Flexibility in Drug Product Development: A Perspective.

The process of bringing a drug to market involves innumerable decisions to refine a concept into a final product. The final product goes through extensive research and development to meet the target product profile and to obtain a product that is manufacturable at scale. Historically, this process often feels inflexible and linear, as ideas and development paths are eliminated early on to allow focus on the workstream with the highest probability of success. Carrying multiple options early in development is both time-consuming and resource-intensive. Similarly, changing development pathways after significant investment carries a high "penalty of change" (PoC), which makes pivoting to a new concept late in development inhibitory. Can drug product (DP) development be made more flexible? The authors believe that combining a nonlinear DP development approach, leveraging state-of-the art data sciences, and using emerging process and measurement technologies will offer enhanced flexibility and should become the new normal. Through the use of iterative DP evaluation, "smart" clinical studies, artificial intelligence, novel characterization techniques, automation, and data collection/modeling/interpretation, it should be possible to significantly reduce the PoC during development. In this Perspective, a review of ideas/techniques along with supporting technologies that can be applied at each stage of DP development is shared. It is further discussed how these contribute to an improved and flexible DP development through the acceleration of the iterative build-measure-learn cycle in laboratories and clinical trials.

Analyzing the Potential Root Causes of Variability of Pharmacokinetics in Preclinical Species.

The purpose of this research was to assess variability in pharmacokinetic profiles (PK variability) in preclinical species and identify the risk factors associated with the properties of a drug molecule that contribute to the variability. Exposure data in mouse, rat, dog, and monkey for a total of 16,592 research compounds studied between 1999 and 2013 were included in the analysis. Both in vivo study parameters and in silico/experimental physicochemical properties of the molecules were analyzed. Areas under the plasma concentration vs time curves (AUC) were used to assess PK variability. PK variability was calculated as the ratio of the highest AUC within a defined set of AUC values (AUCmax) over the lowest AUC within that set (AUCmin). Both intra- and inter-animal variability were analyzed, with intra-animal exposures found to be more variable than inter-animal exposures. While several routes of administration were initially studied, the analysis was focused on the oral route, which corresponds to the large majority of data points and displays higher variability than the subcutaneous, intraperitoneal, or intravenous routes. The association between inter-animal PK variability and physical properties was studied, and low solubility, high administered dose, high preclinical dose number (PDo), and pH-dependent solubility were found to be associated with high variability in exposures. Permeability-as assessed by the measured permeability coefficient in the LLC-PK1 cell line-was also considered but appeared to only have a weak association with variability. Consistent with these findings, BCS class I and III compounds were found to be less prone to PK variability than BCS class II and IV compounds. A modest association of PK variability with clearance was observed while the association with bioavailability, a higher PK variability for compounds with lower bioavailability, appeared to be more pronounced. Finally, two case studies that highlight PK variability issues are described, and successful mitigation strategies are presented.

Preclinical dose number and its application in understanding drug absorption risk and formulation design for preclinical species.

In the drug discovery setting, the ability to rapidly identify drug absorption risk in preclinical species at high doses from easily measured physical properties is desired. This is due to the large number of molecules being evaluated and their high attrition rate, which make resource-intensive in vitro and in silico evaluation unattractive. High-dose in vivo data from rat, dog, and monkey are analyzed here, using a preclinical dose number (PDo) concept based on the dose number described by Amidon and other authors (Pharm. Res., 1993, 10, 264-270). PDo, as described in this article, is simply calculated as dose (mg/kg) divided by compound solubility in FaSSIF (mg/mL) and approximates the volume of biorelevant media per kilogram of animal that would be needed to fully dissolve the dose. High PDo values were found to be predictive of difficulty in achieving drug exposure (AUC)-dose proportionality in in vivo studies, as could be expected; however, this work analyzes a large data set (>900 data points) and provides quantitative guidance to identify drug absorption risk in preclinical species based on a single solubility measurement commonly carried out in drug discovery. Above the PDo values defined, >50% of all in vivo studies exhibited poor AUC-dose proportionality in rat, dog, and monkey, and these values can be utilized as general guidelines in discovery and early development to rapidly assess risk of solubility-limited absorption for a given compound. A preclinical dose number generated by biorelevant dilutions of formulated compounds (formulated PDo) was also evaluated and defines solubility targets predictive of suitable AUC-dose proportionality in formulation development efforts. Application of these guidelines can serve to efficiently identify compounds in discovery that are likely to present extreme challenges with respect to solubility-limited absorption in preclinical species as well as reduce the testing of poor formulations in vivo, which is a key ethical and resource matter.

Dose Number as a Tool to Guide Lead Optimization for Orally Bioavailable Compounds in Drug Discovery.

