Allopregnanolone as regenerative therapeutic for Alzheimer's disease: translational development and clinical promise

Abstract

Herein, we review a translational development plan to advance allopregnanolone to the clinic as a regenerative therapeutic for neurodegenerative diseases, in particular Alzheimer's. Allopregnanolone, an endogenous neurosteroid that declines with age and neurodegenerative disease, was exogenously administered and assessed for safety and efficacy to promote neuro-regeneration, cognitive function and reduction of Alzheimer's pathology. Allopregnanolone-induced neurogenesis correlated with restoration of learning and memory function in a mouse model of Alzheimer's disease and was comparably efficacious in aged normal mice. Critical to success was a dosing and treatment regimen that was consistent with the temporal requirements of systems biology of regeneration in brain. A treatment regimen that adhered to regenerative requirements of brain was also efficacious in reducing Alzheimer's pathology. With an optimized dosing and treatment regimen, chronic allopregnanolone administration promoted neurogenesis, oligodendrogenesis, reduced neuroinflammation and beta-amyloid burden while increasing markers of white matter generation and cholesterol homeostasis. Allopregnanolone meets three of the four drug-like physicochemical properties described by Lipinski's rule that predict the success rate of drugs in development for clinical trials. Pharmacokinetic and pharmacodynamic outcomes, securing GMP material, development of clinically translatable formulations and acquiring regulatory approval are discussed. Investigation of allopregnanolone as a regenerative therapeutic has provided key insights into mechanistic targets for neurogenesis and disease modification, dosing requirements, optimal treatment regimen, route of administration and the appropriate formulation necessary to advance to proof of concept clinical studies to determine efficacy of allopregnanolone as a regenerative and disease modifying therapeutic for Alzheimer's disease.

Keywords: 2-hydroxypropyl β-cyclodextrin; 3-hydroxy-3-methyl-glutaryl-CoA-reductase; 3xTgAD; 3α-HSD; 3α-hydroxysteroid dehydrogenase; 3β-HSD; 3β-hydroxysteroid dehydrogenase; 5α-DHP; 5α-R; 5α-dihydroprogesterone; 5α-reductase; ABAD; AD; ANT; AUC; Allo; Alzheimer's disease; Aβ; CMC; CYP3A; CYP450scc; EMA; European Medicines Agency; FDA; GLP; Good Laboratory Practices; Good Manufacturing Practices; HMG-CoA-R; HβCD; IM; IN; IND; IV; LXR; MED; MTD; Neurogenesis; Neurosteroid; PKA; PKCɛ; PXR; Pharmacodynamics; Pharmacokinetics; Regenerative medicine; SBEβCD; SC; SGZ; SVZ; StAR; TD; TSPO; Treatment regimen; United States Food and Drug Administration; VDAC; adenine nucleotide transporter protein; allopregnanolone; amyloid-beta binding alcohol dehydrogenase; area under the curve; cGMP; chemistry, manufacturing, and controls; cytochrome P450 3A; cytochrome P450 side-chain cleavage; intramuscular; intranasal; intravenous; investigational new drug; liver-X-receptor; maximum tolerated dose; minimal effective dose; non-transgenic; nonTg; pregnane-X-receptor; protein kinase A; protein kinase C epsilon; steroidogenic acute regulatory protein; subcutaneous; subgranular zone; subventricular zone; sulfobutyl ether β-cyclodextrin; transdermal; translocator protein; triple transgenic mouse model of AD; voltage-dependent anion channel protein; β-amyloid.

Fig. 1.

