Clinico-Pathologic Function of Cerebral ABC Transporters – Implications for the Pathogenesis of Alzheimer’s Disease

Abstract

In recent years it has become evident that ABC transporters fulfill important barrier functions in normal organs and during disease processes. Most importantly, resistance to drugs in cancer cells led to intense oncological and pharmacological investigations in which researchers were able to highlight important pharmacological interactions of chemotherapeuticals with ABC transporter function. Recently, the development of neurodegenerative diseases and the maintenance of neuronal stem cells have been linked to the activity of ABC transporters. Here, we summarize findings from cell culture experiments, animal models and studies of patients with Alzheimer’s disease. Furthermore, we discuss pharmacological interactions and computational methods for risk assessment.

THE ABC TRANSPORTER SUPERFAMILY

The superfamily of ATP-binding cassette (ABC) transporters regulates drug bioavailability, metabolism, and distribution in cells and the extracellular matrix; the transporters thus play an important role by limiting drug accumulation in tissues, and have been implicated in drug resistance [1, 2]. Some family members are currently being regarded also as markers for neural stem/progenitor cells (NSPC) [3–7]. The superfamily includes several multidrug transporters (see also http://nutrigene.4t.com/humanabc.htm) belonging to one of the seven ABC A to ABC G subfamilies. At the blood-brain barrier (BBB), these subfamilies include the ABCB subfamily corresponding to P-glycoprotein (P-gp) (which mediates multidrug resistance [MDR]), the ABCC subfamily of multidrug resistance-associated proteins (MRP), and the ABCG2 subfamily (breast cancer resistance protein - BRCP) [2]. Multidrug transporters have in common large and partially overlapping substrate specificities for numerous hydrophobic compounds. They are specifically distributed in organs with secretory or barrier functions, including gut, kidney, liver, lung, and, more importantly for neurodegenerative diseases, also at the BBB and the choroid plexus. In endothelia or epithelia, they are sorted either to the apical membrane, facing the luminal surface of vessels, tubuli or canaliculi, or to the basolateral membrane. This functional and polarized arrangement at the cell membrane enables guided transport through these cellular barriers [2]. Thus, ABC transporters fulfill an important function in excreting various compounds and peptides. Especially in diseases with compartmental involvement of organs, this barrier function plays an eminent role in connecting the intra-compartmental cellular and intercellular environment with the extra-compartmental systemic bloodstream. As the brain must be regarded as one of the largest compartments surrounded by a barrier, this function may also play a role in the pathophysiology of neurodegenerative diseases.

ABC TRANSPORTERS AT THE BLOOD-BRAIN BARRIER

The BBB is a physical and metabolic barrier between the brain and the systemic circulation that serves to regulate and protect the microenvironment of the brain. The BBB is composed of a monolayer of brain capillary endothelial cells. The restriction of brain uptake by the BBB is due to the presence of tight junctions (zonulae occludens) between adjacent endothelial cells and a relative paucity of fenestrae and pinocytotic vesicles within the endothelium of cerebral arterioles, capillaries, and venules [8]. The brain capillary endothelial cells are bordered by an extracellular matrix (basal membrane), pericytes, and astrocytic foot processes. Because of the presence of the BBB, circulating molecules gain access to the brain only via lipid-mediated transport of small nonpolar molecules through the BBB by free (passive) diffusion, or, for some substances, by catalyzed transport [8]. The endothelial cells of the BBB contain numerous membrane transporters involved in the influx or efflux of various essential substrates such as electrolytes, nucleosides, amino acids, peptides, and glucose [9]. It was originally believed that membrane carriers localized at the BBB were solely responsible for the transport of such endogenous substances into and out of the brain, and that drug transport across the BBB was largely dependent on the physicochemical characteristics of the drug such as lipophilicity, molecular weight, and ionic state. However, ABC efflux transporters at the BBB limit the brain uptake of a variety of therapeutic agents, including compounds that are relatively lipophilic and would be predicted to permeate the endothelial lining of the brain microvasculature [9].

