Address correspondence and reprint requests to David Schubert, The Salk Institute, Laboratories for Cellular Neurobiology, 10010 North Torrey Pines Road, La Jolla, CA 92037-1099, USA. E-mail: Schubert@salk.edu
The plant polyphenolic curcumin alters the response of nerve cells to some forms of toxic stress. The steroid-like compound, cyclohexyl bisphenol A, has broad neuroprotective properties that are very distinct from those of curcumin. To incorporate both families of biological activities into a single molecule, a pyrazole derivative of curcumin, called CNB-001, was synthesized. CNB-001 acquires a new activity and is far superior in neuroprotection assays to either parental molecule, but retains some of the properties of both. It is neuroprotective in cell culture assays for trophic factor withdrawal, oxidative stress, excitotoxicity, and glucose starvation, as well as toxicity from both intracellular and extracellular amyloid. While the creation of CNB-001 was based upon an uncommon approach to drug design, it has the potential of a lead drug candidate for treating multiple conditions involving nerve cell death.
Neurodegenerative diseases, ischemia, and trauma are among the major causes of morbidity and death, yet there are no reliable methods of preventing nerve cell loss in these conditions. Although there are innumerable insults that lead to nerve cell death, ranging from aberrant protein aggregation in the amyloid diseases (Lansbury and Lashuel 2006) to the loss of neurotrophic support (Pezet and Malcangio 2004), it is very likely that there is a much smaller subset of mechanisms responsible for the ultimate demise of the cell. If this assumption is valid, then it should be possible to identify drugs that block the common pathways leading to nerve cell death. A good starting point for the identification of such lead compounds is the enormous family of natural products which form the basic scaffolds for the majority of our most widely used drugs (Harvey 1999; Paterson and Anderson 2005; Corson and Crews 2007).
During an investigation on the role of steroids in the cellular response to Alzheimer’s amyloid-beta peptide (Aβ) (Liu and Schubert 1998; Liu et al. 1998), we found that a steroid related compound, cyclohexyl bisphenol A (CBA), had a very unique property in that it not only protected nerve cells from Aβ toxicity, but also from glutamate toxicity and the loss of trophic factor support (Schubert and Liu 2002). At about the same time, it was reported that curcumin, a polyphenolic compound found in the Indian spice turmeric, was protective in a transgenic animal model for Alzheimer’s disease (AD) (Lim et al. 2001). However, unlike CBA, curcumin does not prevent the nerve cell death associated with the loss of trophic support, and the EC50 of both of these compounds in the cell culture assays where they are effective is high, between 1 and 10 μM. Therefore, we sought to improve the effectiveness of both CBA and curcumin by synthesizing a hybrid molecule that is active in a wide range of different assays for neuroprotection. This approach of drug selection based upon biological efficacy is quite distinct from the current more widely employed drug development strategy of first identifying a molecular target and then developing the drug (Pangalos et al. 2007). In this study, we describe the broad biological activity of a CBA-curcumin hybrid molecule called CNB-001.
Materials and methods
High glucose Dulbecco’s modified Eagle’s medium (DMEM) and fetal calf serum were from Invitrogen (Carlsbad, CA, USA). All other reagents were purchased from Sigma (St Louis, MO, USA). The MC65 cell line was from Dr L.-W. Jin (U. C. Davis). The transmembrane receptor kinase B (TrkB)-expressing HT22 cells were obtained from Dr Gerard Thiel (Hamburg, Germany).
Synthesis of CNB-001
CNB-001 was originally synthesized in our laboratory, but larger quantities have recently been made by Craig McNair at High Force Ltd (Bowburn, England). Curcumin (20 g, 0.054 mol) was slurried in toluene (600 mL), and phenyl hydrazine (10.8 mL, 2 equivalents), and trifluoroacetic acid (1 mL) added. The slurry was refluxed for 24 h and the water formed was removed. The reaction was complete as determined by TLC after this time. The mixture was cooled to ambient temperature and the toluene phase washed successively with water (500 mL), saturated sodium bicarbonate solution (300 mL), ammonium chloride solution (300 mL), and then saturated sodium chloride solution (300 mL). The organic phase was dried over MgSO4, filtered, and the solvent removed under reduced pressure to give an orange/yellow solid (24 g). The crude material contained several spots by TLC (EtOAc). CNB-001 was purified by silica gel chromatography, eluting with 10% ethyl acetate in heptanes going to 70% ethyl acetate in heptanes. The appropriate fractions were combined and the solvent removed under vacuum. The yield was 16.2 g, 68%. 1H NMR was recorded on a Bruker 270 MHz and chemical shifts are reported downfield from tetramethylsilane. 1H NMR (CDCl3): δ 3.9 (s, 3H, −OMe), δ 3.8 (s, 3H, −OMe), δ 5.7 (d, 2H, −OH), δ 7.5 (m, 5H, Ph), δ 6.6–7.1 (m, 11H).
