Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Interest in the role of basal forebrain cholinergic neurons in cognitive processes, including attention, learning, and memory, has been sustained over the past two decades largely because of the marked degeneration of these neurons in Alzheimer's disease (Davies and Maloney, 1976; Perry et al., 1977; Whitehouse et al., 1982). Although the relevance of numerous studies to the basal forebrain cholinergic hypothesis of learning and memory has been questioned (Dunnett et al., 1991; Fibiger, 1991), most recent data clearly suggest their involvement in cognitive processes (Winkler et al., 1995; Baxter et al., 1997; Fisahn et al., 1998). Accordingly, there has been revitalized interest in the detailed functioning of basal forebrain cholinergic neurons, as well as their relation to pathological features of Alzheimer's disease (Auld et al., 1998; Quirion et al., 1998).

Basal forebrain neuron acetylcholine (ACh) release is usually studied directly with in vivo microdialysis, in vitro slice (superfusion and static incubation) preparations, or synaptosomal preparations. Each system presents various disadvantages as well as advantages. For example, in vivo microdialysis (e.g., hippocampus or cortex) is not particularly well suited for mechanistic and cellular studies, or for the study of certain molecules, considering the caveats of probe passage and tissue penetration. Furthermore, it is currently impossible to reliably clear exogenous probe-delivered molecules from brain tissue in vivo. In in vitro slice preparations (e.g., hippocampus or cortex) cholinergic terminals are disconnected from their cell bodies and, as such, do not consist of intact cholinergic neurons and have limited viability, thereby restricting experiment duration. Most slice studies consider only stimulated ACh release because nonstimulated release is most often thought to be nonspecific leakage from ruptured neurons. Slice methods also suffer from slow tissue penetration and drug clearance. Hippocampal or cortical synaptosome preparations are very useful for mechanistic studies; however, the rather disruptive preparation may alter membrane protein properties, the completely isolated terminal is rather nonphysiological, and the experiment is constrained by the limited viability of the synaptosomes.

Because our laboratory has regularly encountered these procedural impediments, we sought to circumvent some of them by evaluating basal forebrain ACh release within another experimental paradigm, namely, rat primary embryonic septal cultures. We felt that this model would be particularly adaptable to a variety of experimental paradigms not possible with other approaches. Surprisingly, in previous literature, the release of endogenous ACh from septal cultures has received limited attention. Although ACh release has been reported in other culture models [e.g., postnatal basal forebrain neurons (Allen and Brown, 1996)], to our knowledge, only one previous study has used primary embryonic septal cultures to investigate endogenous ACh release and, in this case, K⁺-stimulated ACh release from mature cultures was contingent on a 3-day nerve growth factor treatment (Suzuki et al., 1994). Indeed, there has never been a detailed characterization of the basic features of endogenous ACh release from primary embryonic septal cultures. Accordingly, it is unknown whether septal cultures release ACh in a manner consistent with that observed in other systems. A detailed characterization is also necessary before further research concerning ACh release from septal cultures can be interpreted in a more complete manner. Furthermore, considering that these cultures are increasingly used as a model of the basal forebrain cholinergic system (Hartikka and Hefti, 1988; Lorenzi et al., 1992; Nonner et al., 1996; Pongrac and Rylett, 1996; Mennicken and Quirion, 1997), we thought it important to evaluate the characteristics of ACh release because release is an event of key importance in neuronal function.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Primary Rat Septal Cultures. Cultures were prepared as described previously (Mennicken and Quirion, 1997). Septal areas (containing cholinergic neurons from the septum, diagonal band of Broca, and substantia innominata) of day 17 rat embryos were dissected in Hanks' solution (Life Technologies, Burlington, Ontario, Canada) containing 0.65% D-(+)-glucose (Sigma Chemical Co., St. Louis, MO), 15 mM HEPES, 10 U/ml penicillin, and 10 mg/ml streptomycin (all obtained from Life Technologies). These were then dissociated enzymatically at 37°C with 0.08% trypsin (Life Technologies) and 0.1% DNase I (Sigma Chemical Co.) for 18 min. This was terminated with 10% fetal bovine serum (FBS; Life Technologies); the dissociation was completed mechanically with several aspirations into a fire-polished small-bore Pasteur pipette. Cells were plated at 600,000 cells/well in four-well tissue culture plates (Nunc, Naperville, IL) previously coated with 25-µg/ml poly(D-lysine) (Sigma Chemical Co.). The growth medium consisted of Dulbecco's modified Eagle's medium (no. 11965; Life Technologies) supplemented with KCl (20 mM), sodium pyruvate (1 mM), D-glucose (35 mM), and HEPES (15 mM), and with 10% FBS (Sigma Chemical Co.) or with the serum-free supplement B27 (2%; Life Technologies), as indicated. Cultures were kept at 37°C (5% CO₂) for up to 2 weeks.