P-gp (MDR1), MRP (MRP1, 2 and 5) and BCRP play a critical role in the BBB by functioning as active efflux pumps that are able to convey a wide range of structurally diverse compounds out of the CNS. Physiologically, their role is to export endogenous substrates, or ingested or acquired and potentially neurotoxic xenobiotics or waste products from nervous tissue; they thus carry out a critical neuro-protective and detoxifying role for the brain. With regard to neurotoxic waste products, it is of major importance to note that the β-amyloid peptide (Aβ) is also transported by these molecules [10].

Many prescribed therapeutics are substrates for these efflux transporters. The activity of the ABC transporters at the BBB therefore becomes crucial in significantly limiting the entry of a large number of therapeutic drugs into the CNS and the excretion of neurotoxic metabolites. The interactions of drug import and export and the excretion of toxic metabolites play a prominent role in determining the risk for the development of neurological diseases due to hampered excretion function (e.g. Alzheimer’s disease [AD]) and the need for treatment of systemic disorders (e.g. hypertonus, hyperlipidemia). This phenomenon presents a considerable challenge to effective drug delivery and to the investigation of the pathogenesis and the progression of CNS proteopathies, e.g. AD, Parkinson’s disease, progressive supranuclear palsy, and multi systems atrophy [11].

The precise cellular and subcellular distribution of ABC transporters at the BBB remains controversial, mainly due to the difficulty in discriminating their localization in these cells with closely apposed cell membranes. Moreover, several monoclonal antibodies raised against human epitopes do not recognize the murine proteins, whereas polyclonal antibodies made against the corresponding mouse epitopes allow sensitive detection [12]. Thus, numerous investigators have used various experimental approaches such as in situ hybridization, immunohistochemistry and multilabeling immunofluorescence strategies to determine the precise location of transporters in capillary endothelial cells [12–14]. However, there is virtually no information about spatial and temporal expression differences in different brain areas of murine models, human subjects, and more importantly during the development of neurodegenerative diseases.

P-Glycoprotein (P-gp, MDR1, ABCB1)

P-gp was the first ABC transporter that was detected in endothelial cells of the human BBB [15]. Subsequently P-gp was localized in brain capillary endothelial cells of several species, including monkeys, rats, mice, cattle, and pigs, suggesting that P-gp may serve as general defense mechanism in the mammalian BBB, protecting the brain from intoxication by potentially harmful lipophilic compounds that otherwise could penetrate the BBB by simple diffusion [16]. P-gp is encoded by a small group of related genes called multidrug resistance genes (MDR). In rodents three P-gp gene products have been identified - Mdr1a, Mdr1b and Mdr2 [17]. In humans there are two gene products known - MDR1 and MDR2 [18]. Only MDR1 confers multi-drug resistance, since MDR2 is predominantly expressed by hepatocytes at the canalicular side for the secretion of phosphatidylcholine into the bile fluid [19]. P-gp is located at the luminal side of cerebral endothelial cells and transports substrates directly into the bloodstream (Fig. 1). It shuffles a variety of substrates with an immense structural diversity out of the brain (Table 1). With respect to AD, P-gp only recently has been recognized to be involved in Aβ transport [10, 20, 21].

Fig. (1).

ABC transporters at the blood-brain barrier. Top: Schematic drawing of a brain capillary. The BBB is comprised essentially of the endothelial cells (E), the basal membrane (BM), and astrocytic processes (AP). Pericytes (not shown) also contribute to the cellular barrier. The cerebral endothelial cells are connected via tight junctions (TJ). Bottom: Localization of ABC transporters at the brain capillary endothelial cells. Transporters localized in either the luminal (apical) or basolateral part of the cell membrane restrict uptake (apical) or excretion (basolateral) of drugs and toxic metabolic products. P-gp – P-glycoprotein, MRP – multidrug resistance protein, BCRP – breast cancer resistance protein. Based on data from [2, 9, 12, 59, 96].

Table 1.

Selected Therapeuticals and Chemical Compounds that are Transported by P-gp. Some of these Drugs are Widely Used for the Treatment of Chronic Diseases, e.g. High Blood Pressure (Beta Blockers) and Hyperlipidemia (Statins). A Majority of these Compounds Inhibit the Excretion Function of P-gp at the BBB for Neurotoxic Metabolites from the Brain. Thus, Life-Long Treatment Using these Compounds Might Increase the Intracerebral Content of Metabolites Such as Aβ. In Contrast, the Antibiotic Rifampicin is the Strongest Known Activator of P-gp Function.