All animal protocols were approved by The Salk Institute Animal Use Committee. CNB-001 was emulsified in 2.5% carboxymethyl cellulose at a concentration of 20 mg/mL and administered by gavage at a dosage of 400 mg/kg body weight. The mice were killed at 0, 1, 2, 4, and 6 h after administration. Plasma was obtained from blood, mixed with K3EDTA to prevent coagulation, centrifuged at 4300 g for 10 min, and extracted twice with ethyl acetate/propanol (9 : 1, v/v). The extracts were centrifuged at 5000 g for 10 min and the organic layer containing CNB-001 was again centrifuged at 20 000 g for 10 min. The extraction recovery from plasma is ∼90%; 30 μL was injected into an HPLC system equipped with a C18 reversed phase column and CNB-001 was detected at 330 nm. The elution solvent system was 50% acetonitrile, 50% water, and 1 g/L trifluoroacetic acid with a flow rate of 1 mL/min.
To assay the distribution of CNB-001 to the brain, mice were anesthetized with chloral hydrate and perfused with phosphate-buffered saline (PBS) to remove blood in the brain. The mice were then decapitated, the brains removed, the meninges removed, and the brains quickly frozen and stored at −80°C before further analysis. To measure CNB-001 levels in the brain, weighed pieces of cortex were homogenized by sonication in three volumes of PBS. The homogenates were then extracted and measured by HPLC as described above. With the fluids of the ventricles and the meninges removed, this procedure is widely used and gives an approximation of drug distribution to the CNS (Keller et al. 2002). The data are presented as micromolar compound based upon brain weight and blood volume.
Trophic factor withdrawal
Primary cortical neurons were prepared from 18-day-old rat embryos according to published procedures (Behl et al. 1994) and cultured at a low cell density of 1 × 106/35 mm dish, in DMEM/F12 (2 mL) containing N2 supplements (Invitrogen) and the compounds to be assayed. Viability was assayed 2 days later using a fluorescent live-dead assay (Molecular Probes, Eugene, OR, USA) and the data presented as the percentage of input cells surviving.
Serum starvation assay
HT22 cells with or without the TrkB brain-derived neurotrophic factor (BDNF) receptor (Rossler et al. 2004) were washed three times in serum-free medium, incubated in serum-free DMEM with or without the indicated compounds for 2 days and cell viability was determined by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay and compared with that of cells grown in 1% serum, which inhibited cell division but maintained viability.
HT22 cells from semi-confluent cultures were dissociated with 0.25% trypsin in Ca2+, Mg2+-free DMEM for 5 min at 37°C and plated at 2 × 103 cells per well in 96-well tissue culture dishes in DMEM plus 10% fetal calf serum. The following day the test compounds were added in triplicate at appropriate concentrations. Thirty minutes after compound addition, 5 mM glutamate was added to initiate the cell death cascade. Twenty hours later, the MTT assay was performed. For the MTT assay, 10 μL of 2.5 mg/mL MTT solution was added and incubated at 37°C for 4 h and then 100 μL of solubilization solution [50% dimethylformamide and 20% sodium dodecyl sulfate (SDS), pH 4.8] was added. The next day, the absorption values at 570 nm were measured. The results are presented as the percentage of the controls with vehicle alone.
