into planar lipid bilayers: the channels formed did not display marked voltage dependence (39). We cannot account for this discrep-ancy at present, but it should be noted that L-type Ca 2ϩ channels (which always display strong voltage dependence in cells), when purified and inserted into lipid bilayers, also display very little voltage dependence over a wide range of potentials (44). Thus, marked functional differences in channel activity when observed in lipid bilayers compared with cell membranes are to be expected.
It is noteworthy that the known blockers of A␤Ps failed to inhibit completely the Cd 2ϩ -resistant secretion in cells previously exposed to chronic hypoxia (Fig. 3). This suggests that formation of Ca 2ϩ -permeable A␤P channels might not account fully for the Cd 2ϩ -resistant Ca 2ϩ influx pathway. Interestingly, these blockers also failed to inhibit fully the Cd 2ϩ -resistant secretion observed in cells pretreated with A␤P-(1-40) (Fig. 5). We cannot at present account for this incomplete blockade. Nevertheless, our present findings indicate that A␤Ps provide a Ca 2ϩ influx pathway that can specifically trigger exocytosis. It should be noted that only Ca 2ϩ influx specifically through N-type voltage-gated Ca 2ϩ channels, but not L-or P/Q-type channels, influences depolarization-evoked exocytosis in control PC12 cells (20,22). This suggests the possibility of an intimate association of A␤Ps with vesicle docking/fusion sites in a manner that may compare with the interaction of known voltage-gated Ca 2ϩ channels and synaptic proteins (43). Most important, we have demonstrated that reduced O 2 levels alone can induce formation of A␤P-mediated Ca 2ϩ influx. As described in the Introduction, A␤Ps are the cleavage products derived from amyloid precursor protein (5,6), and levels of this precursor have been shown to increase following ischemia (9,10). The enhanced Ca 2ϩ influx reported here may therefore be an important contributory factor to the increased incidence of dementias and amyloid deposition following cerebral ischemia, leading to neurotoxicity through excessive transmitter release. I Effects of ␤-amyloid treatment on secretory responses and their modulation by Ca 2ϩ removal or application of Cd 2Ϫ Cells were exposed for 21-26 h to 100 nM A␤P-(1-40) under normoxic conditions. The mean Ϯ S.E. exocytotic frequency, evoked by acute hypoxia (PO 2 ϳ20 mm Hg) or 50 mM K ϩ , was determined from the number of cells indicated. The incomplete inhibition of secretion caused by Ca 2ϩ in A␤P-treated cells was statistically significant (see Footnotes a and b).