Ca++ channel blockers Verapamil, Diltiazem, Mibefradil, Nefedipine, Nitrendipine	HIV protease inhibitors Ritonavir, Saquinavir, Amprenavir	Antitumor drugs Etoproside, Imatinib, Lonafarmib, Vincaalkaloids, Taxanes
Immunosuppressants Cyclosporine A, Tacrolimus, Sirolimus	Statins Atorvastatin, Lovastatin, Simvastatin	Antibiotics Erythromycin, Rifampicin
Opoids Morphine-6-glucuronide Methadone, Loperamide, Fentanyl	Beta blockers Carvedilol, Celiprolol, Talinolol, Bunitrolol, Carazolol	Diverse Digoxin, Midazolam, Ivermectin, Octeotride, Ranitidine, Methotrexate, Rhodamine-123, St John’s Wort

Multidrug Resistance-Associated Proteins (MRP, Mrp)

In contrast to P-gp, data on other ABC transporters at the BBB are much more limited and controversial. Mrp1, Mrp2 and Mrp5 can be detected in brain endothelial cells, with Mrp5 being located at the luminal side, Mrp1 on the abluminal side (basal), and Mrp2 in between. Recent studies of six murine ABC transporters (Mdr1a, Mdr3, Mrp1, Mrp2, Mrp3, and Mrp5) highlighted different cellular and subcellular localizations in different mouse strains [12]. These data emphasize the need for precisely specifying the type of sample, species, and strain analyzed. The authors were able to show endothelial expression of Mrp2 in the brains of C57Bl/6, Swiss, and SvJ mice, but not in FVB mice. It is worth noting that several ABC transporter knockout mice were established in FVB or FVB-related backgrounds [22–25]. Thus, it should be recognized in the experimental setting that ABC transporters might already be expressed differentially in the particular mouse strain employed in the investigations. MRPs comprise a group of closely related gene products that show less than 15% homology to P-gp [26]. The substrate preferences of a number of MRPs and P-gp exhibit overlap, but MRPs show a preference for organic anions and many drugs that are conjugated to glutathione, glucuronate or sulphate due to CNS enzyme activities [27].

Breast Cancer Resistance Protein (BCRP, ABCG2, Bcrp-1)

BCRP appears to be expressed in the luminal membrane of cerebral endothelial cells in a similar manner to P-gp [13]. Interestingly BCRP is essentially a half transporter when compared to P-gp, with only six transmembrane domains and one nucleotide binding site, and it is therefore assumed to form a homodimer in the cell membrane so that it can function as a transporter [28, 29]. P-gp and BCRP seem to share some functional parallels. BCRP mRNA is upregulated 3-fold in P-gp (Mdr1a) knockout mice [30]. This finding highlights the care that must be taken in the analysis of different mouse models in which the lack of the expression of a specific ABC transporter may lead to compensatory changes in gene expression and protein synthesis. BCRP was primarily isolated from human breast tumor cell lines that were negative for P-gp and MRPs, yet displayed a high resistance to mitoxantrone [31].

ABC TRANSPORTER ABLATION ANIMAL MODELS

The majority of multidrug transporter genes have one rodent orthologue. However some of the human genes are encoded in two genes. P-gp, discovered by Juliano and Lin in 1976 [32], is encoded in two genes in mice, Mdr1a and Mdr1b, with ubiquitous, complementary tissue distribution [22, 23]. In recent years, new models with ablation of various BBB ABC transporters have been reported (Table 2). These animal models have been analyzed primarily with respect to alterations in the bioavailability of pharmaceuticals [9, 16, 22, 33–35]. However, the models also provide information about compensatory expression, e.g. triple knockout mice lacking Mdr1a, Mdr1b and Mrp1 show an upregulation of Bcrp-1 during treatment with mitoxantrone, topotecan or doxorubicin [36]. Recently, we established new animal models for the investigation of the involvement of BBB ABC transporters during the initiation of Aβ deposition in the parenchyma and vasculature of AD mouse models. These mice overexpress mutated β-amyloid precursor protein (APP) and presenilin-1 (PS1) [37], or Dutch-type APP [38], or lack the expression of neprilysin [39] and are combined knockouts for one, two, three or four different ABC transporters (Mdr1a & b [23], Mrp1 [24], Bcrp-1 [25]). These different experimental models enable the collection of quantitative data, providing the basis for mathematical models.

