Early Drug Discovery Research Articles

Solvatochromic peptidic binder obtained via extended phage display acts as a fluororeporter for fragment-based drug discovery (FBDD).

We have previously established a selection system to obtain a solvatochromic protein binder from a peptidic fluoroprobe library via the extended T7 phage display. Here, we use the peptidic binder as a fluororeporter in this proof-of-concept study of fragment-based screening approach to drug discovery. The binder is released from the target protein on mixing with an appropriate lead compound, thereby altering its fluorescence color/intensity under 365nm ultraviolet wavelength irradiation. By this instant screening outcome, the affinity of the lead compound is apparent to the naked eye, and quantified with a portable microvolume fluorophotometer. We envision that our simple and affordable screening system will provideopportunities for early stage drug discovery, especially for non-experts in academia and education because expensive hardware is not required for qualifying the measurements.

Application of empirical scalars to enable early prediction of human hepatic clearance using IVIVE in drug discovery: an evaluation of 173 drugs.

The utilization of in vitro data to predict drug pharmacokinetics (PK) in vivo has been a consistent practice in early drug discovery for decades. However, its success is hampered by mispredictions attributed to uncharacterized biological phenomena/experimental artifacts. Predicted drug clearance (CL) from experimental data (i.e. hepatocyte intrinsic clearance: CLint, fraction unbound in plasma: fu,p) is often systematically underpredicted using the well-stirred model (WSM). The objective of this study was to evaluate using empirical scalars in the WSM to correct for CL mispredictions. Drugs (N=28) were used to generate numerical scalars on CLint (α), and fu,p (β) to minimize the error (AAFE) for CL predictions. These scalars were validated using an additional dataset (N=28 drugs) and applied to a non-redundant AstraZeneca (AZ) dataset available in the literature (N=117 drugs) for a total of 173 compounds. CL predictions using the WSM were improved for most compounds using an α value of 3.66 (~64%<2-fold) compared to no scaling (~46%<2-fold). Similarly, using a β value of 0.55 or combination of α and β scalars (values of 1.74 and 0.66, respectively) resulted in a similar improvement in predictions (~64%<2-fold and ~65%<2-fold, respectively). For highly bound compounds (fu,p{less than or equal to}0.01), AAFE was substantially reduced across all scaling methods. Using the β scalar alone or a combination of α and β appeared optimal; and produce larger magnitude corrections for highly-bound compounds. Some drugs are still disproportionally mispredicted, however the improvements in prediction error and simplicity of applying these scalars suggests its utility for early-stage CL predictions. Significance Statement In early drug discovery, prediction of human clearance using in vitro experimental data plays an essential role in triaging compounds prior to in vivo studies. These predictions have been systematically underestimated. Here we introduce empirical scalars calibrated on the extent of plasma protein binding that appear to improve clearance prediction across multiple datasets. This approach can be used in early phases of drug discovery prior to the availability of pre-clinical data for early quantitative predictions of human clearance.

Impedance-Based Phenotypic Readout of Transporter Function: A Case for Glutamate Transporters.

Excitatory amino acid transporters (EAAT/SLC1) mediate Na+-dependent uptake of extracellular glutamate and are potential drug targets for neurological disorders. Conventional methods to assess glutamate transport in vitro are based on radiolabels, fluorescent dyes or electrophysiology, which potentially compromise the cell’s physiology and are generally less suited for primary drug screens. Here, we describe a novel label-free method to assess human EAAT function in living cells, i.e., without the use of chemical modifications to the substrate or cellular environment. In adherent HEK293 cells overexpressing EAAT1, stimulation with glutamate or aspartate induced cell spreading, which was detected in real-time using an impedance-based biosensor. This change in cell morphology was prevented in the presence of the Na+/K+-ATPase inhibitor ouabain and EAAT inhibitors, which suggests the substrate-induced response was ion-dependent and transporter-specific. A mechanistic explanation for the phenotypic response was substantiated by actin cytoskeleton remodeling and changes in the intracellular levels of the osmolyte taurine, which suggests that the response involves cell swelling. In addition, substrate-induced cellular responses were observed for cells expressing other EAAT subtypes, as well as in a breast cancer cell line (MDA-MB-468) with endogenous EAAT1 expression. These findings allowed the development of a label-free high-throughput screening assay, which could be beneficial in early drug discovery for EAATs and holds potential for the study of other transport proteins that modulate cell shape.

Measuring Protein Aggregation and Stability Using High-Throughput Biophysical Approaches.

Structure-function relationships of biological macromolecules, in particular proteins, provide crucial insights for fundamental biochemistry, medical research and early drug discovery. However, production of recombinant proteins, either for structure determination, functional studies, or to be used as biopharmaceutical products, is often hampered by their instability and propensity to aggregate in solution in vitro. Protein samples of poor quality are often associated with reduced reproducibility as well as high research and production expenses. Several biophysical methods are available for measuring protein aggregation and stability. Yet, discovering and developing means to improve protein behaviour and structure-function integrity remains a demanding task. Here, we discuss workflows that are made possible by adapting established biophysical methods to high-throughput screening approaches. Rapid identification and optimisation of conditions that promote protein stability and reduce aggregation will support researchers and industry to maximise sample quality, stability and reproducibility, thereby reducing research and development time and costs.