ACh Release. Most release experiments were conducted 8 days after dissociation and plating [day in vitro (DIV) 8]. Before the experiment, the growth medium was removed and the cells were rinsed (500 µl) with a Krebs-like buffer [125 mM NaCl, 4.8 mM KCl, 1.2 mM KH₂PO₄, 25 mM HEPES, 1.2 mM MgSO₄(7H₂O), 2.2 mM CaCl₂(2H₂O), 10 mM glucose (Sigma Chemical Co.), 10 µM choline (Sigma Chemical Co.), and 100 nM neostigmine (Sigma Chemical Co.)] containing 6 mM K⁺ (note the presence of 4.8 mM KCl and 1.2 nM KH₂PO₄). After a 60-min equilibration period at 37°C and 5% CO₂, this buffer was removed and replaced with fresh buffer (250 µl). After 10 min to establish a basal (nonstimulated) level of ACh release, this buffer was removed and replaced with buffer (250 µl) containing 25 mM K⁺ to stimulate ACh release. The stimulating buffer was replaced at 10-min intervals, the resulting samples were placed on dry ice, and the plates with fresh buffer were immediately returned to the incubator (37°C; 5% CO₂). All samples were kept at 80°C until HPLC estimation of ACh content (<2 weeks). As indicated in Results, several experiments involved varying the concentration and presence of components of the assay buffers. Similarly, drugs [i.e., veratridine (Sigma Chemical Co.), vesamicol (Research Biochemicals, Natick, MA), tetrodotoxin (TTX; Sigma Chemical Co.), AF-DX 384 (K. Thomae GmbH, Biberach, Germany), and oxotremorine (Sigma Chemical Co.)] were applied by inclusion in the assay buffers. Data were expressed as means ± S.E., and, in most cases, one-way ANOVAs with Tukey's post hoc tests were used to determine statistical significance.

ACh Quantification by HPLC. ACh was assayed by HPLC with electrochemical detection in conjunction with an enzyme reactor (Damsma et al., 1987); the separation column, enzyme reactor, and electrode were connected in series. Samples (100 µl) were injected either manually via a 100-µl loop on a two-position valve (Valco, Houston, TX) or by a WISP 710B automatic sample injector (Waters, Milford, MA). ACh and choline, separated on a reversed phase column (75 × 2.1 mm) pretreated with lauryl sulfate, pass through an enzyme reactor (10 × 2.1 mm) containing acetylcholinesterase (AChE; EC 3.1.1.7; type VI-S; Sigma Chemical Co.) and choline oxidase (1.1.3.17; Sigma Chemical Co.), covalently bound to glutaraldehyde-activated Lichrosorb NH₂ (10 µm; Merck, Darmstadt, Germany). All column hardware and packing materials were purchased from Chrompack (Raritan, MA). The resultant hydrogen peroxide is electrochemically detected at a platinum electrode at a potential of +500 mV versus an Ag/AgCl reference electrode (Antec VT-03/Decade; Leiden, the Netherlands). The mobile phase, 0.2 M aqueous potassium phosphate buffer, pH 8.0, containing 1 mM tetramethylammonium hydroxide (Sigma Chemical Co.) is delivered at 0.4 to 0.5 ml/min by a dual piston pump (ESA 580; ESA, Chelmsford, MA) connected to a degasser (CMA 260; Carnegie Medicin, Stockholm, Sweden) and pulse dampener (Lo-Pulse; Scientific Systems, Inc., State College, PA). ACh elutes at ~4 min, and the best detection limit of the assay is ~10 fmol/injection. Sample concentrations were calculated by comparison to known standards.

