Aβ Mediated Diminution of MTT Reduction—An Artefact of Single Cell Culture?

The 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazoliumbromide (MTT) reduction assay is a frequently used and easily reproducible method to measure beta-amyloid (Aβ) toxicity in different types of single cell culture. To our knowledge, the influence of Aβ on MTT reduction has never been tested in more complex tissue. Initially, we reproduced the disturbed MTT reduction in neuron and astroglia primary cell cultures from rats as well as in the BV2 microglia cell line, utilizing four different Aβ species, namely freshly dissolved Aβ (25-35), fibrillar Aβ (1-40), oligomeric Aβ (1-42) and oligomeric Aβ (1-40). In contrast to the findings in single cell cultures, none of these Aβ species altered MTT reduction in rat organotypic hippocampal slice cultures (OHC). Moreover, application of Aβ to acutely isolated hippocampal slices from adult rats and in vivo intracerebroventricular injection of Aβ also did not influence the MTT reduction in the respective tissue. Failure of Aβ penetration into the tissue cannot explain the differences between single cells and the more complex brain tissue. Thus electrophysiological investigations disclosed an impairment of long-term potentiation (LTP) in the CA1 region of hippocampal slices from rat by application of oligomeric Aβ (1-40), but not by freshly dissolved Aβ (25-35) or fibrillar Aβ (1-40). In conclusion, the experiments revealed a glaring discrepancy between single cell cultures and complex brain tissue regarding the effect of different Aβ species on MTT reduction. Particularly, the differential effect of oligomeric versus other Aβ forms on LTP was not reflected in the MTT reduction assay. This may indicate that the Aβ oligomer effect on synaptic function reflected by LTP impairment precedes changes in formazane formation rate or that cells embedded in a more natural environment in the tissue are less susceptible to damage by Aβ, raising cautions against the consideration of single cell MTT reduction activity as a reliable assay in Alzheimer's drug discovery studies.


Introduction
Deposits of beta-amyloid (Ab) and neurofibrillary tangles are the two pathological hallmarks of Alzheimer's disease. There is recent evidence that soluble Ab aggregates can impair function, morphology and subsequently the viability of neuronal cells [1]. Based on NADH dependent reduction activity, cells are able to reduce the tetrazolium salt MTT [3-(4,5-dimethylthiazol-2-yl)-2,5diphenyltetrazolium bromide] into a formazane [2]. Thus, it is widely accepted that the amount of formazane production correlates with both the number and the viability of the cells. The MTT assay is well established for investigations of cellular viability in single cell cultures [3] and tissue slices [4,5]. The MTT assay is frequently used to evidence Ab related changes in membrane properties and disturbed cellular viability [6,7]. The question how Ab inhibits cellular MTT reduction is still a matter of debate. Based on their findings that Ab potently inhibits cellular reduction of MTT in cultured rat hippocampal neurons and HeLa cell lines, Kaneko et al. (1995) have hypothesized that Ab specifically suppresses mitochondrial succinate dehydrogenase [8]. Studies on rat brain tumor cells [9] and astrocytes [10], on the other hand, indicated that Ab decreases cellular MTT reduction by accelerating the exocytosis of MTT formazan.
Although many in vitro findings on Ab toxicity and competing, protective agents are based on the MTT assay [11][12][13][14], the influence of Ab on MTT reduction has never been tested in more complex models than single cell cultures. Organotypic hippocampal slices (OHC) are an in vitro model that retains the three dimensional structure of in vivo systems and ranges in complexity between primary cell cultures and intact animals [15]. OHCs represent a well established tool for the investigation of brain damage due to oxygen glucose deprivation (OGD) [16] or epilepsy [17]. When OHCs were exposed to very high doses of Ab ($10 mM) neuronal apoptotic cell death [18,19] and a pronounced activation of astrocytes [20] occurred. More subtle submicromolar Ab concentrations caused a retraction of neuronal dendrites and a degeneration of dendritic spines [21]. Although it has been shown that MTT is appropriate to evaluate the viability of brain tissue slices and its reduction is impaired after detrimental treatment, such as OGD [5,22], the influence of Ab on MTT reduction in OHCs has never been tested before.
Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), Ab  and Ab  failed to impair MTT reduction in OHC Compared to single cells, the MTT reduction in OHCs was less frequently investigated. Therefore, we characterized the MTT assay in our system and examined its practicability to measure cell toxicity in OHCs. Similar to single cells, OHC produced the first formazane crystals immediately after MTT application and the reaction was saturated within 3 hours (Figure 2A). The MTT activity was diminished to 17.563.4% by application of 15 mM glutamate ( Figure 2B). Since this is an approved model for excitotoxicity related cell damage [17], we considered the MTT reduction assay to be suitable for the detection of cell damage in OHCs.
In order to rule out that diffusion problems due to the size of the Ab aggregates impede toxicity in the OHCs, we immunostained cross sections of OHCs after Ab (1-40) treatment. Ab was clearly marked within the slice ( Figure 2D). Furthermore and in line with the literature [20,34], Ab (1-42) caused an activation of astroglia, as demonstrated by an increased expression of GFAP ( Figure 2E). These results indicate that Ab was able to affect the astroglia within the OHC, although Ab failed to disturb the MTT reduction of the slice.