Small molecule developability challenges have been well documented over the last two decades. One of these critical developability parameters is aqueous solubility. In general, more soluble compounds have improved oral absorption. While enabling formulation technologies exist to improve bioperformance for low solubility compounds, these are often more complex, expensive, and challenging to scale up. Therefore, to avoid these development issues, medicinal chemists need tools to rapidly profile and improve the physicochemical properties of molecules during discovery. Dose number (Do) is a simple metric to predict whether a compound will be reasonably absorbed based on solubility at an expected clinical dose and represents a valuable parameter to the medicinal chemist defining a clinical candidate. The goal of this mini-Perspective is to present the background of the Do equation and how it can be effectively used to rapidly predict oral absorption potential for molecules in the discovery space.

Discovery of MK-4688: an Efficient Inhibitor of the HDM2-p53 Protein-Protein Interaction.

Identification of low-dose, low-molecular-weight, drug-like inhibitors of protein-protein interactions (PPIs) is a challenging area of research. Despite the challenges, the therapeutic potential of PPI inhibition has driven significant efforts toward this goal. Adding to recent success in this area, we describe herein our efforts to optimize a novel purine carboxylic acid-derived inhibitor of the HDM2-p53 PPI into a series of low-projected dose inhibitors with overall favorable pharmacokinetic and physical properties. Ultimately, a strategy focused on leveraging known binding hot spots coupled with biostructural information to guide the design of conformationally constrained analogs and a focus on efficiency metrics led to the discovery of MK-4688 (compound 56), a highly potent, selective, and low-molecular-weight inhibitor suitable for clinical investigation.

The Discovery of Two Novel Classes of 5,5-Bicyclic Nucleoside-Derived PRMT5 Inhibitors for the Treatment of Cancer.

Protein arginine methyltransferase 5 (PRMT5) is a type II arginine methyltransferase that catalyzes the post-translational symmetric dimethylation of protein substrates. PRMT5 plays a critical role in regulating biological processes including transcription, cell cycle progression, RNA splicing, and DNA repair. As such, dysregulation of PRMT5 activity is implicated in the development and progression of multiple cancers and is a target of growing clinical interest. Described herein are the structure-based drug designs, robust synthetic efforts, and lead optimization strategies toward the identification of two novel 5,5-fused bicyclic nucleoside-derived classes of potent and efficacious PRMT5 inhibitors. Utilization of compound docking and strain energy calculations inspired novel designs, and the development of flexible synthetic approaches enabled access to complex chemotypes with five contiguous stereocenters. Additional efforts in balancing bioavailability, solubility, potency, and CYP3A4 inhibition led to the identification of diverse lead compounds with favorable profiles, promising in vivo activity, and low human dose projections.

Reversible bridge-mediated excited-state symmetry breaking in stilbene-linked DNA dumbbells.

The excited-state behavior of synthetic DNA dumbbells possessing stilbenedicarboxamide (Sa) linkers separated by short A-tracts or alternating A-T base-pair sequences has been investigated by means of fluorescence and transient absorption spectroscopy. Electronic excitation of the Sa chromophores results in conversion of a locally excited state to a charge-separated state in which one Sa is reduced and the other is oxidized. This symmetry-breaking process occurs exclusively via a multistep mechanism-hole injection followed by hole transport and hole trapping-even at short distances. Rate constants for charge separation are strongly distance-dependent at short distances but become less so at longer distances. Disruption of the A-tract by inversion of a single A-T base pair results in a pronounced decrease in both the rate constant and efficiency of charge separation. Hole trapping by Sa is highly reversible, resulting in rapid charge recombination that occurs via the reverse of the charge separation process: hole detrapping, hole transport, and charge return to regenerate the locally excited Sa singlet state. These results differ in several significant respects from those previously reported for guanine or stilbenediether as hole traps. Neither charge separation nor charge recombination occur via a single-step superexchange mechanism, and hole trapping is slower and detrapping faster when Sa serves as the electron donor. Both the occurrence of symmetry breaking and reversible hole trapping by a shallow trap in a DNA-based system are without precedent.

Compatible injection and detection systems for studying the kinetics of excess electron transfer.

[reaction: see text] A design for fast kinetic studies of electron transfer in radical anions is reported. alpha-Hydroxy radicals formed by 355 nm laser flash photolysis of alpha-phenacyl alcohols are deprotonated under basic conditions to give ketyl radical anions that serve as electron injectors in inter- and intramolecular electron-transfer reactions. The 2,2-diphenylcyclopropyl group serves as a reporter. When an electron is injected and transferred such that spin character is adjacent to the reporter, cyclopropyl ring opening gives a readily detected diphenylalkyl radical.

Detailed study of precipitation of a poorly water soluble test compound using methodologies as in activity and solubility screening - mixing and automation effects.

