The use of multi-target drugs in the treatment of neurodegenerative diseases

Keywords::

• Alzheimer’s disease
• CNS drugs
• designed multiple ligands
• multifactorial pathoetiologies
• multifunctional drugs
• neurodegenerative diseases
• Parkinson’s disease
• polypharmacology

Neurodegenerative disorders lead to increased mortality and morbidity in older patients, and are a great burden on society, where there is currently no approved treatment to prevent the progression of these diseases. Many neurodegenerative diseases, such as Parkinson’s disease (PD) and Alzheimer’s disease (AD), Lewy body disease, cerebrovascular disease, frontotemporal dementias and neuronal deterioration following stroke, often display comorbidity with neurodegenerative symptoms and behavioral alterations. These include depressive illness and delusional characteristics often associated with schizophrenia [1,2]. Several studies employing a variety of methodological approaches have led to the identification of specific molecular mechanisms underlying cell death and consequential behavioral changes in neurodegenerative disorders [3]. It has thus become increasingly clear that these alterations must be addressed through strategies aimed at the development of therapeutic agents for neurodegenerative diseases that apply entirely novel concepts [4].

Current therapeutic approaches for the treatment of neurodegenerative diseases merely offer limited, and even worse, transient symptomatic benefits to patients, with no mitigation of the insidious loss of neuronal cells encountered in these conditions [5]. Neurodegenerative diseases have a multifactorial pathoetiological origin [6], and it is not surprising that conventional drug-discovery approaches that embrace a ‘one gene, one drug, one disease’ philosophy may not offer the best pathway towards the development of much needed disease-modifying therapeutics for these diseases [7,8]. Drugs that target only one protein are susceptible to resistance, one reason (among several others) being that even a single mutation in the target active site often substantially reduces compound binding affinity and, hence, efficacy. On the other hand, resistance to drugs that target multiple proteins would require the unlikely event of concurrent mutations appearing in multiple protein targets. In view of these concerns, many clinicians and basic scientists have become persuaded that a strategy aimed at the simultaneous targeting of multiple proteins (and therefore etiologies) involved in the development of a disease should be more beneficial than the currently accepted ‘silver bullet’ approach. This approach may be useful in designing drug treatments for a range of diseases [9]. Diseases of the CNS for which such an approach has been suggested include, among others, movement disorders, cognitive deficit disorders, negative symptoms in schizophrenia, Lewy body disease and depressive illness [10–13].

Polypharmacy versus polypharmacology

Polypharmacy

Multiple targeting using a combination of drug entities has been used in the clinical setting for several years through a polypharmaceutic approach. Polypharmacy is achieved simply by combining several drugs that independently act on different etiological targets of a disease or disease-causing organism. For example, such an approach is commonly used in AIDS therapy, where HIV reverse transcriptase inhibitors and HIV protease inhibitors are coadministered as a cocktail (reviewed in [14]). Another example is the approved combination – in one preparation – of fluticasone, a corticosteroid, and salmeterol, a bronchodilator, to simultaneously target the underlying inflammation and bronchoconstriction associated with asthma [15], or the combination of the calcium channel blocker amlodipine and the cholesterol-reducing agent atorvastatin in the treatment of cardiovascular disease [16]. However, a pharmaceutical combination of several drug molecules raises many challenges, not the least of which is the associated complexities encountered when combining drug entities that have potentially different degrees of bioavailability, pharmacokinetics and metabolism [17,18]. Even worse, a combined or even multiplied toxicity and side-effect profile may be experienced, while it is likely that unforeseen drug–drug interactions may occur [19]. In aged patients – the at-risk population for neurodegenerative diseases such as AD [20] – certain side effects generated in this fashion may be life-threatening [21]. It is therefore not surprising that research has focused increasingly on the advantages associated with the design of single drug molecules that act on two or more specific etiological targets of a particular disease. The likelihood of encountering unwanted side effects is less and any side effects may be easier to ‘design out’ when only one ligand is used, as opposed to using two or more ligands.