Choline Acetyltransferase (ChAT) Activity. ChAT activity was determined according to the method of Fonnum (1975), with modifications described previously (Mennicken and Quirion, 1997). Cultures were homogenized in 100 µl of ice-cold homogenizing buffer (40 mM sodium phosphate buffer, pH 7.4, 200 mM NaCl, and 0.5% Triton X-100). Aliquots in duplicate were assayed for ChAT activity with [¹⁴C]acetyl-CoA (New England Nuclear/DuPont, Markham, Ontario, Canada) and choline (Sigma Chemical Co.) as substrate. After 60 min at 37°C, the reaction was stopped with ice-cold 10 mM sodium phosphate buffer, pH 7.4, containing 0.2 mM acetylcholine chloride (Sigma Chemical Co.). Radioactive ACh was extracted with 3-heptanone (Aldrich, Milwaukee, WI) containing 15 mg/ml sodium tetraphenylborate (Aldrich) and quantified with liquid scintillation spectrometry (Tri-Carb 4550; Packard, Downers Grove, IL). ChAT activity was expressed as picomoles of ACh synthesized per hour per well.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

ChAT activity as well as stimulated ([K⁺] = 25 mM) endogenous ACh release (Table 1) increased with culture maturation over several days after plating (DIV 4-DIV 10) in a growth medium containing FBS (10%). Cultures grown in serum-free medium supplemented with B27 (2%) also released ACh (evaluated on DIV 10), although to a lesser extent (595 ± 32 fmol; n = 4) after a stronger stimulation (100 mM K⁺) than cultures grown in the presence of FBS. All experiments for subsequent characterization were performed on DIV 8 with cultures grown in the presence of FBS.

Endogenous ACh release was evaluated under several stimulating conditions (Fig. 1). K⁺ stimulated release in a concentration-dependent manner (6-100 mM) and the Na⁺ channel opener veratridine (25 µM) also increased ACh release. The concentration of K⁺ chosen to be the standard stimulating condition for all other experiments (25 mM) stimulated ACh release robustly but not maximally (Fig. 1).

View larger version (32K): [in this window] [in a new window]   Fig. 1.   Concentration-dependent stimulation of endogenous ACh release from rat primary septal cultures by extracellular K+. DIV 8 cultures were exposed to Krebs-like buffer containing K+ (6-100 mM) or veratridine (25 µM veratridine in Krebs-like buffer containing 6 mM K+). The columns represent the ACh concentration after 10 min of stimulation expressed as femtomoles ACh/well/10 min (mean ± S.E.; n = 3). ACh release from cultures exposed to 15 to 100 mM K+ or 25 µM veratridine was increased significantly compared with release in the presence of 6 mM K+ (one-way ANOVA with Tukey's multiple comparison test, *P < .05, **P < .001).

Stimulated (K⁺ = [25 mM]) endogenous ACh recovered from the cultures was increased by the reversible cholinesterase inhibitor neostigmine (0.001-10 µM) (Table 2). Neostigmine (0.1 µM) was included in the release buffers for all other experiments because it was the lowest concentration providing excellent recovery.

