Delivery of therapeutic agents to the central nervous system: the problems and the possibilities

Abstract

The presence of a blood-brain barrier (BBB) and a blood-cerebrospinal fluid barrier presents a huge challenge for effective delivery of therapeutics to the central nervous system (CNS). Many potential drugs, which are effective at their site of action, have failed and have been discarded during their development for clinical use due to a failure to deliver them in sufficient quantity to the CNS. In consequence, many diseases of the CNS are undertreated. In recent years, it has become clear that the blood-CNS barriers are not only anatomical barriers to the free movement of solutes between blood and brain but also transport and metabolic barriers. The cell association, sometimes called the neurovascular unit, constitutes the BBB and is now appreciated to be a complex group of interacting cells, which in combination induce the formation of a BBB. The various strategies available and under development for enhancing drug delivery to the CNS are reviewed.

Introduction

All animals with a complex nervous system require a blood-brain barrier (BBB). The BBB allows the creation of a unique extracellular fluid environment within the central nervous system (CNS) whose composition can, as a consequence, be precisely controlled. The extracellular fluid compartments of the CNS comprise the brain and spinal cord parenchymal interstitial fluid (ISF) and the cerebrospinal fluid (CSF), contained within the ventricles of the brain and the cerebral and spinal subarachnoid spaces. The structural BBB is created by the cerebral endothelial cells forming the capillaries of the brain and spinal cord (Fig. 1). The endothelial cells at their adjacent margins form tight junctions (zona occludens [ZO]; Brightman & Reese, 1969), produced by the interaction of several transmembrane proteins that project into and seal the paracellular pathway. The interaction of these junctional proteins, particularly occludin and claudin, is complex and effectively blocks an aqueous route of free diffusion for polar solutes from blood along these potential paracellular pathways and thus denies these solutes free access to brain interstitial (extracellular) fluid. The molecular structure and function of the BBB junctional proteins is beyond the scope of this review, but several recent reviews exist (Morita et al., 1999, Kniesel & Wolburg, 2000, Wolburg et al., 2001, Bauer et al., 2004, Hamm et al., 2004). The impediment to free diffusion is, of course, bidirectional and therefore does not allow a free diffusional movement of polar solutes out of the CNS. Because the tight junctions effectively seal off the brain to polar solutes, the endothelial cells are required to maintain a high level of expression of transport proteins for essential polar metabolites such as glucose and amino acids to facilitate their entry into brain (Begley & Brightman, 2003). Thus, the tight junctions between the endothelial cells form an efficient gate in the paracellular pathway, preventing the diffusional entry of polar solutes to the brain via this route.

Electron microscopic studies of the BBB suggest a lower incidence of observable endocytic profiles in these endothelial cells compared with peripheral endothelial cells. However, transcytosis involving vesicular transport across the BBB is a significant factor in the BBB transport of many macromolecules, such as peptides and proteins, and both receptor-mediated transcytosis (RMT) and nonspecific absorptive-mediated transcytosis (AMT) pathways exist across the cerebral endothelium (Begley & Brightman, 2003).

The endothelial cells forming the BBB also exhibit a polarized expression of transport proteins in the luminal and abluminal membranes of the endothelial cells, with some transporters expressed exclusively in one of these interfacial membranes and some in the other, whereas some are inserted into both membranes (Betz et al., 1980, Begley, 1996, Mertsch & Maas, 2002, Begley & Brightman, 2003). As some transporters are unidirectional and some bidirectional in their transport of solutes across the cell membrane, this polarization means that some solutes can be preferentially transported into the brain and some out of the brain and that the transport of some solutes can be facilitated in either direction depending on whether the concentration gradient across the BBB is directed into, or out of, the CNS. This latter aspect can become important for some drugs with affinity for BBB transporters, when after systemic administration the pharmacokinetic profile can cause the concentration gradient to reverse across the BBB. It is thought that the formation of tight junctions between the endothelial cells may also act as a fence in the cell membrane preventing both transport proteins and lipid rafts in the membrane from exchanging between the luminal and the abluminal membrane domains and thus preserving the polarity of the BBB. Potential routes across the BBB for drugs and other solutes are shown in Fig. 2.

