Blood–brain barrier permeability considerations for CNS-targeted compound library design

doi:10.1016/j.cbpa.2008.03.019

Current Opinion in Chemical Biology

Volume 12, Issue 3, June 2008, Pages 318-323

https://doi.org/10.1016/j.cbpa.2008.03.019 Get rights and content

A further refinement of the concept of drug-likeness is required for compound libraries intended for central nervous system (CNS) targets to account for the limitations imposed by blood–brain barrier permeability. This review describes criteria and processes that can be applied in the de novo design and assembly of libraries to increase the odds of compounds residing within CNS-accessible chemical space. A number of published examples where CNS activity and/or penetration characteristics have been a factor in library design are discussed.

Introduction

The fundamental philosophy underpinning early phase drug discovery underwent a dramatic reshaping in the early 1990s with the advent of parallel synthesis and high-throughput screening [1]. Accompanying advances in automation and informatics enabled the generation, management and screening of large expanses of new chemical space and seemed to offer a panacea for the demands of the expanding biological target landscape. The initial surge of excitement stemming from the potential of interlacing the ability to generate large libraries and screen them has however since been tempered by the emergence of important pragmatic considerations. The early enthusiasm and, to a certain extent, hubris fueled by the power of numbers has given way to a realization that library size must be balanced with compound quality, both in terms of the nature of the structures and their analytical fidelity. Consequently, the science of library design has evolved from one where chemical diversity and synthetic practicality were the primary drivers to a more holistic approach, incorporating both activity-based and property-based design elements [2^••]. Smaller, target-focused compound arrays which are preferably amenable to iterative optimization and incorporate drug-like and lead-like features better align with current drug discovery needs. A substantial body of literature has been devoted to the concept of lead-like and drug-like chemical space and its utilization in the assessment of virtual structures has become a routine precursor to the synthesis of libraries [3, 4, 5•, 6]. For targets residing in the central nervous system (CNS), the blood–brain barrier (BBB) represents an additional hurdle to be surmounted relative to peripheral targets [7]. CNS-accessible chemical space is recognized as representing a small subset of the synthetically accessible chenome [8••, 9]. Therefore, a further refinement of the concept of drug-likeness is required for CNS-targeted therapeutics. Surprisingly however, there are very few published examples where CNS penetration characteristics have been factored into library design. The purpose of this article is to: first, review select examples where BBB permeability has been a guiding factor in de novo design of libraries and second, provide recommendations for future design of CNS-permeable compound libraries.

Section snippets

Issues with BBB permeability in library design

The BBB is characterized by an extensive network of capillaries lined with endothelial cells that possess a number of unique morphological and biochemical features [10]. The continuous tight junctions between these cells tightly restrict paracellular diffusion of molecules. Active transport at the BBB by both uptake and efflux proteins and the presence of metabolizing enzymes conspire to present a formidable and complex biological barrier to small molecule entry into the brain. An issue that

BBB permeability-biased library design

A recent literature example describing of the use of PAMPA and IAM chromatography as tools in iterative library design is illustrated in Figure 3 [20]. A serendipitous observation that a diketopiperazine (DKP) degradation product 1, formed from a tripeptide, possessed good passive permeability in a BBMEC assay prompted the design and investigation of additional DKP libraries.

The first iteration of the DKP library contained 15 analogs prepared on solid phase in greater than 90% purity. The

CNS-activity-based and property-based library design

An impressive example of the incorporation of both CNS-activity-based and property-based design features into library design has been reported by researchers from Organon [22]. Analysis of fragments present in 373 CNS drugs indicate that aromatic and amino groups occur with high frequency, being found in 83% and 63% of CNS drugs, respectively. A 3042 compound library was designed, employing a set of 13 different scaffolds. Each scaffold possessed two sites of modification, and was selected to

Physicochemical descriptors in CNS-biased library design

In response to the paucity of available BBB permeability data, data mining approaches relying on reported pharmacological activity or disease indication as a surrogate gauge of CNS entry have been employed to develop predictive methods for CNS activity [23•, 24, 25, 26, 27]. An advantage of these methods is the large body of data housed in databases such as Comprehensive Medicinal Chemistry (CMC), MDL Drug Data Report (MDDR), and World Drug Index (WDI). However, the caveat that indication for a