The cultures were able to sustain stimulated endogenous ACh release for 180 min (Fig. 2A). Furthermore, ACh release responded rapidly to alternating basal and stimulating conditions (Fig. 2B). Neither stimulated ACh release (Fig. 2, A and B) nor basal ACh release (Fig. 2B) changed significantly over time.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Sustainable and rapidly responding stimulated endogenous ACh release from rat primary septal cultures on DIV 8. Data are expressed as a percentage of the ACh concentration in the first 10-min stimulated sample (femtomoles ACh/well/10 min; mean ± S.E.; n = 4). A, ACh release does not change significantly over 180 min of stimulation with Krebs-like buffer containing 25 mM K+ (one-way repeated-measures ANOVA). B, ACh release responds dynamically to alternated basal and stimulating releasing buffers (6 or 25 mM K+). The effect of the 25 mM K+ was acute, and no residual effect on ACh release was apparent during the subsequent 6 mM K+ exposure period. ACh release during both the basal and stimulated conditions did not change significantly over time (one-way repeated-measures ANOVA).

The release of endogenous ACh was dependent on extracellular choline (Fig. 3). ACh released within the first 10 min of stimulation was not significantly altered by exposure to different concentrations of extracellular choline (0-10 µM). However, after 20 and 30 min of stimulation, endogenous ACh released from cultures exposed to lower concentrations of extracellular choline (0-1 µM) was significantly reduced compared with release from cultures exposed to 10 µM choline (Fig. 3). Choline (10 µM) was thus included in the releasing buffers in all other experiments.

View larger version (36K): [in this window] [in a new window]   Fig. 3.   Dependence of stimulated endogenous ACh release from rat primary septal cultures on extracellular choline concentrations. DIV 8 cultures were exposed to choline (0-10 µM) during the 60-min incubation period, as well as during the release period. Data are expressed as a percentage of the ACh concentration in the first 10-min stimulated sample in the presence of 10 µM choline (femtomoles ACh/well/10 min; mean ± S.E.; n = 4). In the first stimulated sample, endogenous ACh release was not affected by choline concentration. However, in the second and third stimulated samples, ACh release in the presence of 0 to 1 µM choline was significantly lower than that observed in the presence of 10 µM choline. Variation in endogenous ACh release is explained by the concentration of choline (P < .0001), the sample time (P < .0001), and the interaction between these two variables (P = .0001) (two-way ANOVA). The interaction was further elucidated with Tukey's multiple comparison post test by comparing all other choline concentrations to 10 µM at each time point (*P < .01, **P < .001).

To determine whether the stimulated ACh release was vesicular in origin, cultures were exposed to different concentrations of the vesicular transport inhibitor vesamicol (0-5 µM). Vesamicol reduced the stimulated release of endogenous ACh in a concentration-dependent manner (Fig. 4). Treatment with the highest concentration of vesamicol (5 µM) blocked almost 100% of ACh released (Fig. 4). In addition, nonstimulated basal release was also sensitive to vesamicol (data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 4.   Stimulated endogenous ACh release from rat primary septal cultures is derived from a vesamicol-sensitive pool. DIV 8 cultures were exposed to vesamicol (0-5 µM) during the 60-min incubation period as well as the release period. Data are expressed as a percentage of the stimulated release in the absence of vesamicol (mean ± S.E.; n = 8). Statistical significance was determined with a one-way ANOVA with Tukey's multiple comparison test and compared all concentrations of vesamicol to the absence of vesamicol (*P < .001).

The muscarinic M₂-like receptor antagonist AF-DX 384 (0.001-10 µM) increased stimulated ACh release and this increase was TTX insensitive (Table 3). The nonselective muscarinic receptor agonist oxotremorine (0.001-10 µM) decreased release (Table 3). Neither AF-DX 384 nor oxotremorine altered ChAT activity under our assay conditions (data not shown) and values were similar whether expressed per microgram of protein or per ChAT activity (data not shown).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The results summarized herein indicate that primary embryonic septal cultures are capable of the release of endogenous ACh. The release appears to be neither artifactual nor nonspecific, but rather the consequence of a functioning cholinergic system. In support of this conclusion, several elements of cholinergic neurotransmission were demonstrated to be functional in these primary septal cultures, including: 1) ChAT activity; 2) AChE activity; 3) concentration-dependent stimulation of ACh by K⁺, as well as stimulation by veratridine; 4) vesicular transport and release of ACh; 5) dependence of release on choline availability; and 6) regulation of release by autoreceptors.

