Address correspondence and reprint requests to M. Grønborg, NsGene, Pederstrupvej 93, DK-2750 Ballerup, Denmark. E-mail: email@example.com
NS 1231 [5-(4-chlorophenyl)-6,7,8,9-tetrahydro-1H-pyrrolo-[3.2-h]naphthalene-2,3-dione-3-oxime] belongs to a chemical series of compounds, which exhibit neurotrophic-like activities. In vitro, NS 1231 rescued nerve growth factor (NGF)- differentiated PC12 cells from death induced by withdrawal of trophic factors. In addition, NS 1231 stimulated NGF-induced neurite outgrowth of undifferentiated PC12 cells. At the molecular level, NS 1231 enhanced NGF-induced signalling events, such as TrkA phosphorylation at the Shc-binding site Tyr490 as well as ERK activation in PC12 cells. Moreover, NS 1231 reduced NMDA-induced excitotoxicity in organotypic hippocampal slice cultures. In a gerbil model of transient global ischaemia, treatment with NS 1231 reduced the delayed loss of neurons in the hippocampal CA1 layer. Furthermore, NS 1231 treatment resulted in a 43% reduction in total infarct volume in the mouse middle cerebral artery occlusion (MCAO) model. The present data thus implicate a therapeutic potential of NS 1231 or structural analogues in treatment of cerebral ischaemia.
Rescue of damaged neurons and stimulation of neurogenesis are theoretically attractive strategies for the treatment of neurodegenerative diseases. Several endogenous neurotrophic factors, including nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF) and glial cell line-derived neurotrophic factor (GDNF), have been identified. These factors are critical for development, differentiation as well as maintenance of distinct populations of neurons. However, their clinical use is limited by their inability to reach the brain after systemic administration. The therapeutic use of neurotrophic factors therefore necessitates intracranial injections, transplantation of cells secreting neurotrophic factors, or gene therapy. Such approaches have resulted in promising results in several animal models of neuronal degeneration (Emerich et al. 1994; Kordower et al. 2000; Schwabitz et al. 2000; Yagi et al. 2000). However, small molecules with the ability to activate or enhance neurotrophic signalling might provide an alternative therapeutic approach.
The neurotrophic factors exert their actions by binding to specific transmembrane receptors with intracellular tyrosine kinase domains. Ligand binding induces receptor dimerization and activation of the intrinsic tyrosine kinase, leading to phosphorylation of specific tyrosine residues located at the intracellular domain. These events result in the recruitment of a number of signalling molecules, leading to activation of pathways including kinases such as ERK, which is one of the important mediators of the neurotrophic response (Cowley et al. 1994; Xia et al. 1995).
We have previously described a molecule with neuroprotective effects, NS 521 (Grønborg et al. 1999). In the present report, we describe another small molecule, NS 1231, with a different chemical backbone structure. Like NS 521, NS 1231 rescues differentiated PC12 cells from death induced by withdrawal of neurotrophic support. In contrast to NS 521, NS 1231 preserves the neurites of surviving PC12 cells. Furthermore, NS 1231 increases neurotrophic factor-induced neurite outgrowth of undifferentiated PC12 cells. In accordance with this, NS 1231 enhances NGF-induced intracellular signalling events in PC12 cells. We also describe the neurotrophic effect of NS 1231 in a primary in vitro model of neuronal degeneration as well as in two different animal models of ischaemia.
Rat PC12 pheochromocytoma cells obtained from Brian B. Rudkin (Ecole Normale Superieure de Lyon, France) were cultured in growth medium consisting of Dulbecco's modified Eagle's medium (DMEM) supplemented with 7.5% fetal bovine serum and 7.5% donor horse serum in a humidified incubator at 37°C and 5% CO2. All cell culture reagents were from Invitrogen A/S (Tåstrup, Denmark).