Primary cultures of cortical neurons that die reproducibly by excitotoxicity were prepared by combining aspects of two published protocols as described in (Schubert and Piasecki 2001). Briefly, BALB/c mouse embryo cortices were minced and treated with 0.1% trypsin for 20 min. After centrifugation, the cells were resuspended in B27 Neurobasal medium (Invitrogen) plus 10% fetal calf serum and were dissociated by repeated pipetting through a 1 mL blue Eppendorf pipette tip. Then the cells were plated at 1 × 105 cells per well in 96-well poly-l-lysine and laminin-coated microtiter plates (Becton Dickinson, Bedford, MA, USA) in B27 Neurobasal plus 10% fetal calf serum and 20% glial growth-conditioned medium. Two days later the medium was aspirated and replaced by serum-free B27 Neurobasal medium plus 10 μg/mL cytosine arabinoside. The cultures were used without media change 11 days after plating and were essentially free of astrocytes. They were exposed to 10 μM glutamate followed by varying concentrations of the compounds. Cell viability was determined 24 h later using the fluorescent live/dead assay. Glial conditioned medium was prepared from confluent rat astrocyte cultures (Schubert and Piasecki 2001). The cells were washed twice with serum-free medium and incubated for 2 days in serum-free Neurobasal medium to produce the growth conditioned medium.
Glucose starvation assay
PC12 cells were washed three times in serum-free, glucose-free medium, and exposed to glucose-free medium plus or minus the indicated compounds and cell viability determined 48 h later using the MTT assay. The data are presented as viable cell number relative to serum-free, glucose-containing medium.
Determination of the Trolox equivalent activity concentration
Trolox equivalent activity concentration values for the compounds were determined according to (Rice-Evans and Miller 1994) but modified for a plate reader. Briefly, 250 μL of 2,2′-azinobis(3-ethylbenzothiazoline 6-sulfonate) was treated overnight with potassium persulfate to produce a radical cation, diluted to an optical density of ∼0.7 at 734 nm and added to 2.5 μL of a solution containing the compounds in ethanol. The change in absorbance because of the reduction of the radical cation was measured at 734 nm for 4 min. To calculate the TEAC, the gradient of the plot of the percentage inhibition of absorbance versus concentration for the test compound was divided by the gradient of the plot for Trolox.
Aβ toxicity to hippocampal neurons
Primary cultures of embryonic day 17 rat hippocampal neurons were used to assay a compound’s ability to directly inhibit Aβ toxicity. The hippocampi were removed, the cells dissociated with 0.25% trypsin in serum-free DME/F12 at 37°C for 20 min, and the trypsin inactivated by the addition of fetal calf serum to 10%. Following centrifugation, the cells were resuspended in DMEM/F12 and a single cell suspension made by repeated pipetting through a 1 mL Eppendorf tip. The cells were plated at 2 × 106 per 35 mm poly-lysine-coated tissue culture dishes in 10% fetal calf serum, and 2 days later 10 μg/mL Ara C was added to inhibit glial proliferation. At day 7, 10 μM Aβ1–42 was added with or without the test compound and viability determined 2 days later by the lactate dehydrogenase release assay (Behl et al. 1994).
Intracellular amyloid toxicity
The induction of intracellular amyloid toxicity in MC65 cells was performed exactly as described (Maezawa et al. 2006). MC65 is a human neuroblastoma cell line that conditionally expresses the C-terminal 99 amino acids (C99) fragment of the human amyloid precursor protein (APP) under the control of a tetracycline promoter (Sopher et al. 1994). In the absence of tetracycline, MC65 cells accumulate Aβ1–40 and Aβ1–40 aggregates, and ultimately die within 4–5 days following C99 induction. Briefly, cells from confluent cultures were dissociated and plated at 4 × 105 per 35 mm tissue culture dish in the presence (no induction) or absence (APP-C99 induced) of 1 μg/mL tetracycline in the presence or absence of the indicated compounds. At day 4, the control cells in the absence of tetracycline were dead, and cell viability was determined by the MTT assay. The data are presented as viability relative to viable controls plus tetracycline. On days 1, 2, and 3, cells were lysed in SDS sample buffer, run on 10–20% SDS–acrylamide gels, and immunoblotted with a monoclonal antibody against Aβ, 6E10 (Covance, San Francisco, CA, USA).
Thioflavin T assay
The assay for Aβ1–42 aggregation was performed according to Reinke and Gestwicki (2007). The compounds were incubated at the indicated concentrations with 25 μM Aβ1–42 (monomer, initially dissolved in hexafluoroisopropanol) in PBS in a final volume of 10 μL. At 48 h, 200 μL of 5 μM thioflavin T in 50 mM glycine, pH 8, was added and the fluorescence measured on a Spectra Max M5 (Molecular Devices, Sunnyvale, CA, USA; ex 446 and 490 nm). The data are presented as percentage of Aβ1–42 control (aggregated, no additive).