Optimal allopregnanolone (Allo) therapeutic regimen applied to therapeutic window to protect from Alzheimer’s disease (AD). Endogenous Allo (blue colored line) is produced both locally within the brain and from peripheral sources, including maternal circulation pre-natal, and post-natal and adult from the gonads and adrenal cortex in both male and female. In this generalized diagram, Allo and neurogenesis (red colored line) correlate throughout the life span and in disease. Most notable is the elevated level of Allo during fetal brain development that peaks and then sharply declines before birth (Hirst et al., 2009; Nguyen et al., 2003; Westcott et al., 2008; Luisi et al., 2000). Not depicted are the cyclic fluctuations in progesterone and closely follows progesterone levels in vivo. Also not depicted are the fluctuations in Allo during puberty. In adults, levels of neurosteroids including Allo gradually decline with advanced age and can be associated with chronological age or more specifically by what is often called endocrine age which considers the unique or categorical endocrine system status. In the presymptomatic and mild cognitive impairment (MCI) stages before AD, circulating blood and brain cortex Allo levels sharply decline correlating with onset of AD (Marx et al., 2006; Naylor et al., 2010). In most animal models of AD, the decline in neurogenesis correlates with temporal progression of AD pathology (Lazarov and Marr, 2010). Human neurogenesis declines with age and doublecortin-positive cells decreased approximately ten-fold from 0 to 100 years of age (Knoth et al., 2010) and human data roughly parallel the rate of neurogenesis in aging rodent models. Another method in which post-mortem brain genomic DNA samples containing unique radiometric signatures due to a spike in atmospheric ¹⁴C levels from global nuclear bomb fallout were used to determine neuronal and nonneuronal cell birthdates (Spalding et al., 2013). The results of these studies revealed that the age of the ¹⁴C-labeled DNA within adult hippocampal neurons were incorporated during adulthood. Each year approximately 1.75 percent of neurons turned over within the self-renewing fraction with only a modest decline during aging. A best-fit scenario model predicted that approximately 35 percent of the hippocampal cells were cycling corresponding to slightly less than the proportion that constitute the entire dentate gyrus region. Spalding and colleagues estimated that the hippocampal dentate gyrus of human brain produces around 700 new neurons per day. At this rate, enough neurons could be regenerated in the hippocampus to replace the dentate gyrus region over the human lifespan suggesting that the potential amount of neurogenesis is very substantial. Interestingly, the rate of neurogenesis in adult humans was similar by comparison to the rate determined in adult rodents but the rate did not decline as steeply as is known in rodents suggesting that humans may rely on neurogenesis more during the aging process. Chronic stress is associated with depressed responsiveness to constant levels of Allo and in correlation, chronic stress inhibits neurogenesis and exacerbates AD and loss of memory (not depicted) (Bengtsson et al., 2012, 2012). Further, neurodegeneration, a hallmark of AD negatively correlates with levels of Allo (also not depicted). The graphical representation suggests a prominent role of Allo in the neuroendocrine axis, its role both in neurogenesis and AD progression.

Fig. 2.

Neurosteroid biosynthesis is a multi-step enzymatic pathway. Cholesterol homeostasis, recruitment to the mitochondrial compartment, enzymatic reduction and transport to neural and glial cells or their precursor cells require regulation of multiple mechanisms critical to Alzheimer’s disease treatment. Allo mechanisms of action promote neurogenesis, oligodendrogenesis, and on a systems-level inhibit excess inflammation and β-amyloidogenesis. In the central and peripheral nervous systems, neurosteroid synthesis occurs in myelinating glial cells, astrocytes, and several neuronal cell types including neural progenitors. Cholesterol is supplied to these cell types and presented to the mitochondria. The mitochondrial membrane translocator protein (TSPO) controls the uptake of cholesterol and the synthesis of neuroactive steroids (Rupprecht et al., 2010). TSPO-associated proteins form a cholesterol transport pore in the mitochondrial inner membrane that include the steroidogenic acute regulatory protein (StAR), voltage-dependent anion channel protein (VDAC), and adenine nucleotide transporter protein (ANT). The mitochondrial pore transports cholesterol to the mitochondrial matrix to be converted into pregnenolone by the cytochrome P450 side-chain cleavage (CYP450scc) enzyme (Liu et al., 2006). Pregnenolone diffuses to the cytosol and then is converted to progesterone by the 3β-hydroxysteroid dehydrogenase (3β-HSD) enzyme. Peripherally derived progesterone and Allo cross the blood-brain barrier to also contribute to the neurosteroid concentration. Allo is synthesized from progesterone in two enzymatic steps by 5α-reductase (5α-R) type-I and 3α-hydroxysteroid dehydrogenase (3α-HSD) in the brain de novo (Mellon, 2007; Mellon et al., 2001). The rate-limiting step in neurosteroidogenesis is the reduction of progesterone to 5α-dihydroprogesterone (5α-DHP) by 5α-R. Subsequently, 3α-HSD catalyzes conversion of 5α-DHP into Allo. Amyloid beta-binding alcohol dehydrogenase or (also known as ABAD; SCHAD; 17βHSD10) is an enzyme that associates with mitochondria and facilitates back conversion of Allo to 5α-DHP (He et al., 2005; Yang et al., 2005). Additionally, anti-depressants such as fluoxetine are pro-neurogenic and have been shown to increase Allo production in the brain (Malberg et al., 2000; Uzunova et al., 2004, 2006). Endogenous Allo or optimal therapeutic dose of Allo with proper drug delivery strategy promote neurogenesis, oligodendrogenesis, and on a systems-level inhibit excess inflammation and β-amyloidogenesis.