Polypharmacology

In a 2005 review, Morphy and Rankovic identified more than 300 reports in drug discovery and development journals published between 1990 and 2004 where compounds were designed as multiple-ligand drugs [22]. The findings of this review were further elaborated on and updated in a more recent publication by the same authors [23]. The actual number of such compounds under development may even be under-represented in these studies, owing to the inconsistency in nomenclature used by researchers when describing ligands designed for multiple drug targets in a disease state. For example, numerous terms are used to describe the mechanisms of these drugs that include: ‘designed multiple ligand’, ‘dual-mechanism’, ‘dual ligand’, ‘bifunctional’, ‘multifunctional’, ‘multimechanistic’, ‘multimodal’, ‘pan-agonist’ or ‘hybrid’ drugs [8,22–27].

The development of disease-modifying therapeutics that address the principal causes of neurodegenerative diseases faces significant obstacles. In addition, the design of multiple ligands is a particularly challenging endeavor for drug development scientists [22,23]. The need to appropriately balance the affinities of one drug molecule towards two or more targets while obtaining physicochemical and pharmacokinetic properties that are consistent with the administration of an oral drug is particularly demanding [23]. In neurodegenerative diseases, current lead generation strategies utilize drug targets that are identified based in large measure on the discovery of disease gene products and brain enzymes controlling their metabolism [25–27]. The properties of designed multiple ligands (DMLs) are dictated, to a large extent, by the proteomic superfamily to which the protein targets belong [23,28]. Therefore, a significant challenge (and in some cases, a barrier) to the development of drugs with appropriate polypharmacological properties for neurodegenerative diseases is the frequent absence of clear evolutionary relationships among the diverse potential protein targets that are aimed for in the design of a given compound [28]. These obstacles require that design approaches should be capable of identifying protein targets (receptors or enzymes) independently of global sequence or structural homology [22,23,28]. Molecular biology and genomics approaches, combined with computational resources and robotic instrumentation applied to genetic analyses and biological data, have united to accelerate the pace of discovery of credible drug targets in the design of DMLs for neurodegenerative diseases [29,30].

Select examples of multi-target drug candidates currently under investigation for neurodegenerative diseases

Anti-amyloid compounds with anticholinesterase activity

In a series of groundbreaking discoveries, the groups of Bolognesi and Melchiorre have explored the polyamine-quinone compound memoquin and a series of related compounds [27,31–34]. These compounds address several mechanisms relevant to AD, including the processing and aggregation of amyloid β (Aβ) peptides, the formation of reactive oxygen species (ROS) and acetylcholinesterase (AChE) inhibitory activity. In animal models, memoquin and several of its derivatives caused a significant decrease in Aβ peptide and ROS formation. Notably, these compounds also were able to elicit a significant reversal of behavioral deficits [31–34]. Recently, a novel series of memoquin derivatives was created by linking the 2,5-diamino-benzoquinone core of these compounds with motifs seen in known amyloid-binding agents, including the naturally occurring polyphenol curcumin [35]. The authors found that weaker AChE inhibitory potencies, with concurrent equipotent anti-Aβ activities of the new compounds compared with memoquin, yielded more balanced biological profiles against the combined targets of Aβ and AChE, thereby addressing one of the great challenges in the design of DMLs (see previously and [23]). Thus, by the appropriate combination of two or more distinct pharmacological properties in one molecule, these authors were able to demonstrate greater effectiveness compared with single-targeted drugs in two of the most important pathoetiological pathways of AD [35].