Tight junction formation, the polar expression of transport proteins and a full differentiation of the cerebral endothelium, appears to be induced by a close association between the endothelial cells, the adjacent pericytes, and the end-feet of astroglia whose cell bodies lie deeper in the brain parenchyma (Kacem et al., 1998, Dore-Duffy, 2003). The astrocytic end-feet form a network surrounding the abluminal surface of the cerebral endothelial cells, with only the extracellular matrix (basement membrane) separating the cells. It is thought that this close association of the endothelial cells and the astrocytes in particular are responsible for inducing BBB properties and differentiation in the cerebral endothelial cells. There is still much debate about the factors involved which induce this differentiation, but it is likely that they are multiple and some are soluble and some depend on cell-to-cell contact involving molecular handshakes; thus, the cells within the association in turn influence each other. Nerve endings also terminate against the abluminal membrane of the BBB endothelial cells and may influence BBB differentiation and permeability (Rennels et al., 1983). In addition, the endothelial cells, the pericytes, and the astrocytes contribute to the proteins of the extracellular matrix and this structure in turn influences the behavior and differentiation of the cells forming the neurovascular unit (Abbott, 2002). The extracellular matrix immediately surrounding the cerebral endothelial cells and the pericytes is distinct in that it contains laminins 8 and 10, whereas the extracellular matrix of the brain parenchyma contains laminins 1 and 2 (Sixt et al., 2001). Perivascular macrophages and microglia derived from blood macrophages may also form a significant part of the BBB neurovascular unit and contribute differentiating and modulatory signals (Zenker et al., 2003).

There is also the analogous blood-CSF barrier formed by the epithelia of the choroid plexuses (Wolburg et al., 2001) and around the other circumventricular organs (CVO). The capillaries in the choroid plexus and the CVO do not form tight junctions and the blood vessels in these structures are freely permeable. Indeed, the CVO possess specialized structures in the capillary wall termed fenestrae where the cytoplasm is attenuated and only the plasma membrane(s) form the diffusional barrier (Prestcott & Brightman, 1998; see Fig. 3). In the choroid plexus, the epithelial (ependymal) cells facing the CSF form tight junctions between adjacent cells, thus preventing a free diffusional pathway across this structure. Similarly, the other CVO are surrounded by ependymal cells, which from tight junctions and prevent diffusion of blood-borne solutes further into the brain. Normally, ependymal cells do not form tight junctions elsewhere in the brain. The origin and the mechanisms of the differentiating factors bringing about tight junction formation in the choroid plexus and in ependymal cells near the CVO are quite unknown. The function of the choroid plexus is to secrete fresh CSF and to regulate the movement of solutes from blood to CSF and vice versa. Hence, the plexus epithelial membrane is also polarized in terms of the transporters expressed in the apical and basolateral membranes. The CVO probably form “windows” in the brain so that a restricted volume of brain extracellular fluid is allowed to rapidly equilibrate with plasma (further diffusion within the brain is limited by the tight junctions in the surrounding ependyma). Dendritic processes and receptors on neurons within this limited area can then interact with these blood-borne solutes and their nervous activity can be modulated. These CVO-proximate neurons can then synaptically influence and initiate quite distant events elsewhere in the CNS depending on where their axons terminate. This function of the CVO becomes vital because of the creation of the BBB, and without their monitoring activity, the brain could not rapidly respond to many blood-borne signals. Some of the CVO are termed neurohemal organs and are specialized to allow neurosecretion in the opposite direction from the brain to blood (Fig. 3). Examples of neurohemal organs are the median eminence, the posterior pituitary, and the pineal, which have no BBB and allow neurohormones to easily escape via their fenestrated and open capillaries into the systemic circulation to exert somatic endocrine functions.

The cells forming the avascular arachnoid membrane, which envelopes the whole CNS, also possess tight junctions that effectively seal the paracellular diffusional pathway between these cells.