Storage of pharmaceutical discovery compounds dissolved in dimethylsulfoxide (DMSO) is commonplace within industry. Often, the DMSO stock solution is added to an aqueous system (e.g. in bioassay or kinetic solubility testing)- since most test compounds are hydrophobic, precipitation could occur. Little is known about the factors affecting this precipitation process at the low (µM) concentrations used in screening analyses. Here, a poorly water soluble test compound (tolnaftate) was used to compare manual and automated pipetting, and explore the effect of mixing variables on precipitation. The amount of drug present in the supernatant after precipitation and centrifugation of the samples was quantified. An unusual result was obtained in three different laboratories: results of experiments performed initially were statistically significantly higher than those performed after a few days in the same lab. No significant differences were found between automated and manual pipetting, including in variability. Vortex mixing was found to give significantly lower supernatant amounts compared to milder mixing types. The mixing employed affects the particle growth of the precipitate. These findings are of relevance to discovery stage bioassay and kinetic solubility analyses.

Discovery of a novel ERK inhibitor with activity in models of acquired resistance to BRAF and MEK inhibitors.

The high frequency of activating RAS or BRAF mutations in cancer provides strong rationale for targeting the mitogen-activated protein kinase (MAPK) pathway. Selective BRAF and MAP-ERK kinase (MEK) inhibitors have shown clinical efficacy in patients with melanoma. However, the majority of responses are transient, and resistance is often associated with pathway reactivation of the extracellular signal-regulated kinase (ERK) signaling pathway. Here, we describe the identification and characterization of SCH772984, a novel and selective inhibitor of ERK1/2 that displays behaviors of both type I and type II kinase inhibitors. SCH772984 has nanomolar cellular potency in tumor cells with mutations in BRAF, NRAS, or KRAS and induces tumor regressions in xenograft models at tolerated doses. Importantly, SCH772984 effectively inhibited MAPK signaling and cell proliferation in BRAF or MEK inhibitor-resistant models as well as in tumor cells resistant to concurrent treatment with BRAF and MEK inhibitors. These data support the clinical development of ERK inhibitors for tumors refractory to MAPK inhibitors.

Dynamics and efficiency of electron injection and transport in DNA using pyrenecarboxamide as an electron donor and 5-bromouracil as an electron acceptor.

The photophysical and photochemical behavior of a series of hairpin-forming DNA conjugates possessing a 5'-tethered pyrenecarboxamide chromophore and one or two bromouracil bases has been investigated. Quenching of the pyrene fluorescence and transient absorption spectra characteristic of the pyrene cation radical are observed only when bromouracil is located at the first or second base pair position nearest to the point of pyrene attachment. These observations are consistent with an intercalated structure for these conjugates in which pyrene is adjacent to the second base pair. Selective quenching of singlet pyrene by bromouracil but not by thymine is consistent with the free energy for charge separation estimated using Weller's equation. Low quantum yields for loss of bromide when bromouracil is not adjacent to pyrene are attributed to inefficient charge separation via either a multistep electron transport or a single-step superexchange mechanism. Quantum yields are only weakly dependent upon the distance between pyrene and bromouracil, as expected for a multistep electron transport mechanism. Loss of bromide from conjugates possessing two bromouracils occurs sequentially. For adjacent bromouracils, competitive loss of bromide from both bromouracils is observed, whereas for nonadjacent bromouracils loss of bromide from the proximal bromouracil occurs prior to any loss from the distal bromouracil, consistent with a slower rate constant for electron transport vs loss of bromide.

Structure and photoinduced electron transfer in DNA hairpin conjugates possessing a tethered 5'-pyrenecarboxamide.

The structure, spectroscopy, and photophysical behavior of a series of hairpin-forming conjugates possessing a 5'-tethered N-alkylpyrenecarboxamide chromophore have been investigated. Comparison of the NMR spectra of the conjugates and analogs lacking the tethered pyrene indicates that the pyrene does not behave as an end-capping group but rather is intercalated between the two terminal hairpin base pairs. An intercalated structure is also consistent with the thermodynamic parameters for hairpin formation and the steady state and transient spectral properties of the conjugates. Quenching of the pyrene fluorescence and transient absorption spectra is observed only when guanine is located in one of the two terminal base pairs and is attributed to hole injection from singlet pyrene to guanine. The fast component of the transient decay is more rapid when guanine is located in the second vs first base pair, consistent with an intercalated but not an end-capped geometry. Spectral broadening of ultraviolet, fluorescence, and transient absorption spectra is attributed to multiple ground state conformations.

Getting to guanine: mechanism and dynamics of charge separation and charge recombination in DNA revisited.

The mechanism and dynamics of charge separation and charge recombination in synthetic DNA hairpins possessing a stilbenedicarboxamide linker and a single guanine-cytosine base pair have been reinvestigated. The combination of femtosecond broad-band pump probe spectroscopy, nanosecond transient absorption experiments, and picosecond fluorescence decay measurements permits analysis of the formation and decay of the stilbene anion radical. Reversible hole injection resulting in the formation of the stilbene-adenine contact radical ion pair is found to occur on the picosecond time scale. The mechanism for charge separation across two or more base pairs is revised from single step superexchange to a multi-step process: hole injection followed by hole transport and hole trapping. The mechanism of charge recombination remains assigned to a superexchange process.