Conclusions

The wealth of in silico and in vitro biological data generated within drug discovery programs has rendered activity-biased library design routine in nature. Whereas, the ability to validate target design hypothesis is readily enabled by high-throughput in vitro assays, in vitro measurement of BBB permeability potential remains limited by assay capacity. Consequently the philosophy exists within many research programs that BBB permeability is better addressed in the lead optimization phase at a

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as:

• of special interest
•• of outstanding interest

References (33)

C.E. Huwe
Synthetic library design
Drug Discov Today
(2006)
I. Muegge
Selection criteria for drug-like compounds
Med Res Rev
(2003)
G. Schneider et al.
Computer-based de novo design of drug-like molecules
Nat Rev Drug Discov
(2005)
Q. Wang et al.
Evaluation of the MDR-MDCK cell line as a permeability screen for the blood–brain barrier
Int J Pharmaceut
(2005)
T.J. Raub
P-glycoprotein recognition of substrates and circumvention through rational drug design
Mol Pharm
(2005)
N.H. Sookhee et al.
Mini review on molecular modeling of P-glycoprotein (Pgp)
Curr Topics Med Chem
(2007)
D. Barn et al.
Design and synthesis of a maximally diverse and druglike screening library using REM resin methodology
J Comb Chem
(2001)
Ajay et al.
Designing libraries with CNS activity
J Med Chem
(1999)
J.A. Gratton et al.
Molecular factors influencing drug transfer across the blood–brain barrier
J Pharm Pharmacol
(1997)
H.M. Geysen et al.
Combinatorial compound libraries for drug discovery: an ongoing challenge
Nat Rev Drug Discov
(2003)

C.A. Lipinski

Lead- and drug-like compounds: the rule-of-five revolution

Drug Discov Today: Technol

(2004)

M.M. Hann et al.

Pursuing the lead-likeness concept in pharmaceutical research

Curr Opin Chem Biol

(2004)

R.A. Goodnow et al.

Library design practices for success in lead generation with small molecule libraries

Comb Chem High Throughput Screen

(2003)

S.A. Hitchcock et al.

Structure–brain exposure relationships

J Med Chem

(2006)

H. Pajouhesh et al.

Medicinal chemical properties of successful central nervous system drugs

NeuroRx

(2005)

Cited by (84)