PC12 survival assay
Estimation of the survival of differentiated PC12 cells, deprived of serum and NGF, was carried out as described previously (Grønborg et al. 1999) with minor modifications. In brief, cells were differentiated in the presence of serum and 2 nm 7S NGF (Alomone Laboratories Ltd, Jerusalem, Israel) for 6 days, and the medium was then replaced with serum-free DMEM supplemented with NS 1231 in the indicated concentrations. NGF (3 nm) was included as a positive control. The number of cells was evaluated after 4 days by measurement of the ability of the cells to reduce the tetrazolium compound MTS (Promega, Southampton, UK) according to instructions given by the manufacturer.
Stimulation of neurite outgrowth in PC12 cells
PC12 cells were seeded in collagen-coated tissue culture plates at a cell density of 1.5 × 104 per cm2 in growth medium. The next day, cells were incubated with the indicated concentrations of NS 1231 and NGF. Cells were fixed in 4% paraformaldehyde after 2 days and stained with Coomassie blue. Neurite length per cell was quantified using unbiased two-dimensional (2D) stereology as previously described (Rønn et al. 2000) using the CAST-grid system connected to an Olympus BH-2 microscope.
PC12 cells were seeded at a density of 7.5 × 104 per cm2 in collagen-coated plates in growth medium. The next day, cells were stimulated for 5 min with NS 1231 in the absence or presence of NGF in serum-free medium. After stimulation, cells were washed in PBS and harvested in heated (96°C) sample buffer [2% sodium dodecyl sulphate (SDS), 0.4 m Tris, pH 8.0, 10 mm dithiothreitol and 0.25 mm Na3VO4]. The cell lysates were electrophoresed on 8–18% gradient SDS gels, which were electroblotted to polyvinylidene difluoride (PVDF) membranes. Signalling molecules were detected using anti-phospho-p44/p42 MAP kinase E10 mAb (1 : 2000) or anti-phospho-TrkA (Tyr490) pAb (1 : 300) from New England BioLabs Ltd. (Hitchin, Hertfordshire, UK) followed by HRP-linked anti-mouse or anti-rabbit antibody. Bands were detected by chemoluminescence using the ECL or the ECL + system (Amersham Pharmacia Biotech, Hørsholm, Denmark). Membranes were stripped and incubated with anti-ERK-1 clone MK12, Transduction Laboratories (1 : 2000) or anti-TrkA, New England BioLabs (1 : 300) to assure that equal amounts of protein were loaded.
NMDA exposure on hippocampal slice cultures
Rat hippocampal slice cultures were prepared and grown on semiporous membranes as previously described (Noraberg et al. 1999). Briefly, 7-day-old rats were decapitated and the dorsal hippocampi isolated. The hippocampi were cut into 300–400 µm thick, transverse slices, and the slices transferred to Gey's balanced salt solution for separation and trimming of the tissue slices. Tissue slices were transferred to membrane inserts and placed in six-well culture trays in culture medium consisting of 25% horse serum, 25% Hank's balanced salt solution (HBSS), 1% of 50% d-glucose and 49% OPTI-MEM. After 3 days, the medium was changed to serum-free neurobasal medium with 2% B27 supplement and 1 mm l-glutamine. Slice cultures, grown for 3 weeks, were exposed to 10 µm NMDA for 48 h (Kristensen et al. 2001). One hour before exposure to NMDA, the regular culture medium was changed to serum-free medium containing NS 1231 (0.3–10 µm). Neuronal degeneration was monitored by densitometric measurements of the cellular uptake of propidium iodide (PI), as previously described (Noraberg et al. 1999). In brief, slice cultures were exposed to 2 µm PI at least 3 h before exposure to 10 µm NMDA. The PI uptake was recorded by a digital camera before the addition of NS 1231 and after 24 and 48 h NMDA exposure and was quantified by densitometric analysis, using NIH Image software (version 1.62). After the last recording of PI uptake, the cultures were fixed in 4% paraformaldehyde. One-half of the cultures were stained with toluidine blue (Nissl) without sectioning. The other half was transferred to 20% sucrose solution for 24 h, frozen, and sectioned at 20 µm (three series) on a cryostat. Two series of culture sections were immunostained for the neuronal markers MAP2 (Noraberg et al. 1999) and NeuN (Blaabjerg et al. 2001).
