Comparability study of monocyte derived dendritic cells, primary monocytes, and THP1 cells for innate immune responses.

Immunogenicity is one major challenge to the successful development of biotherapeutics because it could adversely affect PK/PD, safety, and efficacy. Preclinical immunogenicity risk assessment strategies and assays have been developed and implemented to screen and optimize discovery molecules. Internalization by antigen presenting cells (APC) and innate immune activation are initial prerequisite steps in eliciting immune responses to biotherapeutics. Dendritic cells (DC)- and monocyte-based assays are employed to interrogate such risks, and their value has been well documented in the literature. However, these assays have limited throughput, exhibit higher variability, and entail lengthy and complex procedures as they are based on primary cells such as peripheral blood mononuclear cells (PBMC) from individual donors. Herein, we investigated THP1 cells as surrogate cells to study APC internalization and innate immune activation. Comparability studies showed that THP1 cells could resemble innate immune responses of monocyte-derived DC and primary CD14+ monocytes using a panel of therapeutic antibodies. In addition, an automated high throughput THP1 internalization assay was qualified to enable risk assessment at pre­lead stages. The results demonstrated that THP1 cells can be utilized to assess immunogenicity risk in a high throughput manner.

The Janus kinase 1/2 inhibitor baricitinib reduces biomarkers of joint destruction in moderate to severe rheumatoid arthritis.

BACKGROUND: Tissue released blood-based biomarkers can provide insight into drug mode of action and response. To understand the changes in extracellular matrix turnover, we analyzed biomarkers associated with joint tissue turnover from a phase 3, randomized, placebo-controlled study of baricitinib in patients with active rheumatoid arthritis (RA). METHODS: Serum biomarkers associated with synovial inflammation (C1M, C3M, and C4M), cartilage degradation (C2M), bone resorption (CTX-I), and bone formation (osteocalcin) were analyzed at baseline, and weeks 4 and 12, from a subgroup of patients (n = 240) randomized to placebo or 2-mg or 4-mg baricitinib (RA-BUILD, NCT01721057). Mixed-model repeated measure was used to identify biomarkers altered by baricitinib. The relationship between changes in biomarkers and clinical measures was evaluated using correlation analysis. RESULTS: Treatment arms were well balanced for baseline biomarkers, demographics, and disease activity. At week 4, baricitinib 4-mg significantly reduced C1M from baseline by 21% compared to placebo (p < 0.01); suppression was sustained at week 12 (27%, p < 0.001). Baricitinib 4-mg reduced C3M and C4M at week 4 by 14% and 12% compared to placebo, respectively (p < 0.001); they remained reduced by 16% and 11% at week 12 (p < 0.001). In a pooled analysis including all treatment arms, patients with the largest reduction (upper 25% quartile) in C1M, C3M, and C4M by week 12 had significantly greater clinical improvement in the Simplified Disease Activity Index at week 12 compared to patients with the smallest reduction (lowest 25% quartile). CONCLUSION: Baricitinib treatment resulted in reduced circulating biomarkers associated with joint tissue destruction as well as concomitant RA clinical improvement. TRIAL REGISTRATION: ClinicalTrials.gov NCT01721057 ; date of registration: November 1, 2012.

The Derivation of a Matched Molecular Pairs Based ADME/Tox Knowledge Base for Compound Optimization.

Matched Molecular Pairs (MMP) analysis is a well-established technique for Structure Activity and Property Analysis (SAR and SPR). Summarizing multiple MMPs that describe the same structural change into a single chemical transform can be a powerful tool for prediction (termed Transform from here on). This is particularly useful in the area of Absorption, Distribution, Metabolism, and Elimination (ADME) analysis that is less influenced by 3D structural binding effects. The creation of a knowledge database containing many of these Transforms across typical ADME assays promises to be a powerful approach to aid multidimensional optimization. We present a detailed workflow for the derivation of such a database. We include details of an MMP fragmentation algorithm with associated statistical summarization methods for the derivation of Transforms. This is made freely available as part of the LillyMol software package. We describe the application of this method to several ADME/Tox (Toxicity) assay data sets and highlight multiple cases where the impact of traditional medicinal chemistry Transforms is contradicted by MMP data. We also describe the internal software interface used by medicinal chemists to aid the design of new compounds via automated suggestion. This approach utilizes the matched pairs database to "suggest" improved compounds in an automated design scenario. A nonvisual script-based version of the automated suggestions code with an associated set of described chemical Transforms is also made freely available along with this paper and as part of the LillyMol software package. Finally, we contrast this knowledge database against a larger database of all MMPs derived from a 2 million compound diversity set and a subset of MMPs seen in historical discovery projects. The comparison against all transforms in the diversity collection highlights the very low coverage of the transform database as compared to all possible transforms involving 15 atom fragments. The comparison against a smaller subset of Transforms seen on internal Medicinal Chemistry projects shows better coverage of the transform database for a small set of common medicinal chemistry strategies. Within the context of all possible transforms available to a medicinal chemistry project team, the challenge remains to move beyond mere idea generation from past projects toward high quality prediction for novel ADME/Tox modulating Transforms.