The standard unit used to quantify endogenous ACh was femtomoles of ACh per well. Under the conditions used in the present experiments (e.g., no long-term drug treatments with trophic or toxic potential, no differential ChAT activity altering treatments), data expressed according to ACh per well were comparable to data expressed according to other units of measurement. For example, there was little relative difference in response to AF-DX 384 or oxotremorine whether stimulated release was expressed as femtomoles of ACh per well, per protein concentration, or per ChAT activity. Considering the carefully controlled environment of the cultures, as well as the current experimental conditions, the methods used in other systems (e.g., in vitro slice preparation) to normalize release values (e.g., by comparison of protein concentration) do not yield a better representation of release.

Considering that previous studies in primary septal culture have established the activity of AChE as well as its presence almost exclusively on ChAT-positive neurons (94%) (Hefti et al., 1985; Hartikka and Hefti, 1988), our finding that a cholinesterase inhibitor increased recoverable ACh was not surprising. Indeed, as with most other methods of evaluating ACh release (e.g., slice superfusion, microdialysis, etc.), we found neostigmine useful to increase the amount of ACh recovered. However, future experiments requiring the absence of neostigmine or a similar AChE inhibitor should be feasible because without neostigmine, we recorded a strong ACh signal after 10 min of stimulation with 25 mM K⁺. Variation of these parameters should generate even stronger signals (e.g., increasing the stimulation time, increasing K⁺ concentration).

The large range of ACh released in response to different concentrations of extracellular K⁺, from <500 to almost 15,000 fmol/10 min of stimulation, suggests that there is a sizable window for evaluating modulation of ACh release by acute or long-term treatments (e.g., growth factors, toxic compounds, classical antagonists, or agonists). ACh release was shown to be sustainable for a period of at least 180 min with the stimulation conditions of 25 mM K⁺ and a 10-min buffer replacement cycle. It is remarkable that the cultured neurons appear to suffer limited or no fatigue during this rather long time frame. The sustainable nature of release indicates that the synthesis capacity of the cultured septal cholinergic neurons can maintain this level of release, suggesting the appropriateness of the stimulus.

The demonstrated dependence of ACh release from embryonic septal cultures on availability of sufficient extracellular choline is consistent with previous literature reporting from other experimental models (Haga, 1971; Yamamura and Snyder, 1973; Mulder et al., 1974; Murrin et al., 1977; Polak et al., 1977; Jope, 1979; Jope and Jendon, 1980). Because choline (29 µM) and K⁺ (25 mM) are present in the growth medium (DIV 0-DIV 8), it is likely that choline uptake and ACh release are occurring throughout the maturation of these neurons in vitro, and that they are present within the neurons at any given time. On evaluation of release (DIV 8), it was interesting that regardless of extracellular choline concentration, the amount of ACh released was not significantly different during the first 10 min of stimulation with 25 mM K⁺. This suggests that the cholinergic neurons have sufficient stores of choline under normal culture growth conditions to maintain a limited amount of ACh synthesis and release. Between 10 and 20 min of stimulation, it was apparent that new ACh is synthesized from extracellular choline taken into the neurons and used in the releasable pool of ACh, as inferred from the observation that lower release is associated with lower concentrations of extracellular choline. It was apparent that sustainable release of ACh was associated with 10 µM choline, which was stable for 180 min. During the time evaluated, low ACh release was apparent even in conditions without extracellular choline. The source of ACh/choline under this condition could include: 1) stored ACh that is not immediately available for exocytosis (Collier et al., 1993), 2) ACh synthesized from membrane phospholipids (Wurtman, 1992), and/or 3) ACh synthesized from choline reinternalized after cholinesterase activity (Collier and MacIntosh, 1969; Collier and Katz, 1974). Hence, data obtained herein are in accordance with the literature in the field obtained with various other models.