Separation of single cells from OHCs and treatment with Ab
Considering our conflicting findings in single cells and OHCs it appeared likely that the susceptibility of cells to Ab mediated diminution of MTT reduction activity depends on their environment. To address this matter, we split one OHC preparation into two groups. One group was cultivated further and the other group was separated into single cells. For the first time we prepared single cells from OHCs. Because of the matured state of the isolated cells only few neurons survived the isolation procedure and thus the cultures consisted largely of astrocytes ( Figure 3). When we exposed the slices to Ab (25-35) before the separation and measured the MTT reduction activity two hours after the preparation, there was no effect of Ab (25-35) on the MTT reduction of both, the slices and the single cells ( Figure 3A). In contrast, when the slices were first separated and then Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) was applied to the two groups for 2 days, Ab diminished the MTT reduction in single cell cultures but not in OHCs (Ab (25-35) 80.1%61.0% control: 100%61,1%; Figure 3B). Ab related impairment of LTP is restricted to a particular Ab species and does not correlate with MTT reduction in acute hippocampal slices To further substantiate the assumption that cells within tissuelike structures react different to Ab than single cells, we exposed acutely isolated hippocampal slices from adult rats to 500 nM Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) or 500 nM oligomeric Ab (1-40) or 1 mM fibrillar Ab  and measured the influence on LTP. When we exposed slices to Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) Figure 4A). Because of their large molecule size Ab (1-40) fibrils were expected to have limited and slow access to neuronal target structures. Therefore, we exposed slices to Ab (1-40) fibrils persistently throughout the experiment and with a relatively high concentration of 1 mM. However, application of fibrillar Ab (1-40) did not alter LTP (fibrillar Ab (1-40): 187.1616.6%, n = 8, of baseline value, 240 min after tetanus application; Figure 4A). To investigate whether the disturbed LTP caused by Ab (1-40) oligomers correlates with a diminished MTT reduction, we applied MTT to acute slices in parallel to LTP recording. There was no difference in the MTT reduction between control, Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) and Ab (1-40) oligomer treated slices (ACSF control: 100.0625.0%, Ab (1-40) 107.9%627.0%; Ab (25-35) 106.7%615.7% Figure 4B).