All of the experiments presented were repeated in triplicate at least four times with similar results. The data are presented as the mean ± SD of triplicate determinations.
Because of the innate complexity of the CNS, essentially all diseases and other insults to the CNS are multifactorial in the sense that there are a large number of mechanisms that can be identified as a cause of nerve cell death. Many, if not most, of these mechanisms can be reproduced in cell culture assays and compounds can be identified that inhibit individual mechanisms. However, in the real world cells are subjected to many insults simultaneously, necessitating either the combined use of different drugs or the identification of a drug that blocks multiple toxic insults. Another possibility is that cell death pathways are shared between different insults and are subject to inhibition by a single pharmacological intervention. For the CNS, these insults can be divided into at least six categories, including (i) oxidative stress (Maher 2006), (ii) glucose starvation (Warren and Frier 2005), (iii) loss of trophic factor support (Pezet and Malcangio 2004), (iv) excitotoxicity (Rothman 1985) and, in the case of the amyloid diseases, (v) extracellular amyloid toxicity (Lansbury and Lashuel 2006), and (vi) intracellular amyloid toxicity (LaFerla et al. 2007). To identify a more universally neuroprotective compound, we have employed cell culture assays for all six toxic insults, and selected a candidate drug that works in all six.
In an initial attempt to create a molecule with these broad biological properties, two molecules, CBA and curcumin, were used as the starting material because between them they are active in all six assays and curcumin has the ability to bind to and dissociate amyloid (Cole et al. 2007). To this end a molecule with the apparent structural components of both CBA and curcumin, a pyrazole derivative of curcumin, CNB-001 (Fig. 1), was synthesized and tested in the different neuroprotection assays.
The neuroprotective properties of CNB-001 are shown in Figs 2–4, and each assay is described individually, starting with the trophic factor withdrawal (TFW) assay. When embryonic day 18 rat cortical cells are dissociated and plated into a minimal medium at 5 × 105 cells per mL, the vast majority of the cells die within 24 h. In contrast, when cells are plated at a twofold higher density, the majority of these cells survive. The cells at low density can be rescued by the growth-conditioned medium of high density cultures. On the basis of these observations as well as the observation that the low-density cells can also be rescued by combinations of exogenous protein growth factors, the assay is considered to be a method to measure the ability of a compound to provide trophic support in the absence of classical trophic factors. It is called a TFW assay (Abe et al. 1990). Figure 2a and b show the ability of CNB-001 to maintain cell viability in this assay relative to vehicle control. The data for CNB-001 are quantified in Fig. 2c along with the results for the parental compounds CBA and curcumin. These data suggest that CNB-001 has neurotrophic factor-like activity that is several fold better than CBA.
Brain-derived neurotrophic factor plays an important role in the maintenance of nerve cell survival following CNS injury and ischemia. For example, CNS cells that make less BDNF recover more slowly from injury (Chen et al. 2005), while increasing BDNF increases nerve cell survival (Kurozumi et al. 2005),(Berger et al. 2004). However, as administered BDNF has poor bioavailability because of its inability to cross the blood–brain barrier, small organic molecules that mimic the effect of BDNF are a viable but as yet unfulfilled target for drug development (Pezet and Malcangio 2004). As CNB-001 is active in the TFW assay, we explored the possibility that it can mimic the effect of BDNF in another assay of cell survival. To do this, we examined derivatives of HT22 cells transfected with either the full-length BDNF receptor, TrkB, or an inactive, C-terminally truncated TrkB (Rossler et al. 2004). When HT22 cells expressing TrkB or truncated TrkB are starved for serum they die within 2 days (Rossler et al. 2004). However, in the presence of BDNF, only the cells expressing full-length TrkB survive. Figure 2d shows that using a maximally effective concentration of each compound based on the HT22 oxidative stress assay (Fig. 3a), both CNB-001 and CBA are able to mimic the effect of BDNF and promote the survival of the cells. CNB-001 is again more active than CBA, and both work on cells expressing either full-length TrkB or the truncated inactive receptor, indicating that their effects are receptor-independent. Curcumin is inactive in this assay. These data further demonstrate that CNB-001 has a neurotrophic-like activity.