Development of a FRET-Based Assay for Analysis of mAbs Internalization and Processing by Dendritic Cells in Preclinical Immunogenicity Risk Assessment.

Treatment-emergent antidrug antibodies (TE-ADA) pose a major challenge to the development of biotherapeutics. The antidrug antibody responses are highly orchestrated and involve many types of immune cells and biological processes. Biological drug internalization and processing by antigen-presenting cells (APCs) are two initial and critical steps in the cascade of events leading to T cell-dependent ADA production. The assays thus far described in literature to evaluate immunogenicity potential/risk as a function of APC activity mainly focus on internalization of labeled drug candidates in vitro. Herein, we describe a high-throughput Förster Resonance Energy Transfer (FRET)-based assay for assessing both internalization and processing using CD14+ monocyte-derived dendritic cells (DCs) as APCs. Antigen-binding fragment F(ab')2 against IgG fragment crystallizable gamma (Fcγ) was labeled with the activatable FRET pair TAMRA-QSY7 as a universal probe for antibodies and proteins with a fragment crystallizable (Fc) domain. The assay was qualified using six mAbs of known clinical immunogenicity and one IgG1 isotype antibody using Design of Experiment (DoE). Correlation analysis of internalization and clinical immunogenicity data showed that this FRET-based internalization assay was able to detect clinically immunogenic antibodies. This method provides a tool for analyzing/screening the immunogenicity risk of biological candidates by assessing one of the critical components of the ADA formation process within the broader context of an immunogenicity risk assessment strategy.

Exploring Tunable Hyperparameters for Deep Neural Networks with Industrial ADME Data Sets.

Deep learning has drawn significant attention in different areas including drug discovery. It has been proposed that it could outperform other machine learning algorithms, especially with big data sets. In the field of pharmaceutical industry, machine learning models are built to understand quantitative structure-activity relationships (QSARs) and predict molecular activities, including absorption, distribution, metabolism, and excretion (ADME) properties, using only molecular structures. Previous reports have demonstrated the advantages of using deep neural networks (DNNs) for QSAR modeling. One of the challenges while building DNN models is identifying the hyperparameters that lead to better generalization of the models. In this study, we investigated several tunable hyperparameters of deep neural network models on 24 industrial ADME data sets. We analyzed the sensitivity and influence of five different hyperparameters including the learning rate, weight decay for L2 regularization, dropout rate, activation function, and the use of batch normalization. This paper focuses on strategies and practices for DNN model building. Further, the optimized model for each data set was built and compared with the benchmark models used in production. Based on our benchmarking results, we propose several practices for building DNN QSAR models.

Arrhenius time-scaled least squares: a simple, robust approach to accelerated stability data analysis for bioproducts.

Defining a suitable product presentation with an acceptable stability profile over its intended shelf-life is one of the principal challenges in bioproduct development. Accelerated stability studies are routinely used as a tool to better understand long-term stability. Data analysis often employs an overall mass action kinetics description for the degradation and the Arrhenius relationship to capture the temperature dependence of the observed rate constant. To improve predictive accuracy and precision, the current work proposes a least-squares estimation approach with a single nonlinear covariate and uses a polynomial to describe the change in a product attribute with respect to time. The approach, which will be referred to as Arrhenius time-scaled (ATS) least squares, enables accurate, precise predictions to be achieved for degradation profiles commonly encountered during bioproduct development. A Monte Carlo study is conducted to compare the proposed approach with the common method of least-squares estimation on the logarithmic form of the Arrhenius equation and nonlinear estimation of a first-order model. The ATS least squares method accommodates a range of degradation profiles, provides a simple and intuitive approach for data presentation, and can be implemented with ease.