Cholinesterase/monoamine oxidase-B inhibitors

The identification of the propargylamine moiety as a key element that confers neuroprotective activity to many compounds designed for neurodegenerative disorder, led to the development of the propylcarbamate, ladostigil (TV3326), a novel neuroprotective agent being investigated for the treatment of neurodegenerative disorders including AD, Lewy body disease and PD [36,37]. Ladostigil is a dual cholinesterase–monoamine oxidase-B (MAO-B) inhibitor, designed by combining the carbamate cholinesterase inhibitory moiety of the rivastigmine molecule, with the MAO inhibitory propargylamine moiety of the rasagiline molecule. Ladostigil has been shown to attenuate 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine toxicity in mice, a model of parkinsonism [38]. Although a poor MAO-B inhibitor, TV3279 (the S isomer of ladostigil) has shown similar neuroprotective activity in laboratory animals [39].

Metal chelators/MAO-B inhibitors with additional cotargeted mechanisms

Degeneration of cholinergic cortical neurons is the main pathological feature of the cognitive deficit in dementia in AD and in dementia with Lewy bodies (DLB). Many DLB subjects have dementia and depression resulting from possible degeneration of cholinergic, noradrenergic and serotonergic neurons. In both AD and DLB, accumulation of iron is found inside some melanin-containing dopaminergic neurons and inside amyloid plaques and neurofibrillary tangles [40]. It has been suggested that iron accumulation may contribute to the oxidative stress-induced apoptosis reported in both diseases [41]. Such oxidative stress may result from increased glial MAO activity, leading to increased hydrogen peroxide production that can generate reactive hydroxyl radicals through Fenton chemistry with intracellular iron. Based on this proposal, Zheng et al. developed neuroprotective compounds with dual iron chelating and MAO-B-inhibitory activity [41]. These authors combined the antioxidant chelator moiety of an 8-hydroxyquinoline derivative of the neuroprotective brain-permeable iron chelator VK-28 [42] with the propargylamine moiety found in compounds such as rasagiline and deprenyl. The resulting compound, HLA20, was identified as a potential lead compound for further studies and has been characterized as a brain-permeable iron chelator and a brain-selective MAO inhibitor that possesses the propargyl neuroprotective moiety [43].

Recently, Lim’s group has developed compounds with structures that – based on known structure–activity relationship experience in MAO inhibition [44] – may possess MAO-inhibitory properties [45]. These compounds have emerged as bifunctional candidate small molecules that chelate Aβ-associated metal species and regulate metal-induced Aβ aggregation and oxidative stress-induced neurotoxicity. The salutary influence of the bidentate ligand clioquinol on metal-involved Aβ aggregation has been explained through its metal chelation, which assists, in part, to disaggregate Aβ aggregates. However, clioquinol does not completely hinder the progression of Aβ aggregation [46]. Thus, these investigators reported the synthesis of two bifunctional small molecules, N-(pyridin-2-ylmethyl)aniline and N(1),N(1)-dimethyl-N(4)-(pyridin-2-ylmethyl)benzene-1,4-diamine, and demonstrated that these compounds interact both with metal ions and Aβ species [44]. N(1),N(1)-dimethyl-N(4)-(pyridin-2-ylmethyl)benzene-1,4-diamine modulated metal-induced Aβ aggregation and neurotoxicity in vitro, as well as in human neuroblastoma cells, while simultaneously disassembling Aβ aggregates in human AD brain tissue homogenates containing metal ions and the Aβ protein. These findings follow an earlier study by the same group with two different compounds: 2-[4-(dimethylamino)phenyl]imidazo[1,2-a]pyridine-8-ol and N(1),N(1)-dimethyl-N(4)-(pyridin-2-ylmethylene)benzene-1,4-diamine, which were found to react with Cu(II)-associated Aβ aggregates more effectively than known metal chelating agents, such as clioquinol and ethylenediaminetetraacetic acid [47]. Of note is the fact that these compounds show great promise as potential MAO inhibitors, an action that will bolster the multimodal promise of these investigational new drugs as potential therapeutics in AD and PD [Geldenhuys WJ, Lim MH, Van der Schyf CJ, Unpublished Data]. Experiments are currently underway and results of these studies will be published in due course [Lim MH, Pers. Comm.].