Many models of programmed cell death have been established to characterize the oxidative mechanisms underlying neuronal degeneration. One of the most robust of these models is the glutamate-induced programmed cell death of immature cortical neurons and of hippocampal HT22 nerve cells (Davis and Maher 1994; Tan et al. 2001). In this assay, elevated levels of extracellular glutamate interfere with cystine uptake through the cystine/glutamate antiporter, which normally carries cystine into cells at the expense of the outflow of glutamate. The decreased cystine uptake leads to the depletion of intracellular GSH, a cysteine-containing tripeptide vital for cell survival because of its ability to act as an enzymatic cofactor and antioxidant. This form of cell death, called oxytosis (Tan et al. 2001), is distinct from excitotoxicity but is important to the understanding of neuronal degeneration in at least two situations. Many neurons contain no ionotropic glutamate receptors, but are killed by excess glutamate in trauma and ischemia, and neurons depleted of GSH by unknown mechanisms are selectively lost in Parkinson’s disease (for review, see Maher 2006). Figure 3a shows that CNB-001 is highly neuroprotective in this assay with an EC50 of 600 nM, while CBA and curcumin are much less active, with EC50S of 8 and 9 μM, respectively.
In addition to inhibiting cystine uptake, the pathological consequences of extracellular glutamate can be mediated by ionotropic glutamate receptors. Excessive extracellular glutamate leads to nerve cell death via the activation of NMDA receptors (Rothman and Olney 1986). This phenomenon, which can be reproduced in cell culture (Rothman 1985), is termed excitotoxicity (Olney 1986). Figure 3b shows that while CNB-001 is able to partially inhibit the excitotoxic response of primary cortical neurons to 10 μM glutamate, neither curcumin nor CBA are effective.
Glucose is the brain’s major energy source and the loss of glucose is one of the causes of nerve cell death in a variety of CNS pathologies, as well as in hypoglycemia associated with diabetes (Warren and Frier 2005). Therefore, drugs that maintain viability under conditions of reduced glucose availability would be of therapeutic value. To assay for neuroprotection from this condition, PC12 cells were washed three times in glucose-free medium and replated in the absence or presence of glucose, and cell viability assayed 48 h later. Figure 3c shows that 2 μM CNB-001, 10 μM CBA, and 50 ng/mL nerve growth factor were able to maintain cell viability with glucose starvation, while 10 μM curcumin was inactive.
The three compounds under discussion all have aromatic hydroxyl groups, and it could be argued that they are neuroprotective because they act as antioxidants by virtue of their ability to scavenge peroxyl-radicals and other reactive oxygen species. To determine if their neuroprotective potency correlated with their antioxidant potential, the TEAC assay was used; their antioxidant potential was compared with Trolox, a water soluble derivative of vitamin E (Rice-Evans et al. 1996). Figure 3d shows that both CNB-001 and CBA have less than 10% of the antioxidant activity of curcumin, which is equivalent to Trolox. Therefore antioxidant activity per se does not explain the neuroprotective action of these compounds.
Amyloid-beta peptide is thought to be one of the toxic entities in AD, and exogenous Aβ can be toxic to hippocampal neurons (Yankner et al. 1990). In addition, the intracellular accumulation of Aβ and C-terminal APP fragments are also toxic (McPhie et al. 1997; LaFerla et al. 2007). To determine if CNB-001 is able to inhibit extracellular Aβ toxicity, hippocampal neurons were treated with 10 μM Aβ1–42 in the presence of increasing amounts of CNB-001, CBA, and curcumin, and cell viability determined 2 days later (Fig. 4a). While CNB-001 had an EC50 of 400 nM in this assay, those of CBA and curcumin were 1 and 5 μM, respectively.
To assay the accumulation and toxicity of intracellular Aβ and Aβ-containing amyloid fragments, a human neuroblastoma cell line that conditionally expresses the C99 derived from the β-secretase cleavage of APP was used. This cell line, called MC65 (Sopher et al. 1994) has recently been used to identify drugs that inhibit amyloid aggregation (Maezawa et al. 2006). The cells are routinely grown in the presence of tetracycline and, following its removal, the expression of C99 is induced and the cells die within 4 days because of the accumulation of intracellular, toxic protein aggregates. Cell death is not caused by the secretion of Aβ and extracellular toxicity, nor by secreted toxins (Sopher et al. 1994). Death is inhibited by γ secretase inhibitors (Maezawa et al. 2006). Compounds that block intracellular Aβ polymerization inhibit cell death (Maezawa et al. 2006), as do some antioxidants (Woltjer et al. 2005). Cell death can be easily measured with any standard viability assay because there is complete cell lysis.
