Loop diuretics as potential CNS therapeutic agents: Limitations by active efflux transport at the blood-brain barrier

Abstract

Brain injury occurs in many diseases and is associated with marked morbidity and mortality. One of the worst brain diseases is stroke, which is the second leading cause of death worldwide and third leading in the United States after cardiovascular disease and cancer. Currently, the only FDA approved drug for the treatment of stroke is tissue plasminogen activator, which has many limitations. Thus, there is a need for development of new therapeutic agents for the treatment of stroke as well as other central nervous system (CNS) disorders associated with injury, including traumatic brain injury, chronic pain, epilepsy, and neurotoxicity. One of the major medical problems in brain injury is the cellular and interstitial fluid edema that arises in part due to activation of the sodium potassium chloride co-transporter! (NKCCl). This transporter is located throughout the CNS at brain capillaries (i.e., blood-brain barrier; BBB), glial cells and neurons. Upon activation, it leads to net movement ofNa+-K+-zcr ions into brain and brain cells, which also brings in water. The resulting elevated intracranial pressure collapses blood vessels and compromises blood flow leading to worsened cerebral ischemia and neuronal death. An agent which blocks cerebral edema and cell swelling under conditions of neural injury would be a major advance. Bumetanide and other loop diuretics are potent inhibitors of NKCC 1 and have the potential to reduce cerebral edema, infarct volume, and neural death associated with NKCC 1 activation. However, most studies have relied on intracerebral drug administration, which is highly invasive, or very high dose peripheral administration. Little is known of the extent to which bumetanide and other loop diuretics cross into brain and reach active concentrations within the CNS. Based upon their structures, we hypothesized that bumetanide and other loop diuretics would show extremely limited distribution to brain and would be kept out by active efflux transport at the BBB. To test this hypothesis, we measured the rate of eHJ-bumetanide uptake into brain using the in situ brain perfusion technique. eH]Bumetanide as well as other loop diuretics were found to have very low permeability at the BBB. Brain bumetanide uptake was limited by high plasma protein binding as well as a low capacity BBB saturable influx process with properties of an organic anion transporting polypeptide ( oatp ), including sensitivity to inhibition by digoxin. In vivo pharmacokinetic analysis in rats demonstrated that bumetanide distribution to brain was limited (brain/serum concentration ratio= <2%) and that at steady state the free bumetanide concentration in brain and cerebrospinal fluid were only 8-20% of that of serum, suggesting net active efflux transport. Further, even at the highest doses (30 mg.kg i.v.), free bumetanide concentration in brain equaled or exceeded the Ki ofNKCCl for only 30-60 min. The results demonstrated a marked delivery problem ifbumetanide were to be considered for future neural therapy. Delivery was limited even in the presence of permanent middle cerebral artery occlusion, where the integrity of the BBB may be partially compromised. Bumetanide, though anionic at physiological pH, is lipophilic and would be predicted to readily cross lipid membranes. Therefore, if a CNS delivery problem exists, it may lie in active BBB efflux. Bumetanide transport out of brain was measured with the brain efflux index method. With this approach, bumetanide clearance from brain was found to exceed uptake by 3 fold and to be saturable upon addition of elevated bumetanide concentration, consistent with the presence of active efflux transport. Inhibitor analysis provided evidence for roles of oatp2, organic acid transporter 3 (OAT3) and possibly multidrug resistance protein 4 (MRP4). Rat oatp2 and OAT3 were confirmed to have the ability to transport bumetanide using Xenopus transportocytes. The results suggest that BBB active efflux transport markedly restricts bumetanide exposure to the CNS and may limit use of this agent as a potential therapy. This problem may be overcome by future development of new NKCC 1 inhibitors which are not substrates for the BBB efflux carriers and potentially may target brain NKCCI over renal NKCC2. Alternatively, it may be wise to explore bumetanide bio-conjugates with BBB drug delivery vectors that preferentially deliver bumetanide to brain, overcoming the efflux transporters and simultaneously limiting NKCC2 inhibition and fluid diuresis in the kidney.