Discussion
In this study we compared the effect of different Ab species on the MTT reduction activity in hippocampal neurons, astrocytes, microglia, OHCs, acutely isolated hippocampal slices from adult animals and the hippocampal formation in vivo. We showed that all tested Ab species impaired MTT reduction activity in all single cell cultures already at high nanomolar concentrations. These findings are in good agreement with various other studies investigating toxic or activating Ab effects in hippocampal neurons [35], astrocytes [10] and microglia [36]. In contrast to our findings in the single cell cultures none of the Ab species affected cellular viability in OHCs, although we could confirm the presence of Ab in the slices by immunostaining and GFAP upregulation. In line with our observations other studies in OHCs also showed no or, at very high concentrations, only very limited toxic effects of Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), Ab  and Ab   [18,19,37,38]. In contrast to that and to our findings Lambert et al. published in 1998 that slice cultures could be injured with as little as 5 nM soluble Ab (1-42) of so called Ab derived diffusible ligands (ADDL) [39]. Later, Chong et al. described in 2006 neuronal cell death in hippocampal brain slices because of Ab (1-42) oligomer treatment [40]. The reason for the difference to our results could be the kind of Ab (1-42) preparation, as both groups used aggregation protocols which resulted in spheres of approximately similar size. However, their contrasting observations render it likely that their mode of preparation resulted in a different internal structure of the aggregates. Future studies should be carried out to extensively compare the different Ab species for their potentially different effects. Nevertheless, we observed comparable detrimental effects of all investigated Ab species on MTT reduction in single cell culture, which could not be seen in any complex tissue. That discrepancy between single cells and OHCs regarding the effect of Ab is difficult to reconcile. As single cell cultures are almost exclusively prepared from embryonic tissue and as OHCs represent juvenile tissue one explanation could be that the respective cells are in different physiological states. Scrutinizing this assumption we show that single cells obtained from juvenile OHCs are only susceptible to Ab effects after being cultured. Similarly, Yankner et al. (1990) reported that dissociated neurons maintained in cultures are resistant to Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) toxicity during the first days in culture and that Ab neurotoxicity increases with the age of the culture [41]. This may indicate that cultured cells and cells that are embedded in the intact hippocampal synaptic circuitry and anatomy differ regarding cell properties which are crucial for Ab toxicity or that the interaction between the neural elements in the relatively intact tissue enables a counteracting protective mechanism. Possible mechanisms may be alterations in the membrane lipid composition [42] or an altered accessibility of lipid rafts for Ab [43]. Similar reasons may account for the Ab effects in studies where OHCs were cultured for several weeks [44]. These findings do not reflect the situation in adult tissue as we and others [45] did not observe a fast toxic effect of Ab after in vivo application. Also consistent with our results Geula et al. (1998) did not observe a significant Ab toxicity in aged rats but found age-dependent Ab toxicity in aged monkeys [46]. This does not exclude that the hippocampal neurons in OHCs, acutely isolated slices and in vivo are physiologically impaired, as LTP was disturbed in the acutely isolated slice preparations at least after Ab oligomer application. Recent studies increasingly indicated that soluble, pre-fibrillar Ab assemblies rather than mature fibrils may induce early neuronal alterations, leading to physiological interruption before cell death is detectable [47]. Our LTP experiments elucidated the effects of distinct Ab species on synaptic potentiation. We show that Ab (1-40) oligomers disturbed LTP, whereas Ab (1-40) fibrils did not impair LTP, although Ab (1-40) fibrils where higher concentrated and permanently exposed to the slices. This is in good agreement with the current view that Ab oligomers are responsible for the early disturbance of brain physiology [48][49][50][51]. Whether or not LTP disturbances are a first sign of neuronal degeneration remains to be elucidated. If so, the MTT assay would evidently be unable to detect such early alterations in cellular physiology, as we demonstrated that Ab (1-40) oligomer mediated LTP disruption was not reflected by MTT reduction in slices. On the other hand, studies utilizing primary neuronal and astroglial cultures showed an inhibition of MTT reduction already 2 h after Ab application [10,52]. This may not necessarily reflect cell death, as Ab-induced alterations in MTT reduction in human cortical cultures could not be confirmed with other cytotoxicity assays like LDH and alamarBlue [53].
Taken together, we showed that single cell cultures are prone to impairment by Ab, whereas cells embedded in the intact hippocampal synaptic circuitry and anatomy are quite resistant, suggesting that results obtained with cell cultures cannot be conferred directly to complex tissue. In addition, we demonstrated that Ab mediated LTP disruption depends on the Ab species and does not correlate with MTT reduction in acute isolated slices, relativizing the MTT assay as a reporter of early physiological disruption and drug testing. Thus, Ab effects observed in single cell cultures should be interpreted cautiously regarding their relevance for more complex brain tissue, independently whether MTT reflects cellular viability or precedes cell death.