The cellular pool of releasable ACh was investigated with vesamicol, a well established blocker of newly synthesized ACh into specialized synaptic vesicles (Marshall, 1970; Anderson et al., 1983; Ricny and Collier, 1986; Marshall and Parsons, 1988). Pre- and costimulation exposure with vesamicol reduced ACh release, with the highest concentration evaluated (5 µM) completely blocking ACh release. These data suggest that synaptic vesicles are the primary source for released ACh from the cultures in our model, rather than nonspecific or nonphysiological leakage, or via mediatophores, observed in other preparations (Israel and Dunant, 1993).

Interestingly, basal release is also vesamicol-sensitive. This suggests that nonstimulated spontaneous ACh release is regulated by vesicular exocytosis, rather than by nonspecific leakage, as is usually assumed in other in vitro models (e.g., slice preparations). That spontaneous release is present in primary cultures is not surprising, considering that action potentials and spontaneous excitatory/inhibitory postsynaptic potential have been well documented (Li et al., 1998; Boulanger and Poo, 1999).

The blockade of M₂-like receptors on septal cholinergic neurons is known to increase ACh release, whereas stimulation by agonists has the opposite effect (Raiteri et al., 1984). Hence, it has been proposed that the muscarinic M₂ receptor acts as a cholinergic autoreceptor regulating ACh release (Raiteri et al., 1984; Mash et al., 1985; Lapchak et al., 1989; Quirion et al., 1995). In basal forebrain cholinergic projection neurons, mRNA and protein for the m₂ receptor subtype are found at high levels (Buckley et al., 1988; Vilaro et al., 1992, 1994; Levey et al., 1995; Rouse and Levey, 1996). In agreement with these findings is the observation that increases of in vivo hippocampal ACh release induced by AF-DX 384 were attenuated by antisense against the m_2, but not the m₄, receptor (Kitaichi et al., 1999). Accordingly, we have evaluated whether such an autoinhibitory mechanism exists in primary septal cultures. The preferential M₂-like receptor antagonist AF-DX 384 enhanced endogenous ACh release stimulated by high K⁺ in a concentration-dependent manner, whereas the nonspecific muscarinic agonist oxotremorine decreased this measure, suggesting that the muscarinic M₂-like receptor functions as an autoreceptor in the septal culture model. Moreover, the insensitivity of the AF-DX 384 effect to TTX suggests a site of action in the vicinity of, or directly on, the nerve terminals. Thus, endogenous ACh release from primary rat embryonic septal cultures is probably autoregulated in a manner similar to that observed in vitro (Raiteri et al., 1984) and in vivo (Quirion et al., 1995).

We have shown that primary embryonic septal cultures release endogenous ACh in a manner functionally consistent with other models, giving further credit to their use in the study of cholinergic system physiology. More detailed studies of septal culture ACh release can now be considered in perspective of the basic characteristics described in this report. Use of this release system should complement the standard models used to evaluate ACh release (e.g., microdialysis, slice and synaptosome preparations). Indeed, septal cultures offer several advantages that could circumvent restrictions associated with other model systems. For instance, they are suitable for mechanistic, cellular, and molecular investigation and the excellent growth medium/assay buffer exposure facilitates large-molecule (e.g., growth factors) studies and rapid drug clearance. Another advantage is that the intact neuron, rather than a portion of it, is studied with a minimum of potentially disruptive preparation. Accordingly, the experimental duration is not limited by neuron viability and could conceivably last weeks, thereby permitting evaluation of both acute and chronic consequences of a treatment in the same neurons. Also, the importance of assessing ACh release from septal cultures is emphasized considering that release and ChAT activity (the marker most commonly quantified) can have dissimilar responses to treatments, for example, with neurotrophins (Auld and Quirion, 1999).

In conclusion, the rat primary embryonic septal culture system for evaluating endogenous ACh release should be a particularly valuable tool for the study of cellular, molecular, and neurochemical features of ACh release, and also should be very useful for elucidating mechanisms of disease states, including Alzheimer's disease.