Figure 4b shows that CNB-001 inhibits amyloid-induced cell death in the MC65 system with an EC50 of ∼300 nM and CBA with an EC50 over 10 μM. In contrast, curcumin is inactive. CNB-001 also inhibits the accumulation of Aβ and APP C-terminal aggregates in MC65 cells (Fig. 4c). Together, these data clearly demonstrate that CNB-001 has the ability to inhibit nerve cell toxicity caused by both intracellular and extracellular amyloid.
Curcumin binds to and dissociates Aβ amyloid aggregates, thereby giving it tremendous potential for the treatment of AD (Yang et al. 2005; Cole et al. 2007). To determine if CNB-001 or CBA are more or less effective than curcumin at inhibiting Aβ1–42 aggregation, aggregation in the presence of increasing concentrations of CNB-001, CBA and curcumin was assayed by the thioflavin T method. Congo red, a potent inhibitor of Aβ aggregation, was used as a positive control. Figure 4d shows that there is about an order of magnitude decreasing difference in the ability to inhibit Aβ aggregation between Congo red, curcumin and CNB-001, while CBA is relatively inactive. A similar sequence of inhibition was observed when aggregation was measured by western blotting (not shown). These data show that curcumin is significantly better than CNB-001 in the direct inhibition of Aβ aggregation.
While a large number of compounds are neuroprotective in cell culture assays, a lesser number can be administered orally and pass through the blood–brain barrier into the brain. It was therefore asked if CNB-001 can enter the brain following gavage. Ten-week-old female BALB/c mice were given a single oral dose of CNB-001 emulsified in 2.5% carboxymethyl cellulose at 20 mg/mL and administered by gavage at 400 mg/kg body weight. Blood concentration was measured as a function of time, and brain concentration determined following perfusion with PBS to remove blood and the removal of the meninges. CNB-001 is rapidly absorbed into the blood and quickly distributes to the brain (Table 1). The maximal plasma concentration was reached around 2 h after administration while the maximal brain concentration was reached 1 h after gavage. Even after 6 h there was a significant amount of CNB-001 in the brain. These results suggest that CNB-001 can be orally administered and pass into the brain.
Table 1. Micromolar plasma and brain content after a single oral dose of CNB-001 at 400 mg/kg
Time after gavage (h)
The data presented above show that a pyrazole derivative of curcumin, CNB-001, has broad neuroprotective activity in cell culture assays that encompass models of TFW, excitotoxicity, oxidative stress, glucose starvation, and amyloid toxicity. CNB-001 was synthesized as a molecular hybrid of curcumin, a natural product that has a clinical history as an anticancer and anti-inflammatory drug, and CBA, a derivative of the estrogenic molecule, bisphenol A (BPA), a compound that is widely used in the synthesis of plastics. Individually, CBA and curcumin work in some but not all of the six assays described above, while the hybrid molecule is effective in all and has a much lower EC50 in the assays than either parental molecule. It follows that CNB-001 is a promising lead compound for the treatment of a variety of CNS neurodegenerative diseases, as well as ischemia and trauma.
Curcumin is an FDA-approved drug for cancer therapy (for review, see Aggarwal et al. 2003). In cell culture systems, curcumin has a wide variety of effects on cell signaling pathways which are both cell type and concentration dependent (see, for example Aoki et al. 2007; Cole et al. 2007; and references therein). Curcumin is also anti-inflammatory and is able to block and, in some cases, reverse several of the pathological and memory problems associated with AD in animal models (Frautschy et al. 2001; Lim et al. 2001; Cole et al. 2007). Curcumin reduces oxidative damage, inhibits glial activation, and has the ability to bind amyloid plaques and inhibit amyloid aggregation both in vivo and in vitro (for review, see Cole et al. 2007). The structure of curcumin, which has two methoxylphenol groups separated by a β-diketone bridge, allows it to chelate iron, and free iron has been implicated in a variety of neurodegenerative diseases (Molina-Holgado et al. 2007). CNB-001, a pyrazole derivative of curcumin, is likely to retain many of the biological activities of curcumin. In addition, CNB-001 is not metabolized as rapidly as curcumin, for the intact molecule can be found in the brain several hours after gavage (Table 1). Curcumin contains a labile β-diketone bridge and its in vivo metabolism is not completely understood (Sharma et al. 2007). Because it is more hydrophobic, the cell permeability aspect of CNB-001 is also likely to be better than that of curcumin. This is perhaps reflected in the results of the assay for intracellular amyloid toxicity (Fig. 4b and c), where CNB-001 is effective while curcumin is not, although curcumin is more effective in preventing Aβ aggregation (Fig. 4d).