Single cell culture
The animals were maintained under constant environmental conditions, with an ambient temperature of 2162uC, a relative humidity of 40%, a 12-h light-dark cycle and free access to food and water. All animal procedures have been approved by the ethics committee of the German federal state of Sachsen-Anhalt, and are in accordance with the European Communities Council Directive (86/609/EEC).
Cells cultures from 1-day-old Wistar rats (Institute breeding stock) were prepared and cultured as described previously [56]. Briefly, newborn rats were decapitated, and the brains were removed and collected in ice-cold Hanks-buffer solution (Biochrom; Berlin, Germany). The brains were gently passed through nylon meshes of 250 mm and 136 mm pore width, in consecutive order. The cell suspension was centrifuged at 4uC for 5 min at 500g. The cells were resuspended in 10 ml growth medium (DMEM supplemented with 10% (v*v 21 ) fetal calf serum, 20 U*ml 21 penicillin and 20 mg*ml 21 streptomycin).
Single cells from OHCs were isolated by gently removing the slices from the membrane and collecting them in ice-cold Hanks-buffer solution (Biochrom; Berlin, Germany). Then the protocol for cell culture preparation described above was applied. Preparation and cultivation of OHCs was done as described below.
For astrocyte-enriched cultures (95% astrocytes), cells were seeded in 48 well plates at a starting density of 2*10 4 cells/ml in DMEM supplemented with 10% (v*v 21 ) fetal calf serum and incubated at 37uC in an atmosphere containing 5% CO 2 . The medium was changed every second day. For neuron-enriched culture (80% neurones), the DMEM was exchanged by Start V Medium (Biochrom) 24 h after seeding.

Ab application/MTT assay
Ab (25-35) (Bachem) was freshly dissolved in bidistilled water to a concentration of 1 mg*ml 21 . For fibril formation, recombinant Ab (1-40) [57] was dissolved in bidistilled water to a concentration of 1 mg*ml 21 and incubated for 5-7 days at 37uC. The formation of fibrils was verified by negative stain electron microscopy. Ab (1-42) oligomers were generated as described [26]. The quality of the oligomer preparation was controlled by negative stain electron microscopy and Sodiumdodecylsulfate-Polyacrylamidgelelectrophoresis (SDS-PAGE). The Ab species were added to the cell culture medium at a concentration of 0.5-10 mM (Ab (1-42) oligomers) or 0.5-20 mM (Ab (1-40) fibrils) and incubated for 1-3 days. Then MTT (Carl Roth) was added to the medium (0.5 mg*ml 21 ) and incubated for 3 hours. The medium was removed and the cells were diluted in 20% SDS/50% Dimethylformamid. The relative formazane concentration was measured by determination of the absorbance at 570 nm (well plate reader, Optima FluoStar).

Organotypic cultures
Organotypic hippocampal interface slice cultures from 10-day-old Wistar rats (Institute breeding stock) were prepared and cultured as interface slices as described previously [59]. Briefly, the slices were placed on membrane inserts in 6-well plates (NUNC, Wiesbaden, Germany) containing 1.2 ml of NB medium/well and were maintained in a humidified incubator for 12-15 days in vitro (DIV).

Immunhistochemistry
For the immunohistochemical staining of Ab and GFAP, the slices were fixed in 0.1 M phosphate buffer containing 4% paraformaldehyde. The slices were stored in the fixative overnight. After cryoprotection in 30% sucrose, the slices were rapidly frozen in methylbutane at 280uC. The whole slices were cut on a sliding microtome and the 20 mm sections were stored at 4uC in cryoprotectant (CPS) containing 25% ethylene glycol, 25% glycerine in 0.1 M phosphate-buffered saline (PBS). The slices were transferred from CPS to 0.1 M phosphate buffer and washed overnight. Unspecific bindings were blocked for 2 h in the corresponding serum and then the slices were incubated with the primary antibodies and stored at 4uC overnight. All secondary antibodies were incubated at room temperature for 2 h. The slices were then coverslipped with 1,3-diethyl-8-phenylxanthine (DPX).