Photoinduced charge separation in pyrenedicarboxamide-linked DNA hairpins.

The synthesis and photophysical properties of the dihydroxypropylamide derivative of pyrene-1,6-dicaboxamide, its aniline dyad, and DNA conjugates are reported. The dicarboxamide serves as a hairpin linker for bis(oligonucleotide) conjugates having short base pair stems. The dihydroxypropyl derivative has a large fluorescence quantum yield and long singlet decay time, as determined by fluorescence and time-resolved broad band pump-probe spectroscopy. The aniline dyad undergoes exergonic charge separation with formation of a radical ion pair which decays via charge recombination. The highly characteristic transient absorption spectrum of the pyrene anion radical is used to monitor the dynamics of its formation and decay. The dicarboxamide-linked hairpin conjugates undergo charge separation with adjacent guanine and adenine bases. Charge separation with guanine is accompanied by efficient pyrene fluorescence quenching. In contrast, reversible charge separation with adenine results in multiple exponential fluorescence decay. The energetics and dynamics of charge separation are compared with those of other arenedicarboxamide DNA hairpin linkers.

Crossover from superexchange to hopping as the mechanism for photoinduced charge transfer in DNA hairpin conjugates.

The mechanism and dynamics of photoinduced charge separation and charge recombination have been investigated in synthetic DNA hairpins possessing donor and acceptor stilbenes separated by one to seven A:T base pairs. The application of femtosecond broadband pump-probe spectroscopy, nanosecond transient absorption spectroscopy, and picosecond fluorescence decay measurements permits detailed analysis of the formation and decay of the stilbene acceptor singlet state and of the charge-separated intermediates. When the donor and acceptor are separated by a single A:T base pair, charge separation occurs via a single-step superexchange mechanism. However, when the donor and acceptor are separated by two or more A:T base pairs, charge separation occurs via a multistep process consisting of hole injection, hole transport, and hole trapping. In such cases, hole arrival at the electron donor is slower than hole injection into the bridging A-tract. Rate constants for charge separation (hole arrival) and charge recombination are dependent upon the donor-acceptor distance; however, the rate constant for hole injection is independent of the donor-acceptor distance. The observation of crossover from a superexchange to a hopping mechanism provides a "missing link" in the analysis of DNA electron transfer and requires reevaluation of the existing literature for photoinduced electron transfer in DNA.

Dynamics and mechanism of bridge-dependent charge separation in pyrenylurea-nitrobenzene pi-stacked protophanes.

Herein are reported the synthesis, structure, and electronic properties of a series of tertiary di- and polyarylureas possessing pyrene and nitrobenzene end groups separated by a variable number of internal phenylenediamine bridging groups. These molecules adopt folded "protophane" structures in which the adjacent arenes are loosely pi-stacked. The behavior of both the pyrene and nitrobenzene singlet states has been investigated by means of femtosecond broadband pump-probe spectroscopy, and the transients have been assigned on the basis of comparison to reference molecules. Femtosecond time resolution permits direct observation of the fast internal conversion process for both the pyrene and nitrobenzene upper singlet states, as well as the intersystem crossing of nitrobenzene. The ultrafast (ca. 100 fs) charge separation of the donor-acceptor urea having no bridging group is attributed to an internal conversion process. The slower charge separation and charge recombination of the donor-acceptor urea having a single bridging group occur via a bridge-mediated superexchange process. Addition of a second bridging unit results in a role reversal for the pyrene singlet state, which now serves as an excited-state acceptor with the bridging units serving as the electron donors. The change in the directionality of electron transfer upon addition of a second bridging phenylenediamine is a consequence of a decrease in the bridge oxidation potential as well as a decrease in the rate constant for single-step superexchange electron transfer.

p-Nitrobenzenesulfenate esters as precursors for laser flash photolysis studies of alkyl radicals.

A series of p-nitrobenzenesulfenate esters was used in laser flash photolysis (LFP) studies to generate alkoxyl radicals that fragmented to give the (2,2-diphenylcyclopropyl)methyl radical. Rate constants for the beta-scission reactions increased as a function of the carbonyl compound produced in the fragmentation reaction in the order CH2O < MeCHO < Me2CO < PhCHO < Ph2CO and increased with increasing solvent polarity. For alkoxyl radicals that fragment to produce benzaldehyde and benzophenone, the beta-scission reactions are faster than 1,5-hydrogen atom abstractions when the incipient carbon radical is as stable as a secondary alkyl radical, and this entry to carbon radicals can be used in LFP kinetic studies.