Synthesis and biological evaluation of novel <sup>18</sup>F-labeled 2,4-diaminopyrimidine derivatives for detection of ghrelin receptor in the brain
2024, Bioorganic and Medicinal Chemistry Letters
The ghrelin receptor (GHSR) is known to regulate various physiological processes including appetite, food intake, and growth hormone release. Its expression is mainly observed in the brain, pancreas, stomach, and intestine. However, the functions of the receptor have not been fully elucidated. GHSR imaging with positron emission tomography (PET) is expected to further understanding of the functions and pathologies of the receptor. In this study, we newly designed and synthesized diaminopyrimidine derivatives ([¹⁸F]BPP-1 and [¹⁸F]BPP-2) and evaluated their utility as novel PET probes targeting GHSR. In in vitro competitive binding assays, the binding affinity of BPP-2 for GHSR (K_i = 274 nM) was comparable to that of the diaminopyimidine lead compound Abb8a (K_i = 109 nM). In a biodistribution study using normal mice, [¹⁸F]BPP-2 displayed low uptake in the brain and moderate uptake in the pancreas, but high radioactivity accumulation in bone was observed due to its defluorination in vivo. Taken together, although further improvement of the pharmacokinetics is needed, the diaminopyrimidine scaffold has potential for the development of useful GHSR-targeting PET probes.
Synthesis and biological evaluation of indane-based fluorescent probes for detection of amyloid-β aggregates in Alzheimer's disease
2023, Bioorganic and Medicinal Chemistry
In this article, the development of fluorescent imaging probes for the detection of Alzheimer's disease (AD)-associated protein aggregates is described. Indane derivatives with a donor-π-acceptor (D-π-A) structure were designed and synthesized. The probes were evaluated for their ability to bind to β-amyloid (Aβ) protein aggregates, which are a key pathological hallmark of AD. The results showed that several probes exhibited significant changes in fluorescence intensity at wavelengths greater than 600 nm when they were bound to Aβ aggregates compared to the Aβ monomeric form. Among the tested probes, four D-π-A type indane derivatives showed promising binding selectivity to Aβ aggregates over non-specific proteins such as bovine serum albumin (BSA). The molecular docking study showed that our compounds were appropriately located along the Aβ fibril axis through the hydrophobic tunnel structure. Further analysis revealed that the most active compound having dimethylaminopyridyl group as an election donor and dicyano group as an electron acceptor could effectively stain Aβ plaques in brain tissue samples from AD transgenic mice. These findings suggest that our indane-based compounds have the potential to serve as fluorescent probes for the detection and monitoring of Aβ aggregation in AD.
Biodistribution and toxicity of innate defense regulator 1018 (IDR-1018)
2022, European Journal of Pharmaceutics and Biopharmaceutics
Innate defense regulators (IDRs) are synthetic host-defense peptides (HDPs) with broad-spectrum anti-infective properties, including immunomodulatory, anti-biofilm and direct antimicrobial activities. A lack of pharmacokinetic data about these peptides hinders their development and makes it challenging to fully understand how they work in vivo since their mechanism of action is dependent on tissue concentrations of the peptide. Here, we set out to define in detail the pharmacokinetics of a well-characterized IDR molecule, IDR-1018. To make the peptide traceable, it was radiolabeled with the long-lived gamma-emitting isotope gallium-67. After a series of bench-top characterizations, the radiotracer was administered to healthy mice intravenously (IV) or subcutaneously (SQ) at various dose levels (2.5–13 mg/kg). Nuclear imaging and ex-vivo biodistributions were used to quantify organ and tissue uptake of the radiotracer over time. When administered as an IV bolus, the distribution profile of the radiotracer changed as the dose was escalated. At 2.5 mg/kg, the peptide was well-tolerated, poorly circulated in the blood and was cleared predominantly by the reticuloendothelial system. Higher doses (7 and 13 mg/kg) as an IV bolus were almost immediately lethal due to respiratory arrest; significant lung uptake of the radiotracer was observed from nuclear scans of these animals, and histological examination found extensive damage to the pulmonary vasculature and alveoli. When administered SQ at a dose of 3 mg/kg, radiolabeled IDR-1018 was rapidly absorbed from the site of injection and predominately cleared renally. Apart from the SQ injection site, no other tissue had a concentration above the minimum inhibitory concentration that would enable this peptide to exert direct antimicrobial effects against most pathogenic bacteria. Tissue concentrations were sufficient, however, to disrupt microbial biofilms and alter the host immune response. Overall, this study demonstrated that the administration of synthetic IDR peptide in vivo is best suited to local administration which avoids some of the issues associated with peptide toxicity that are observed when administered systemically by IV injection, an issue that will have to be addressed through formulation.
In vitro blood brain barrier models: An overview
2022, Journal of Controlled Release
Citation Excerpt :
Molecules have a high solubility in water and permeability in the intestines because of their physical features. These qualities are commonly found in around 90% of orally-active medicinal drugs in Phase II clinical trials [135–138]. One of the earliest indicators of BBB permeability in diverse data sets has been the octanol/water partitioning coefficient (logPoct).