Transient global ischaemia in gerbils
Gerbils were anaesthetized with halothane, and the right and left carotid arteries were located and occluded for 4 min. Animals were kept at normal body temperature before and after the operation using heating lamps. During the operation the gerbils were placed on heating pads, and the body temperature was controlled and maintained at 37 ± 0.5°C. NS 1231 (10 mg/kg) was administrated intraperitoneally (i.p.) 2 min post occlusion insult and again the following day. Four days later, the animals were killed, and the brains were removed and cooled to − 70°C. The brains were then sectioned in 20-µm thick slices of which 5–7 with hippocampal tissue were selected and stained with haematoxylin–eosin (HE). The degree of hippocampal damage was categorized into one of four groups: group 1, no damage in the CA1-layer; group 2, the CA1-layer partly damaged; group 3, the CA1-layer completely damaged; and group 4, damage in more than just the CA1-layer (Jensen and Møller 1992). The total ischaemia score was obtained as the sum of scores in the right and left hemispheres, and Kendall's tau test was used for statistic evaluation.
Mouse middle cerebral artery occlusion (MCAO)
Female NMRI-mice 27–38 g were anaesthetized with halothane (2% halothane in 30% O2−70% NO2). During the operation the body temperature were maintained at 37 ± 0.5°C by placing the mice on heating pads connected to a CMA/150 temperature controller, and further maintained at this temperature for approximately 12 h after the operation by placing the animals under heating lamps. MCAO was performed as earlier described (Møller et al. 1995). NS 1231 (15 mg/kg) was administrated i.p. 2 min after MCAO. The control group was given vehicle (20% Tween-80). On the fourth day, the mice were killed and the brains removed. The brains were frozen on dry ice and cut into 20 µm sections. With random start, every 40th sections were sampled and stained with HE, thus the distance between the tissues analyzed at was 800 µm (t). The infarct volume was estimated using the Cavalieri's volume estimator (Gundersen et al. 1988). In short, the sections were placed in an Olympus CAST-grid microscope system, with which a point grid can be generated on the screen. The total number of points hitting the infarcted volume (ΣP) was counted. Each point represent an area (apoint), and the total volume of the infarct was calculated using the formula:
This volume estimator gives the unbiased estimate of the infarcted volume within few minutes, with a precision of the estimator (CE) around 5% in the above set-up (Gundersen et al. 1988).
Survival of PC12 cells after growth factor deprivation
NS 1231 (Fig. 1) increases survival of NGF-differentiated PC12 cells, deprived of trophic support. After 4 days incubation in serum-free medium, viability was determined by estimating the ability of the cells to reduce the tetrazolium compound MTS. Addition of NS 1231 at the time of serum and NGF withdrawal prevented cell loss in a dose-dependent way (Fig. 2). Significant survival effects of NS 1231 were observed at concentrations between 1 and 10 µm. In addition, NS 1231 preserved the neurites of the surviving cells (data not shown).
NS 1231 potentiates NGF-induced neurite outgrowth and intracellular signalling
The effect of NS 1231 on neurite outgrowth in undifferentiated PC12 cells was assessed. Cells were incubated in full growth medium with NS 1231 (1 or 3 µm) in the absence or presence of NGF. After 48 h cells were fixed, and the mean neurite length was estimated using unbiased 2D stereology. NS 1231 significantly potentiated neurite outgrowth induced by NGF (Fig. 3). In order to investigate whether this was correlated with intracellular signalling events, the effect of NS 1231 (1 or 3 µm) in the absence or presence of NGF was assessed in undifferentiated PC12 cells. NS 1231 had no significant effect alone on TrkA autophosphorylation at Tyr490 but significantly potentiated NGF-stimulated TrkA activation (Fig. 4a). In accordance with this result, NS 1231 also potentiated NGF-induced activation of ERK (Fig. 4b).
In accordance with previous studies (Kristensen et al. 2000), cultures exposed to 10 µm NMDA displayed a fairly selective CA1 lesion after NMDA exposure, as illustrated by PI uptake predominantly in this region (Fig. 5). The basic PI uptake before exposure to NMDA was low. NS 1231 in a concentration of 1 µm significantly reduced the PI uptake after 1 and 2 days of exposure to NMDA (Fig. 6). In agreement with this, histological investigations showed NMDA-induced neurodegenerative changes in CA1, which were significantly reduced in the presence of 1 µm NS 1231 (Fig. 6).