In contrast to curcumin, there is no information about the biological properties of the other structural component of CNB-001, CBA. However, an analog of CBA that lacks the cyclohexyl group linking the two aromatic rings (Fig. 1) is BPA. BPA is an estrogenic compound commonly used in the production of phenol resins and multiple plastics. BPA has received a great deal of notoriety recently because it leaches from the food-contact surface lacquer coating for cans as well as plastic bottles, with potentially harmful consequences (Kang et al. 2006). It is unlikely that CNB-001 has steroid-like properties, for its structure is clearly much more related to that of curcumin than the basic steroid scaffold. However, the estrogenic properties of CNB-001 should be examined.
The antioxidant activity of CNB-001 is probably not responsible for its broad neuroprotective activity for the following reasons. First, while potent lipid-soluble antioxidants are effective at high concentrations (10–100 μM) in several of the assays described above, they are not protective in the excitotoxicity and glucose starvation assays. Second, when CNB-001, CBA and curcumin are assayed for direct antioxidant activity using the TEAC assay, curcumin is at least a 20-fold better antioxidant than CNB-001, yet it is much less active in promoting cell survival.
The above data demonstrate that a relatively simple structural modification of a natural product (curcumin) with desirable biological properties can lead to a molecule with unique properties related to, but distinct from the original. This approach, which does not require knowledge of a molecular target, is the foundation of much of our pharmaceutical industry, for about half of our modern drugs are based upon natural product scaffolds (Koehn and Carter 2005). Recently, there has been a trend that requires the identification of a specific target and then the synthesis of a high affinity ligand for this target (Pangalos et al. 2007). However, this approach may not be successful for the treatment of complex neurological diseases where several factors are involved in nerve cell death. Perhaps the identification of low affinity drugs with multiple targets such as curcumin and CNB-001 is a better approach. Curcumin and other groups of natural products such as the flavones have multiple biological functions, including the ability to reduce inflammation by inhibiting cyclooxygenases and lipoxygenases, to act as antioxidants, and the ability to bind to and inactivate toxic amyloid (see, for example, Goodman et al. 1994; Kim et al. 2005; Yang et al. 2005; van Leyen et al. 2006; Cole et al. 2007). In addition, these compounds have the ability to stimulate innate neuroprotective pathways in mammalian cells via the activation of phase 2 enzymes (Mattson and Cheng 2006). Therefore, compounds like curcumin have the ability to utilize a broad range of defense mechanisms as opposed to modifying the activity of only a single molecule. CNB-001 may share many of these positive pharmacological traits with curcumin with the added advantage of being more broadly neuroprotective by yet to be determined mechanisms.
The complete inhibition of a target molecule by a high affinity ligand within a disease associated pathway may not work in chronic diseases such as AD because essentially all enzymes and signaling molecules have multiple roles within the body, some of which are necessary to maintain viability. Perhaps a recent example of this problem is the toxicity of secretase inhibitors (Barten et al. 2006). It is also likely that a ligand that dissociates Aβ with high affinity will also be toxic, for the assumption that the molecular interactions stabilizing amyloid fibrils are unique to amyloid is certainly not valid; similar protein–protein interactions necessary for survival will also be affected. Therefore, the identification of drugs with more modest affinity for multiple targets may be the most effective against complex neurological diseases. Finally, while the majority of the drug development work for AD has been focused upon well defined targets such as the secretases and Aβ aggregation, much less work has been directed toward the ultimate problem, nerve cell death (Scatena et al. 2007). The synthesis of a small molecule that is protective in a variety of neurotoxicity assays is a small step in this direction.
This work was supported by U.S. Public Health Service Grants to DS and PM, the Bundy Foundation, and the Shiley Trust for Alzheimer’s Research. We thank Dr. Lee-Way Jin (University of California, Davis) for the MC-65 cells.