Ab application/MTT assay/PI staining
The Ab species were added to the slice culture medium at the respective concentrations (1-10 mM) and incubated for 3-6 days. For the application ''on top of the slice'', 1 ml of the Ab stock solution was directly applicated onto the surface of the slice. 1 ml of the solvent was applicated onto the control slices. Then MTT was applied to the medium (0,5 mg*ml 21 ) and incubated for 3 hours. The slices were quickly removed from the membrane and completely diluted in 20% SDS/50% dimethylformamid (incubation for 24 h at RT). After centrifugation, the relative formazane concentration of the supernatant was measured by determination of the absorbance at 570 nm (well plate reader, Optima FluoStar).
Electron microscopy was done as previously described by [60]. Cell death was evaluated by cellular incorporation of propidium iodide (PI) 3d and 6d after Ab treatment. Cultures were incubated with PI-containing medium (10 mM) for 2 h at 33uC. Fluorescent images were acquired semiautomatized (Nikon motorized stage; LUCIA software) and analyzed by densitometry to quantify necrotic cell death (LUCIA Image analysis software).

Acute hippocampal slices/LTP
Hippocampal slices (400 mm thick) were prepared from 7-to 8week-old male Wistar rats (Institute breeding stock) as described previously [61]. Briefly, both hippocampi were isolated and transferred into a submerged-type recording chamber where they were allowed to recover for at least 1 h before the experiment started. The chamber was constantly perfused with artificial cerebrospinal fluid (ACSF) at a rate of 2.5 ml/min at 3361uC. Synaptic responses were elicited by stimulation of the Schaffer collateral-commissural fibers in the stratum radiatum of the CA1 region using lacquer-coated stainless steel stimulating electrodes. Glass electrodes (filled with ACSF, 1-4 MV) were placed in the apical dendritic layer to record field excitatory postsynaptic potentials (fEPSPs). The initial slope of the fEPSP was used as a measure of this potential. The stimulus strength of the test pulses was adjusted to 30% of the EPSP maximum. During baseline recording, single stimuli were applied every minute (0,0166 Hz) and were averaged every 5 min. Once a stable baseline had been established, long-term potentiation was induced by applying 100 pulses at an interval of 10 ms and a width of the single pulses of 0.2 ms (strong tetanus) three times at 10 min intervals.
Ab  oligomers and fibrils were prepared as described previously [27] and visualized by negative stain electron microscopy. Immediately after the slice preparation, fibrillar Ab (1-40) was persistently applied to the slices at a concentration of 1 mM. Ab (1-40) oligomers and Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) were applied to the slice for 30 min before tetanus application at a concentration of 500 nM. The Ab (1-40) solvent HFIP was removed from the ACSF by exposure to a gentle stream of carbogen for 1h. For control experiments we added the same amount of HFIP used for the Ab (1-40) experiment to the ACSF and removed it by gasification. There was no difference between the potentiation in the HFIP-deprived ACSF and pure ACSF and, therefore, these experiments were pooled. In parallel to the experiments, some slices of the same preparation were separately exposed to Ab for 3-4 hours and analyzed with MTT assay as described above.

In vivo infusion of Ab
In vivo infusion was performed as described previously [62]. Briefly, anaesthesia of 10-week-old male Wistar rats (Institute breeding stock) was induced with halothane in a mixture of nitrous oxide and oxygen (50:50) and maintained with 2-3% halothane (Sigma, Deisenhofen, Germany) via a rat anaesthetic mask (Stölting). The animals were placed in a Kopf stereotaxic frame. Following a midline incision, a burr hole (1 mm in diameter) was drilled into the skull (coordinates: posterior, 0.9 mm from bregma; lateral, 1.7 mm to satura sagittalis) and a 29-gauge cannula was lowered to 4.5 mm below the skull, according to the rat brain atlas of Paxinos and Watson [63]. Ab (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) (1 mg*ml 21 ) or Ab  oligomer (1 mg*ml 21 ) was injected intracerebroventricularly in 3ml sterile solvent over 5 min. After 5 min the cannula was slowly withdrawn. Ab  (1 mg*ml 21 ) was used as inactive peptide control. After three days, acute hippocampal slices were prepared as described above, then directly placed on cell culture membranes and the MTT reduction activity analyzed as described above.

Statistics
Values of LTP recording are given as mean6S.E.M. Values of MTT reduction are given as mean6S.D. As indicated in Results, the Mann-Whitney U-test or the analysis of variance (ANOVA) with repeated measures was used to compare the field potentials between two groups of differentially treated cells or slices, respectively (i.e., control vs. Ab treatment), where appropriate.