Lipophilic solutes can diffuse through the endothelial cell membrane and enter the CNS passively (Levin, 1980). There is well-established relationship between lipid solubility, either calculated or determined as an oil-water partition coefficient, with brain penetration, which increases with increasing lipid solubility. However, some lipid-soluble molecules do not enter the brain as readily as their lipophilicity solubility might suggest (Fig. 4). These substances and many of their metabolites are removed from the brain and the cerebral endothelium by active efflux transporters, which hydrolyze ATP and can move their substrates against a concentration gradient from blood to brain (Begley, 2004a, Begley, 2004b). These active transporters are generally called ATP-binding cassette (ABC) transporters.

The ability of the BBB, the choroid plexus, and the pericytes to transform and detoxify many substances entering the CNS has probably been underestimated in the past, as the transforming enzyme activity of the brain as a whole has usually been considered, which is low, rather than the activity of specific cell types or associations. For several hydrolyzing and conjugating enzymes, the enzyme activity in the choroid plexus per unit weight of tissue is similar to that in the liver (Minn et al., 2000).

Thus, the function of the BBB is essentially 2-fold. It enables the creation of a separate and extremely stable intracerebral extracellular fluid compartment consisting of the CSF and the brain ISF, the composition of which can be maintained distinct from the somatic extracellular fluid. Within these protected compartments, the composition of the extracellular fluid can be precisely regulated in terms of solute concentrations (Table 1). This stability is essential as the CNS relies on accurate synaptic transmission and inhibition, and spatial and temporal summation, to perform its complex integrative functions. Unless the synapse can operate against an extremely stable background, accurate synaptic transmission and nervous integration becomes impossible (Begley & Brightman, 2003). The somatic extracellular fluid contains many potential neurotransmitters and other neuroactive substances whose concentrations in this fluid may vary widely within short periods of time. The CNS could not tolerate, and continue to function, against such a background of significant variation in the concentrations of neuroactive substances that occurs in the general extracellular fluid on a regular basis. Amino acids that are present in blood in high concentration (e.g., glycine, glutamic acid, and aspartic acid) are potent excitatory neurotransmitters; thus, their background concentration in brain extracellular fluid must be maintained stable at very constant levels.

Secondly, the BBB has a neuroprotective function. In a highly complex tissue such as the CNS, where neuronal cell division is either absent or a rare event, any acceleration in cell death and neuronal attrition will cause premature degenerative disease and pathology. Many potentially neurotoxic substances are being continuously ingested in the diet or are generated by metabolism. The BBB is therefore crucial in limiting the access of these potentially damaging xenobiotics and metabolites to the CNS by either blocking their entry or actively removing them from brain via the ABC transporters (Begley & Brightman, 2003, Begley, 2004a, Begley, 2004b).

The BBB clearly changes in several brain diseases and a variety of pathological processes may either alter the quality of the barrier or contribute to the development of the disease process (Neuwelt, 2004). Examples of BBB dysfunction in disease states are defective transport of amyloid-β by the BBB in Alzheimer's disease (Zlokovic, 2004); leptin resistance in obesity, where leptin feedback is impaired as the result of reduced leptin transport across the BBB (Banks & Lebel, 2002); and opening of the barrier in active CNS lesions in multiple sclerosis (Werring et al., 2000) and in brain tumors, where the effects on the cerebral endothelium appear to be varied within the tumor type; it may appear normal and continuous with tight junctions or remain continuous but develop fenestrations or become discontinuous, with or without the development of fenestrations (Schlageter et al., 1999). Different regions within the same tumor may indeed show markedly different changes in microvessel morphology.

The robustness of the BBB may decline with age, although few studies are available (Preston, 2001). Tight junction integrity appears to be maintained with age and a study in aged Fischer 334 rats indicates that P-glycoprotein (Pgp) expression in the BBB is maintained (Warrington et al., 2004). However, well-documented changes in drug pharmacokinetics with age may well alter brain penetration of many drugs and enhance drug-drug interactions.