Selectivity data: assessment, predictions, concordance, and implications.

Could high-quality in silico predictions in drug discovery eventually replace part or most of experimental testing? To evaluate the agreement of selectivity data from different experimental or predictive sources, we introduce the new metric concordance minimum significant ratio (cMSR). Empowered by cMSR, we find the overall level of agreement between predicted and experimental data to be comparable to that found between experimental results from different sources. However, for molecules that are either highly selective or potent, the concordance between different experimental sources is significantly higher than the concordance between experimental and predicted values. We also show that computational models built from one data set are less predictive for other data sources and highlight the importance of bias correction for assessing selectivity data. Finally, we show that small-molecule target space relationships derived from different data sources and predictive models share overall similarity but can significantly differ in details.

What general conclusions can we draw from kinase profiling data sets?

Understanding general selectivity trends across the kinome has implications ranging from target selection, compound prioritization, toxicity and patient tailoring. Several recent publications have described the characterization of kinase inhibitors via large assay panels, offering a range of generalizations that influenced kinase inhibitor research trends. Since a subset of profiled inhibitors overlap across reports, we evaluated the concordance of activity results for the same compound-kinase pairs across four data sources generated from different kinase biochemical assay technologies. Overall, 77% of all results are within 3 fold or qualitatively in agreement across sources. However, the agreement for active compounds is only 37%, indicating that different profiling panels are in better agreement to determine a compound's lack of activity rather than degree of activity. Low concordance is also found when comparing the promiscuity of kinase targets evaluated from different sources, and the pharmacological similarity of kinases. In contrast, the overall promiscuity of kinase inhibitors was consistent across sources. We highlight the difficulty of drawing general conclusions from such data by showing that no significant selectivity difference distinguishes type I vs. type II inhibitors, and limited kinase space similarity that is consistent across different sources. This article is part of a Special Issue entitled: Inhibitors of Protein Kinases (2012).

Development and validation of an antibody-dependent cell-mediated cytotoxicity-reporter gene assay.

Humanized monoclonal antibodies (mAbs) are the fastest growing class of biological therapeutics that are being developed for various medical indications, and more than 30 mAbs are already approved and in the market place. Antibody-dependent cell-mediated cytotoxicity (ADCC) is an important biological function attributed to the mechanism of action of several therapeutic antibodies, particularly oncology targeting mAbs. The ADCC assay is a complicated and highly variable assay. Thus, the use of an ADCC assay as a lot release test or a stability test for clinical trial batches of mAbs has been a substantial challenge to install in quality control laboratories. We describe here the development and validation of an alternate approach, an ADCC-reporter gene assay that is based on the key attributes of the PBMC-based ADCC assay. We tested the biological relevance of this assay using an anti-CD20 based model and demonstrated that this ADCC-reporter assay correlated well with standard ADCC assays when induced with the drugable human isotypes [IgG1, IgG2, IgG4, IgG4S > P (S228P) and IgG4PAA (S228P, F234A, L235A)] and with IgG1 isotype variants with varying amounts of fucosylation. This data demonstrates that the ADCC-reporter gene assay has performance characteristics (accuracy, precision and robustness) to be used not only as a potency assay for lot release and stability testing for antibody therapeutics, but also as a key assay for the characterization and process development of therapeutic molecules.

A general approach to confidence regions for optimal factor levels of response surfaces.

For a response surface experiment, an approximate hypothesis test and an associated confidence region is proposed for the minimizing (or maximizing) factor-level configuration. Carter et al. (1982, Cancer Research 42, 2963-2971) show that confidence regions for optimal conditions provide a way to make decisions about therapeutic synergism. The response surface may be constrained to be within a specified, bounded region. These constraint regions can be quite general. This allows for more realistic constraint modeling and a wide degree of applicability, including constraints occurring in mixture experiments. The usual assumption of a quadratic model is also generalized to include any regression model that is linear in the model parameters. An intimate connection is established between this confidence region and the Box-Hunter (1954, Biometrika 41, 190-199) confidence region for a stationary point. As a byproduct, this methodology also provides a way to construct a confidence interval for the difference between the optimal mean response and the mean response at a specified factor-level configuration. The application of this confidence region is illustrated with two examples. Extensive simulations indicate that this confidence region has good coverage properties.