Table 2.

ABC Transporters with Known Function at the Blood Brain Barrier and/or Involvement in AD

Human transporter	Mouse transporter	BBB expression	Knockout model published by
ABCA2	ABCA2	no	[97]
ABCB1/P-gp/MDR1-P-gp	Mdr1a & b Mdr1a	Luminal	[23]
ABCC1/MRP1	Mrp1	Luminal a/o Basal	[98]
ABCC2/MRP2	Mrp2	Luminal a/o Basal	[99]
ABCC4/MRP4	Mrp4	Luminal a/o Basal	[33]
ABCC5/MRP5	Mrp5	Luminal	[100]
ABCG2/BCRP	Bcrp-1	Luminal	[25]

PHARMACOLOGICAL INTERACTIONS WITH ABC TRANSPORTER FUNCTION

The role of ABC transporters in neurodegenerative diseases is intriguing, as the activity of these transporters can be modulated by several drugs [40]. Recently, a number of modulators of ABC transporter-mediated drug resistance have been developed, primarily for P-gp activity. Observations made on many of these inhibitors have been largely incidental. Additionally, there are different groups of therapeutics that are currently used for the treatment of cardiovascular diseases, e.g. hypertonus and hyperlipidemia. These drugs also coincidentally interfere with ABC transporter function at the BBB. Some of the drugs do not interfere with just a single ABC transporter, and moreover, the vast majority of the drugs change the transport kinetics of a variety of related transport proteins. These interactions have already been recognized, since the multi-drug treatment of various disorders can raise plasma and tissue drug levels to toxic amounts. Numerous problems may be associated with general inhibition of ABC transporter activity. An impressive example is congestive heart failure caused by a combined therapy of verapamil and doxorubicin at sufficient doses to inhibit P-gp [41]. Inhibitors of efflux transporters, when applied coincidently for general treatment or to enhance CNS uptake of therapeutics, are better suited to acute treatment regimes where the object is to achieve short maximal concentrations, for example in the treatment of brain tumors, than for chronic administration where long-term inhibition of ABC transporter-mediated efflux will disturb normal homeostasis. With respect to the cerebral proteopathies, this means that extended treatment with ABC transporter inhibitors will lead to raised intracerebral concentrations of a range of neurotoxic substances that enter the brain from the bloodstream or are being produced within the brain. These changes of the transport kinetics and the accumulation of neurotoxic peptides or peptide fragments must also be taken into account when evaluating the pathogenesis of neurodegenerative diseases such as AD, Lewy body disease, and Parkinson’s disease (PD). Most importantly, the extent of transport kinetics changes that take place during long-term treatment, in relation to the intracerebral concentration of neurotoxic peptide moieties, has yet to be investigated intensively in vivo.

ABC TRANSPORTERS ARE MARKERS FOR NEURAL STEM/PROGENITOR CELLS

Stem cells can be purified based on the efflux of fluorescent dyes such as rhodamine 123 and Hoechst 33342. This method is widely used to extract hematopoietic stem cells [42, 43]. However, these dyes also can be used to evaluate ABC transporter function [44]. Just recently, it has become clear that ABC transporters are expressed in a variety of stem/progenitor cells. Various groups have demonstrated expression of P-gp, BCRP, and ABCA2 in human and murine NPSCs [3–5]. The expression of different ABC transporters in NSPCs of adult mouse and rat brains appears to be region-dependent. For example, it was found that ABCA2 expression is limited to the oligodendrocytic lineage (unambiguously distinguished from astrocytes), and to a subset of cortical GABAergic interneurons and pyramidal glutamatergic neurons where it could be localized to lysosome-related organelles [4]. Considering that the latter are targets of Alzheimer’s lesions, these data provide a new rationale to explore the neuropathological consequences of ABC transporter functional dysregulation. ABCA2 also has been suggested to be a marker of neural progenitors, as it is expressed in the subventricular zone and the dentate gyrus, sites of ongoing neurogenesis in the adult brain. Moreover, it was found by another research group that all nestin-positive cells within neurospheres simultaneously express P-gp [4]. Using fluorescence-activated cell sorting (FACS) with an antibody against P-gp, it was possible to extract even more human NSPCs than the nestin-positive population [7]. In an additional investigation, the same researchers showed that both BCRP and P-gp are expressed functionally at the plasma membrane of NSPCs. Furthermore, these experiments highlighted that the expression of P-gp, BCRP and nestin was reduced during differentiation, while expression of the astrocytic marker glial fibrillary acid protein (GFAP) was upregulated [3]. Thus, ABC transporters may have a functional role in maintaining the undifferentiated status of NSPCs, as the downregulation of nestin and upregulation of GFAP are considered to be indicators of NSPC differentiation or maturation. ABC transporters have emerged as an important new field of investigation in the regulation of stem cell biology and the development of neurodegenerative diseases. It will be of major interest for the AD research community to highlight the interactions and pathways involved therein – regulation of compound excretion (including Aβ efflux) and stem cell maintenance function. Altered neurogenesis in AD patients has been recognized by various groups [45, 46]. Suggestive as a key player is the Wnt signaling pathway and its key enzyme glycogen synthase kinase-3 beta (GSK3β), as this enzyme is known to be involved in pathways that contribute to the progression of AD [47–51].