Adenosine A_2A antagonists with MAO-inhibitory activity

In AD and PD, dual inhibition of MAO-B, as well as adenosine A_2A receptor blockade, may be a novel therapeutic approach to prevent neuronal cell death. Caffeine, a nonselective adenosine receptor antagonist, is under some scrutiny as a potential drug to counteract age-related cognitive decline [48]. Caffeine, in fact, has been suggested to protect against Aβ neurotoxicity [49], while acute treatment with caffeine and the A_2A receptor antagonist ZM241385 was found to reverse age-related olfactory deficits and memory decline in rats [50], clearly suggesting involvement of A_2A, and not A₁, receptors in cognitive decline and, possibly, neurodegenerative processes. Evidence such as this, and other evidence for neuroprotection also in parkinsonian models, led to the evaluation of (E)-8-styryl-xanthinyl-derived adenosine A_2A receptor antagonists also for inhibition of MAO-B in several studies (reviewed in [51]). All of the compounds tested demonstrated MAO-B inhibition in the low micromolar to high nanomolar range, suggesting that the neuroprotective properties of known adenosine A_2A receptor antagonists may be in part due to MAO-B inhibition, in synergism with their A_2A antagonism [51].

NMDA antagonists that also act as calcium channel blockers

Several studies have shown that NMDA receptor antagonists, such as dizocilpine (MK-801) and the clinically used polycyclic cage amine memantine, display neuroprotective effects in the clinic and in in vitro experiments in neurons [52]. An alternative pathway for calcium to enter into neuronal cells is through voltage-gated ion channels, such as L-type calcium channels. Animal experiments with the brain-permeable L-type calcium channel antagonist nimodipine have suggested that calcium channel antagonists may be neuroprotective in ischemia by antagonizing the influx of calcium into neuronal cells [53]. The importance of calcium overload during cell death suggests that a dual calcium channel and NMDA receptor antagonists might be useful as neuroprotective drugs in neurodegenerative diseases such as AD.

The polycyclic cage amine, 8-benzylamino-8,11-oxapentacyclo[5.4.0.0^2,6.0^3,10.0^5,9] undecane (NGP1-01), has been investigated as an L-type calcium channel blocker and as an NMDA receptor antagonist [54]. Structurally similar to memantine (an uncompetitive NMDA receptor antagonist used clinically to treat AD), NGP1-01’s favorably fast on-off binding kinetics should afford this compound an improved side-effect profile compared with powerfully binding NMDA antagonists such as MK-801 [55]. NGP1-01 was shown to also be an uncompetitive NMDA antagonist in murine whole-brain synaptoneurosomes [Geldenhuys WJ, Bloomquist JR, Van der Schyf CJ, Unpublished Data] and blocked NMDA-mediated ⁴⁵Ca²⁺ uptake with an IC₅₀ of 2.98 µM. This dual mechanism of modulating calcium entry into neuronal cells (via L-type calcium channels and the NMDA receptor) suggests that NGP1-01 may have utility as a neuroprotective agent in AD, stroke and other neurodegenerative diseases with cognitive decline [56].

Conclusion

The treatment of neurodegenerative diseases poses perplexing challenges. These challenges are partly due to the complex pathology involved in the etiology of these diseases. The utility of an emerging approach that purports to use a single polypharmacological drug molecule (i.e., one that acts on more than one drug target in the etiological pathway of the disease), may offer new hope in the treatment of many neurodegenerative diseases, including those associated with cognitive decline. Such a transition from a ‘magic bullet’ to a ‘magic shotgun’ approach may be key in designing future treatment regimens for neurodegenerative diseases and associated cognitive decline. Drug-discovery endeavors will have to shift focus from the design of selective agents that target only one pathophysiological pathway, to the design of agents (such as DMLs) that operate through manifold mechanisms and are intended to target the very complexity of the disease state.

Financial & competing interests disclosure

The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.