NS 1231 reduces hippocampal damage after global ischaemia
To determine the neuroprotective potential in vivo, NS 1231 was tested in a gerbil model of global ischaemia. Animals received vehicle or NS 1231 (10 mg/kg) 2 min after 4 min of transient global ischaemia, and again the following day. All animals included in the study survived the ischaemic insult and treatment with NS1231. All of the control animals showed total damage in the hippocampal CA1-layer (Fig. 7). In contrast, eight of 13 animals receiving NS 1231 after the ischaemic insult had no or only partial damage in the CA1 neurons of the hippocampus. This neuroprotective effect was significant (p = 0.01) using the non-parametric Kendall's tau test. No effect on body temperature was seen after administration of NS 1231 (data not shown).
NS 1231 reduces infarct volume after MCAO in mice
In the mouse MCAO model, NS 1231 administered i.p. (15 mg/kg) at 2 min post-ischaemia resulted in a 43% reduction of total infarct volume (Fig. 8). All animals included in the study survived the ischaemic insult and treatment with NS1231. In both the control group and the NS 1231-treated group, the infarct volumes were smaller than in earlier trials.
In the present study, we show evidence that the compound NS 1231 has neuroprotective activity in different neuronal culture systems and in two in vivo models of neuronal degeneration. We have previously described a molecule with neuroprotective effects, NS 521 (Grønborg et al. 1999). Like NS 521, NS 1231 was identified in an in vitro screening program based on differentiated PC12 cells deprived of trophic support. After several days in the presence of NGF, PC12 cells acquire numerous properties of mature sympathetic neurons, including outgrowth of neurites (Greene and Tischler 1976). Withdrawal of trophic support induces loss of cell viability, and this was inhibited by addition of NS 1231. Importantly, NS 1231 also preserved the neurites of the surviving cells. In contrast, cells rescued by NS 521 loose their neurites (Grønborg et al. 1999). Addition of NS 1231 to undifferentiated PC12 cells induced the outgrowth of short neurite-like spikes and strongly potentiated neurite outgrowth induced by NGF. In accordance with this, NS 1231 enhanced NGF-induced autophosphorylation of TrkA at the Shc binding site (Tyr490) as well as activation of ERK in undifferentiated PC12 cells. NS 1231 alone had no detectable effect on activation of these molecules. Increased survival of NGF-differentiated PC12 cells after withdrawal of NGF in the presence of NS1231 may be explained by the presence of trace amounts of NGF after the change to serum-free medium.
It is likely that NS 1231 activates other intracellular pathways than ERK activation mediated through phosphorylation of TrkA at Tyr490. This could involve phosphorylation of other sites in TrkA for example the PLCγ binding site (Tyr785). The activated PLCγ cleaves phosphatidylinositol-4,5-bisphosphate, yielding diacylglycerol and inositol trisphosphate. These two second messengers activate protein kinase C and trigger a rise in intracellular calcium, respectively, and may contribute to the neurotrophic-like effect of NS 1231 without directly enhancing ERK activity. It is also possible that NS1231 activates pathways that do not require TrkA activation and may act both independently and synergistically with NGF. A structural analogue of NS 1231 enhances intracellular signalling induced by NGF as well as bFGF and EGF in PC12 cells (Dagøet al. 2001). However, if also NS1231 is able to potentiate signalling of other neurotrophic factors than NGF remains to be established.
NS 1231 was also tested in an in vitro model of excitotoxicity. In this model a significant reduction of NMDA-induced PI uptake by 1 µm NS 1231 in organotypic hippocampal slice cultures was observed in three independent experiments. Though there was no effect of NS 1231 at higher concentrations, this might be explained by decreased solubility of NS 1231 in these particular medium conditions. Alternatively, NS 1231 is toxic to the cultures at higher concentrations. This has been shown to be the case for high concentrations of neurotrophic factors in these cultures (Meyer and Widmer, personal communication). Another reason for the discrepancy between rescuing doses of NS 1231 in PC12 cells and hippocampal slice cultures could be differences in expression of receptors or other molecules involved in cell survival. Furthermore, hippocampal slice cultures contain several cell types, which might interact with each other in a more complex way than is the case in a PC12 cell culture.