INVOLVEMENT OF ABC TRANSPORTERS IN ALZHEIMER’S DISEASE

The development of Aβ plaques in the extracellular compartment of the brain parenchyma is a hallmark of AD, and biochemical and genetic findings underscore the importance of the Aβ peptide in the pathogenesis of AD [52]. Amyloid plaques consist predominantly of Aβ peptides, mainly Aβ40 and Aβ42. The Aβ peptides are produced continuously in cells of the nervous system and systemic tissues by the cleavage of the amyloid precursor protein (APP). With age, soluble Aβ accumulates over time in the brain extracellular space, multimerizes into oligomers, protofibrils and insoluble fibrils, and eventually forms diffuse and dense-cored amyloid plaques [53–55]. Studies in AD patients have detected increased levels of peripherally circulating Aβ [56, 57]. It was suggested that Aβ in blood plasma and cerebrospinal fluid exists in equilibrium (which is controlled by an unknown mechanism) that shifts the concentration toward the brain during plaque development [58]. Normally, the brain endothelium, a key element of the BBB in vivo, does not allow free exchange of polar solutes such as Aβ between brain and blood, or blood and brain, due to the presence of a continuous cellular monolayer of tightly junctioned endothelial cells [59]. Thus, specialized transporters for Aβ must exist in brain endothelium to expel the peptide from the CNS into the bloodstream, and/or to shuttle circulating Aβ into the CNS. It has been proposed that the Aβ equilibrium between plasma and CSF is regulated at the BBB by an influx receptor (RAGE, receptor for advanced glycation end products) and an efflux receptor (LRP1, low-density lipoprotein receptor-related protein) [60]. Apolipoprotein E is considered to be involved in the RAGE/LRP1 shuffling mechanism [61], and polymorphisms in apolipoprotein E (ApoE) are known to markedly influence the risk of late-onset AD [53, 62–68].

In 2001, Lam and colleagues [10] investigated ABC transporter-mediated Aβ efflux in vitro in transfected HEK293 cells. By applying the P-gp inhibitors RU486 and RU49953, they showed that P-gp serves as an Aβ efflux pump. Additionally, fluorescence quenching binding affinity determinations and transport competition experiments supported their data. The results were confirmed in in vitro cell culture experiments using MDR1-transfected polarized kidney epithelial cells (LLC-PK1) [44]. In patients, differences in cerebrovascular P-gp expression were shown to influence Aβ deposition rates, although the determination of three known polymorphisms failed to be of significance [69]. Interestingly, the inverse correlation between Aβ deposition and P-gp expression was most striking with regard to Aβ40 positivity. Investigations of the correlation between cerebrovascular β-amyloid angiopathy (CAA) and P-gp function indicated a loss of P-gp in vessels with abundant cerebrovascular Aβ in non-demented elderly humans [20]. However, in that study, it was unclear whether the correlation was due to a primary loss of P-gp expression or a diminution of P-gp expression secondary to Aβ deposition. Kinetic studies of Aβ influx/efflux in wild type and double transgenic AD mice [70] using ¹²⁵I-labeled Aβ40 yielded conflicting results [71, 72]. Interestingly, Kandimalla and colleagues demonstrated that higher Aβ40 levels in the peripheral circulation in AD mice compared to wild-type mice are not due to differences in the plasma, hepatic, or renal metabolism of the peptide.