Fig. 3.

The neurosteroid release rate pyramid describes the relationship between release rate of active neurosteroid and aqueous cyclodextrin vehicle. Molar complexation ratios greatly alter the release rate or dissolution profile of neurosteroids including allopregnanolone (Allo) as illustrated by the release rate pyramid. Typical cyclodextrin carrier formulations utilize the cyclodextrin such as hydroxypropyl beta-cyclodextrin, at 5–30% in aqueous solution such as water or normal saline with a maximal achievable cyclodextrin concentration between 45–60% weight to volume. Generally, a 1:1 stoichiometry inclusion complex of neurosteroid and a cyclodextrin exists at the apex of the pyramid hypothetically set at a threshold dose that requires maximal delivery rate of the neurosteroid. For Allo, the molar complexation ratio (suitable cyclodextrin:Allo) that is optimal for release rate is closer to 1:2. The same Allo dose would exhibit a slower release rate with a cyclodextrin:Allo of 1:10 complexation ratio since most of the Allo would be suspended in a depot formulation and less accessible for release and absorption into the biological system. Likewise, a complexation ratio that possesses cyclodextrin:Allo of 10:1 at the same dose of Allo would be readily soluble in an aqueous environment similar to the 1:10 suspension formulation that sits on the opposite side of the solubility spectrum. The 10:1 formulation would also be less accessible for absorption compared to the 1:1 formulation as the Allo is held within the overly abundant empty cyclodextrin molecules governed by the binding and dissociation rate. We hypothesize that the 10:1 formulation displays a re-uptake phenomenon whereby the neurosteroid is relatively less likely to be released into the biological system and therefore slower to release and interact with the receptor targets. A desired pharmacokinetic profile can be titrated by understanding the complexation ratio and release rate to optimize a measurable biological effect such as neurogenesis, anxiolytic activity, or sedation. For Allo, ataxia or sedation can be used as a measurable biomarker of target engagement and can establish the maximally tolerated dose in relation to formulation and delivery route. The lowest possible dose to experience sedation for example at 1:1 formulation would elicit little or no effect with 1:10 or 10:1 formulation complexes.

Fig. 4.

Neurosteroid pharmacokinetic profiles with consideration of formulation and route of administration. A. Pharmacokinetic profiles in blood plasma are generalized within the therapeutic range between the maximum tolerated dose (MTD) and the minimal effective dose (MED) to highlight the importance of formulation release rates and maximum exposure levels following a single equal neurosteroid bolus infusion dose. Neurosteroids readily cross the blood–brain barrier after absorption into the blood stream. The profiles: (1) Intravenous (IV) soluble dose (red colored line) – rapid release rate to reach T_max; relatively highest concentration C_max, rapid elimination rate; optimal for rapid cell signaling and non-genomic mechanisms within minutes and a 5–30 min distribution phase and rapid clearance. (2) Subcutaneous (SC)/intramuscular (IM) soluble dose (green colored line) – medium release; medium elimination; optimal for rapid cell signaling and non-genomic mechanisms; activates some slow mechanisms in distribution phase of 10–60 min range. (3) SC/IM suspension dose (blue colored line) – slow release; slow elimination; less optimal for rapid cell signaling, optimal for mechanisms that require slow or prolonged activation. (B) Intravenous (IV) soluble dose (red colored bar) is plotted on dual axes as the systemic amount of neurosteroid exposure or area under the curve (AUC) required to activate therapeutically relevant mechanisms and the duration of action. For example, a short burst of the neurosteroid Allo reaches MTD defined by level of conscious-sedation with relatively low exposure AUC via a bolus IV infusion of fully soluble neurosteroid. SC/IM soluble dose (green colored bar) depicts the AUC of neurosteroid that could be required for longer duration of action compared to IV. With soluble SC/IM injection, relatively less total AUC is required to activate the biological system compared to a hypothetical SC/IM suspension dose (blue bar) with relatively low solubility and slow release rate and minimal peak level at a given dose. Depending on the desired biological effect, each route has advantages and disadvantages that should be considered when developing a therapeutic strategy.