Addition of neurotrophic factors has previously been shown to protect against excitotoxic insults in these cultures (Pringle et al. 1996). The protective effect of NS 1231 in hippocampal slice cultures could be due to potentiation of neurotrophic factor signalling, as astrocytes in the cultures might secrete low amounts of neurotrophic factors.
Little is known about the exact mechanisms underlying the protective properties of the trophic factors in this model. One possible explanation is an effect on the expression of glutamate receptors. BDNF and bFGF has been shown to down-regulate NMDA receptor function in cerebellar granule cells by altering the expression of NMDA receptor subunits (Brandoli et al. 1998).
In accordance with the in vitro data, we also present data showing a significant neuroprotective effect of NS 1231 in a gerbil model of global ischaemia and a mouse MCAO model. In gerbils, the damage after transient global ischaemia in the hippocampal CA1 layer was absent or reduced in a significant fraction of the NS 1231-treated group. Likewise, NS 1231 treatment decreased infarct volumes significantly in NMRI mice after focal ischaemia as compared with untreated controls although the infarct volume was smaller in the control group than observed in previous in-house trials. The reason for this discrepancy is not clear but strain variance may play a role. It is unlikely that the neuroprotective effect of NS 1231 observed in any of the models is caused by an effect on brain temperature. No effect on body temperature was seen after administration of NS1231 in gerbils and the body temperature was maintained constant during the operation. In the MCAO model, the body temperature was maintained constant during and after the operation.
The CA1 pyramidal neurons in the hippocampus have been shown to undergo selective and delayed cell death in experimental animal models, as well as in humans, after transient cerebral ischaemia. The expression of neurotrophic factors as BDNF, bFGF, CNTF and GDNF in the hippocampus and cerebral cortex has been shown to be up-regulated after ischaemic insults (Lin et al. 1997; Tsukahara et al. 1998; Park et al. 2000; Wei et al. 2000; Yanamoto et al. 2000). The elevated expression may play an important role in protection of ischaemic injured neuronal cells. In accordance with this hypothesis, BDNF and TrkB are expressed in about 95% of CA1 neurons surviving ischaemic insults (Ferrer et al. 1997). Tyrosine phosphorylation of Trk has been shown to be elevated in dentate granule cells after transient forebrain ischaemia, but to a lesser extent in the vulnerable CA1 region (Hu et al. 2000). Furthermore, increasing the levels of neurotrophic factors as BDNF, NGF, EGF or GDNF by exogenous delivery has been shown to protect against neuronal damage in the hippocampus after ischaemic insults in animal models (Beck et al. 1994; Peng et al. 1998; Schabitz et al. 2000; Yagi et al. 2000). Though these ligands activate different tyrosine kinase receptors, they all induce activation of the ERK pathway. The ERK signalling pathway has been suggested to be neuroprotective for dentate granule cells after transient forebrain ischaemia, as ERK activation was detected in the nuclei and dendrites of surviving dentate gyrus cells, but not in dying CA1 neurons (Hu et al. 2000). Thus it is possible that NS 1231 protects the neurons by enhancing the intracellular signalling of endogenously produced neurotrophic factors. However, at present it is not completely clear whether the effect of NS 1231 is dependent on low concentrations of neurotrophic factors, or the compound also activates independent signalling, different from the ERK pathway, with survival promoting effects. Further studies to clarify the molecular mechanism of action are in progress.
We conclude, that NS 1231 is a potential anti-ischaemic drug, as indicated by the ability of the compound to reduce hippocampal damage in a gerbil model of global ischaemia and infarct volume in the mouse MCAO model. If NS 1231 enhances intracellular signalling induced by neurotrophic factors, it may also be a potential drug for treatment of other neurodegenerative diseases.
We thank Brian B. Rudkin for providing the PC12 cells.