ABC transporters, in particular P-gp, may also play a role in the intracerebral/intracellular distribution of Aβ, as they have been detected in different intracerebral (neurons, astrocytes, microglia) and intracellular compartments [73–79]. These findings are supported by our own observation of early intracellular accumulations of Aβ in small vesicles in Mdr1a/b double knockout mice expressing the Dutch-type variant of APP F (Fig. 2). New APP-overexpressing AD mouse models (APPPS1 [37]) that were depleted of P-gp (Mdr1a and Mdr1b [23]) show that BCRP function partially compensates for P-gp-mediated Aβ clearance. BCRP functions as a clearance transporter with limited capacity relative to P-gp. BCRP compensates for rising soluble Aβ concentrations up to an intracerebral concentration of 200ng/mg brain protein, thus leading to double sigmoidal deposition kinetics in these murine models.

Fig. (2).

Accumulation Aβ-immunoreactive material (antibody clone 4G8) in a 175 day-old Mdr1a/b-deficient mouse (B) expressing the Dutch-type variant of APP. Large pyramidal neurons in layer 5 (D – cortex & F – layer 5) and neurons in CA1-3 of the hippocampal formation (E) show small, Aβ-positive endolysosomes. Control DTAPP-type mice (A), which are not deficient for Mdr1a/b, do not show intracellular deposits (C – CA1 of hippocampal formation). Scale bars: A, B and D: 500μm; E: 100μm; C and F: 50μm.

Large genetic studies using 326 microsatellite markers in 466 late-onset AD families have demonstrated various new chromosomal regions with linkage to AD [80]. Matching ABC transporter locations with these new regions, a number of transporters can be found in close proximity to genetic hotspots, for example ABCA2 located on chromosome 9 (marker D9S1818). The most exciting finding was on chromosome 9, where the nonparametric affected sibpair maximum LOD score (MLS) in the combined sample was 2.97 [77]. The results for chromosome 9 were enhanced when only the autopsy-confirmed subset with an MLS of 4.31 was analyzed. This was the highest LOD score yet seen in a genomic screen for AD, even larger than the score for D19S246, the closest screening marker to apolipoprotein E in that study. Later, an in vitro cell culture investigation using neuroblastoma cells demonstrated a close relationship of APP and ABCA2 in endolysosomal compartments [81]. Analyzing brain tissue samples, Chen and colleagues demonstrated that ABCA2 levels were highest in temporal and frontal regions of the AD brain, with lower but significant levels in parietal, occipital, and cerebellar regions. However, similar ABCA2 protein expression patterns were observed in normal and schizophrenia brain samples. Thus, it will be important to define precisely whether there is indeed a role of ABCA2 in the pathogenesis of AD.

Unlike ABCA2, ABCA1 and ABCG2 have been suggested to play a significant role in the regulation of neuronal cholesterol efflux to apoE discs, and in the regulation of APP processing to generate Aβ peptides [82]. ABCA1 serves as a major regulator of cholesterol efflux and high density lipoprotein metabolism. It was demonstrated that targeted disruption of ABCA1 increases Aβ deposition in APP23 mice [83]. The lack of ABCA1 also considerably increased levels of CAA and exacerbated CAA-related hemorrhages in these mice. Remarkably, the augmentation of parenchymal and vascular Aβ was accompanied by a dramatic decrease in the level of soluble brain apoE, whereas insoluble apoE was not changed. Just recently, genetic analysis of polymorphisms in the promoter and coding region of ABCA1 indicated that the development of AD might be influenced by either a qualitative change of the ABCA1 protein caused by coding region variants (219K, 883I, and 1587R), or by a quantitative change in ABCA1 expression caused by promoter region variant (−14T) in concert with the APOE epsilon4 allele [84].

LONG-TERM INHIBITION OF ABC TRANSPORTER FUNCTION – WHERE DO WE GO?