Understanding the composition and function of the blood brain barrier (BBB) enables the development of novel, innovative techniques for administering central nervous system (CNS) medications and technologies for improving the existing models. Scientific and methodological interest in the pathology of the BBB resulted in the formation of numerous in vitro BBB models. Once successfully studied and modelled, it would be a valuable tool for elucidating the mechanism of action of the CNS disorders prior to their manifestation and the pathogenic factors. Understanding the rationale behind the selection of the models as well as their working may enable the development of state-of-the-art drugs for treating and managing neurological diseases. Hence, to have realistic simulation of the BBB and test its drug permeability the microfluidics-based BBB-on-Chip model has been developed. To summarise, we aim to evaluate the advanced, newly developed and frequently used in vitro BBB models, thereby providing a brief overview of the components essential for in vitro BBB formation, the methods of chip fabrication and cell culturing, its applications and the recent advances in this technological field. This will be critical for developing CNS treatments with improved BBB penetrability and pharmacokinetic properties.
Nanostructured lipid base carrier for specific delivery of garlic oil through blood brain barrier against aggressiveness of glioma
2021, Journal of Drug Delivery Science and Technology
Glioblastoma, GBM, is a primary malignant brain tumor and is one of the most difficult-to-treat cancers. The inability of drugs to cross the blood brain barrier (BBB) is the main obstacle for treating brain cancers. The development of drug delivery systems which improve the permeation across the BBB is the focus of this research, specifically the benefits and use of nanostructured lipid carriers (NLC). The anti-inflammatory and anti-tumor activity of garlic oil have been widely studied. However, direct treatment with garlic oil is restricted due to its low bioavailability and the application of drug for brain cancer to improve the efficiency of garlic oil, a NLC was developed as a carrier of garlic oil. The garlic NLC was formulated using the hot high-pressure homogenization technique. The physicochemical properties of the garlic NLC were analyzed using dynamic light scattering. Size of the garlic NLC was 136.8 ± 0.56 nm and the zeta potentials of each formulation were −36.27 ± 0.96 mV and PDI 0.163 ± 0.03. The effect of garlic NLC on the cell viability of the glioma cell was determined using a MTT assay. The garlic NLC had higher efficiency towards inhibiting cell viability of glioma cells compared to free garlic oil. In addition, the effects of the garlic NLC on cell migration and cell invasion were assessed using a Boyden chamber assay. The migration and invasion abilities of garlic NLC treated cells were significantly decreased compared to the control and free garlic oil. Moreover, in vitro BBB was used to test permeation of garlic oil across BBB. Interestingly, garlic NLC could cross the BBB while the free garlic oil could not. Taken together, garlic NLC might be applicable as an alternative therapeutic treatment strategy for brain cancer.
Methods to optimize CNS exposure of drug candidates
2020, Bioorganic and Medicinal Chemistry Letters
Citation Excerpt :
Establishment of these assays early in the screening funnel allows medicinal chemists to triage potential new chemical entities (NCEs) before they can progress into the costlier in vivo assays and animal models. The pharmaceutical industry has learned to address the attrition that has been witnessed during human clinical trials by identifying areas for improvement, for example around pharmacokinetics and bioavailability during the 1990s20 and is continuing on such path through the adoption of pre-clinical screens to respond to potential inadequate brain penetration21 and challenges associated thereof. Since the discovery of the first drug efflux transporter more than 45 years ago, the role of these proteins in the BBB has been extensively studied.
Among the most devastating disorders of our time, neurologic and psychiatric diseases combine to cause more disability than any other disease area. One of the key objectives within the medicinal chemistry discipline is to design molecules that penetrate into the target tissue. The objective of tissue specificity can be to gain or restrict drug access to the compartment of interest. This article briefly reviews the progress of CNS drug discovery over the past few decades. Included are the most recent efforts to harness structural and physicochemical properties assessment coupled with the impact of efflux transporters in determining brain penetration and the translation from rodent to human brain tissue targeting.

View all citing articles on Scopus

View full text

Blood–brain barrier permeability considerations for CNS-targeted compound library design

Introduction

Section snippets

Issues with BBB permeability in library design

BBB permeability-biased library design

CNS-activity-based and property-based library design

Physicochemical descriptors in CNS-biased library design

Conclusions

References and recommended reading

Drug Discov Today

Med Res Rev

Nat Rev Drug Discov

Int J Pharmaceut

Mol Pharm

Curr Topics Med Chem

J Comb Chem

J Med Chem

J Pharm Pharmacol

Combinatorial compound libraries for drug discovery: an ongoing challenge

Nat Rev Drug Discov

Lead- and drug-like compounds: the rule-of-five revolution

Drug Discov Today: Technol

Pursuing the lead-likeness concept in pharmaceutical research

Curr Opin Chem Biol

Library design practices for success in lead generation with small molecule libraries

Comb Chem High Throughput Screen

Structure–brain exposure relationships

J Med Chem

Medicinal chemical properties of successful central nervous system drugs

NeuroRx