Various treatment regimes for cardiovascular diseases include therapeutics that interfere with ABC transporter function. As described earlier, ABC transporters can be found at the BBB, in membranes of intracellular compartments, and in the choroid plexus. This complex system of membrane-bound “locks and elevators” is highly vulnerable to interference. The large number of patients with obesity and high blood pressure is not limited to the elderly, but more importantly begins to rise even in the younger population of <40 years of age. These patients will be treated with beta-blockers, anti-hyperlipidemic drugs and calcium antagonists for many years before they eventually reach the age at which AD and other proteopathies become increasingly common. There are several questions arising from this scenario, for example: 1) Does a reduction of ABC transporter function over a long period of time put the patients at risk for developing neurodegenerative diseases? 2) If so, how much reduction in transporter function is required to significantly raise the risk − 10%, 5%, 0.5%, or less? 3) What types of mechanisms play a role therein? 4) What types of toxic agents harm cerebral homeostasis – primarily intracerebral or primarily extracerebral (exogenous)? 5) How can we assess the risk of certain therapeutic drugs? and 6) Is stem cell maintenance altered by agents that affect transporters?

Environmental and genetic factors have been discussed as causes for late-onset AD. Genetic disposition and diet also influence ABC transporter function. A large number of polymorphisms in coding and non-coding regions of ABC transporter genes have been described that affect drug clearance [85–87].

AD RISK ASSESSMENT USING SYSTEMS BIOLOGY COMPUTATIONAL MODELING

A basic problem is that the many risk factors for complex diseases are poorly understood, let alone the way in which they interact [88]. The very fact that there are multiple drivers for these conditions suggests that a reductionist approach focusing on individual entities in isolation is no longer appropriate, and may even be misleading. Investigating the possible combinations by trial and error in humans is onerous, but feasible, when only two components are considered. However, it becomes extremely complicated with three components, and well nigh impossible with four or more. Such circumstances will require a systems approach and the use of sophisticated and progressively refined models. Systems Biology, therefore, promises to assist in the development of more specific compounds and in the identification of optimal drug targets on the basis of their importance as key ‘nodes’ within an overall network, rather than on the basis of their properties as isolated components.

Only very recently have systems biology approaches been applied towards elucidating the dynamics underlying complex diseases such as AD. The systems biology approach includes mathematical modeling of cellular biochemical reaction networks and the investigation of their properties through computer simulations [89, 90]. This strategy allows us to explore the structure and function of networks using methods from dynamical systems theory.

A number of different factors or presumed ‘drivers’ that are involved in AD pathophysiology can be included in these analyses: Aβ, tau, apoE, inflammation, oxidative stress, protein degradation, and protein transport systems. Since aging is the most important risk factor for AD, and aging is associated with multiple changes that might influence the probability of developing AD, a systems biological approach is particularly valuable in this instance. Computer simulations enable fast and large-scale quantitative predictions of networks dynamics. An important application is to understand disease mechanisms and to support the development of therapies. Mathematical modeling of AD has, so far, focused on the kinetics of Aβ aggregation, a downstream consequence of the decision between the amyloidogenic and non-amyloidogenic pathways. Up to now, systems biology has been one-directional, aiming at the mathematical modeling of biochemical reactions. Only recently has feedback to the wet-lab, in the form of quantitative predictions of pathway dynamics, been possible. A systems biology approach supports not only the formulation and testing of hypotheses, but could also form the basis for a risk assessment of ABC transporter involvement in AD mouse models. The phenomena described here are of a dynamic nature, most likely involving nonlinear feedback mechanisms. In the long run, once the relevant cellular components are identified and characterized at the molecular level, it will be necessary to investigate dynamic features for an understanding of transport mechanisms. A systems biology approach does, however, require quantitative time-series data, which for many omics-type technologies remains a challenge to this day.

Clearance kinetics and toxicity evaluation of various compounds, pharmaceuticals and peptides can be assessed in cell culture experiments and animal models. These data help to evaluate the risk to prevent side effects or other diseases. Risk assessment in AD research can be supported with systems biology methods. Systems biology aims to unravel the complex dynamics of cellular biochemical reaction networks by mathematical modeling and computer simulations. These simulations can help to analyze interaction networks with respect to internal critical events or to the impact of external influences. Aβ aggregation and fibrillization have been extensively analyzed in silico and in vivo with respect to different APP mutations [91, 92]. Computational models beyond the scope of Aβ aggregation are rare.

Simplified interaction schemes (Fig. 3) and quantitative data from different ABC transporter knockout mouse models can be used to describe the chemical kinetics of Aβ aggregation and its clearance through the BBB in mathematical models. The resulting differential equations describe the interaction between Aβ moieties and different ABC transporters, and allow predictions of common regulation mechanisms. Knowledge and quantification of toxic aggregates, e.g. Aβ*56, can be included and used to assess critical constellations [55]. The application of computer simulations helps to predict the influence of drug interactions within the network. Due to the fact that processes that underlie AD are of a nonlinear dynamic nature, different strategies to those from genomics and bioinformatics are required. Systems biology has emerged as a paradigm to address such phenomena, but relies on accurate, quantitative, sufficiently rich time series. The systems approach investigates the dynamic behavior of a system through systematic perturbations. Given that an appropriate set of stimulus-response data is available, methods from dynamical systems theory allow the identification of mathematical models from time-course experiments. These computational techniques support the formulation of hypotheses about alternative network structures and through simulation studies, metabolic control analysis, sensitivity and bifurcation analysis, key elements of a network can be investigated with respect to their role within the overall network [93, 94]. Systematic perturbation experiments are necessary if the underlying processes involved feedback mechanisms to realize a robust or adaptive system. Apart from the study of intracellular data, a mathematical approach could also be useful for phenomenological data, capturing relationships between conditions and helping extract rules for AD risk assessment. This area of research is in its infancy, but given the increasing recognition and success of such approaches in the study of metabolic and signal transduction pathways, the time has come to strengthen interdisciplinary links between the medical and clinical domain with systems biology. The establishment of interaction maps and models of AD-related pathology with a correlation to the inhibition of ABC transporter function, especially those transporters that exhibit BBB excretion function (e.g. P-gp), will help us to evaluate clinical information of different drug therapies.

Fig. (3).

Schematic model for the mathematical evaluation of clearance, degradation, and aggregation of Aβ involving ABC transporters at the BBB. Aβ aggregates into oligomers, protofibrils and fibrils. The kinetics of aggregation and disaggregation vary, thus the timepoints of plaque development and the onset of neurotoxic effects are highly dependent on these factors. The aggregates and monomers can be degraded by different proteases. Additionally, Aβ monomers and small Aβ multimers are transported by BBB transporters. The excretion capacity and its velocity are dependent on the charge, size and conformation of the Aβ peptides. Charge changes as known from Dutch-Aβ (E22Q) lead to hampered transport. After several months, the restricted clearance of Dutch-Aβ results in deposition at the BBB. Intracerebral toxic levels of metabolites and peptide oligomers impair the normal function of neurons and disturb the maintenance of NSPCs. Importantly, long-term inhibition of the ABC transporter excretion function at the BBB by therapeuticals (e.g. statins and beta-blockers) could put the patients at higher risk for developing neurodegenerative diseases. The knowledge of clearance, degradation, aggregation and excretion kinetics as well as the amount of a defined Aβ moiety needed to exert toxicity for specific neuronal or stem cell populations and its clinical outcome, can be used to calculate individual risk profiles.

CONCLUSIONS

Given that most idiopathic AD, which accounts for the vast majority of AD cases, cannot be explained simply by increased Aβ production [52], it is important to elucidate fully the mechanisms whereby the concentrations of AD-related proteins are regulated in the brain. To this end, a detailed understanding of the molecular and genetic basis of ABC transporter function for clearance mechanisms at the BBB and neural stem cell maintenance holds at least a part of the key to understanding the pathogenesis of AD. Using model organisms [95], the kinetics of Aβ transport/exchange between blood and brain can be computed and matched for healthy and diseased persons. Furthermore, these systems-biology computational models can be used to integrate known interactions of currently used therapeutics with the excretion function at the BBB. It is of emerging importance to investigate the effect of long-term pharmacological treatment of systemic diseases on the removal of APP-associated peptides, since some of these cleavage products have been noted for their neuronal toxicity in vitro and in vivo [55, 91, 92]. ABC transporters may also play a crucial role in NSPC self-renewal by preventing cellular differentiation, and hence could be vital in future stem cell-based therapies for neurodegenerative disorders such as AD and PD.