One Assay, Two Vital Values‐Estimating Intrinsic Permeability and Efflux Using MDCK‐MDR1 Assay

BackgroundIntestinal intrinsic permeability and transporter mediated efflux are two major determinants of oral absorption. Typically, wild type Madin‐Darby canine kidney cells (MDCK) and MDCK cells transfected with the human MDR1 gene (MDCK‐MDR1) are used to evaluate intrinsic permeability and efflux, respectively. The MDCK assay provides intrinsic permeability but lacks transporter mediated efflux. Conversely, the MDCK‐MDR1 assay provides efflux ratio and “apparent permeability”. However, the apparent permeability can differ significantly from intrinsic permeability due to efflux. Traditionally, both assays (MDCK, MDCK‐MDR1) are applied separately and sequentially, which can be time‐consuming and costly.In this study, we present a novel approach to estimate intrinsic permeability and efflux simultaneously using the MDCK‐MDR1 assay alone. Intrinsic permeability can be calculated using a scaling method by combining the 3 parameters (bi‐directional apparent permeability values and efflux ratio) determined from the MDCK‐MDR1 assay.MethodsTest compounds of varying permeability and efflux ratios were used to assess the validity of the hypothesis. Results from the MDCK, MDCK‐MDR1 and caco‐2 assays were analyzed. Efflux ratio is estimated by dividing the apparent permeability from the basolateral to apical direction (PappBA) by the apparent permeability from the apical to basolateral direction (PappAB). Different averaging methods were used to calculate intrinsic permeability from MDCK‐MDR1 assay, and the measured intrinsic permeability from the MDCK assay was used as a comparison to find the most accurate averaging method.ResultsOf the different scaling and averaging methods used for estimating intrinsic permeability from MDCK‐MDR1 assay, the best method we found was a “quotient” scheme where calculated intrinsic permeability was adjusted based on the magnitude of efflux ratio. With this adjustment, both the intrinsic permeability and efflux ratio can be reliably obtained from the MDCK‐MDR1 assay for compounds with a wide range of efflux ratios. The calculated intrinsic permeability from the MDCK‐MDR1 assays aligns within 2 fold of the permeability values determined from the MDCK assay. Results from the caco‐2 cell line were also evaluated and applicable to the method mentioned above. Several hypotheses were formulated regarding why the “quotient” method was a better fit than simple apparent permeability averaging.ConclusionThe strategy presented here provides an efficient and effective method to estimate both intrinsic permeability and efflux solely based on the MDCK‐MDR1 assay. Compared to the traditional approach of using MDCK and MDCK‐MDR1 assays separately, the method used here requires less time, cost, and compound for testing. This is particularly attractive for early drug discovery screening needs.

Evaluation of the Success of High-Throughput Physiologically Based Pharmacokinetic (HT-PBPK) Modeling Predictions to Inform Early Drug Discovery.

Minimizing in vitro and in vivo testing in early drug discovery with the use of physiologically based pharmacokinetic (PBPK) modeling and machine learning (ML) approaches has the potential to reduce discovery cycle times and animal experimentation. However, the prediction success of such an approach has not been shown for a larger and diverse set of compounds representative of a lead optimization pipeline. In this study, the prediction success of the oral (PO) and intravenous (IV) pharmacokinetics (PK) parameters in rats was assessed using a “bottom-up” approach, combining in vitro and ML inputs with a PBPK model. More than 240 compounds for which all of the necessary inputs and PK data were available were used for this assessment. Different clearance scaling approaches were assessed, using hepatocyte intrinsic clearance and protein binding as inputs. In addition, a novel high-throughput PBPK (HT-PBPK) approach was evaluated to assess the scalability of PBPK predictions for a larger number of compounds in drug discovery. The results showed that bottom-up PBPK modeling was able to predict the rat IV and PO PK parameters for the majority of compounds within a 2- to 3-fold error range, using both direct scaling and dilution methods for clearance predictions. The use of only ML-predicted inputs from the structure did not perform well when using in vitro inputs, likely due to clearance miss predictions. The HT-PBPK approach produced comparable results to the full PBPK modeling approach but reduced the simulation time from hours to seconds. In conclusion, a bottom-up PBPK and HT-PBPK approach can successfully predict the PK parameters and guide early discovery by informing compound prioritization, provided that good in vitro assays are in place for key parameters such as clearance.

Identifying Acute Cardiac Hazard in Early Drug Discovery Using a Calcium Transient High-Throughput Assay in Human-Induced Pluripotent Stem Cell-Derived Cardiomyocytes.

Introduction: Early identification of cardiac risk is essential for reducing late-stage attrition in drug development. We adapted the previously published cardiac hazard risk-scoring system using a calcium transient assay in human stem cell-derived CMs for the identification of cardiac risks recorded from the new hiPSC-CM line and investigated its predictivity and translational value based on the screening of a large number of reference and proprietary compounds. Methods: Evaluation of 55 reference drugs provided the translation of various pharmacological effects into a single hazard label (no, low, high, or very high hazard) using a Ca2+-sensitive fluorescent dye assay recorded by -by FDSS/µCell Functional Drug Screening System (Hamamatsu on hiPSC-CM line (FCDI iCell Cardiomyocytes2). Results: Application of the adapted hazard scoring system in the Ca2+ transient assay, using a second hiPS-CM line, provided comparable scoring results and predictivity of hazard, to the previously published scoring approach, with different pharmacological drug classes, as well as screening new chemical entities (NCE’s) using a single hazard label from four different scoring levels (no, low, high, or very high hazard). The scoring system results also showed minimal variability across three different lots of hiPSC-CMs, indicating good reproducibility of the cell line. The predictivity values (sensitivity and specificity) for drug-induced acute cardiac risk for QT-interval prolongation and Torsade de pointes (TdPs) were >95% and statistical modeling confirmed the prediction of proarrhythmic risk. The outcomes of the NCEs also showed consistency with findings in other well-established in vitro and in vivo cardiac risk assays. Conclusion: Evaluation of a large list of reference compounds and internal NCEs has confirmed the applicability of the adaptations made to the previously published novel scoring system for the hiPSC-CMs. The validation also established the predictivity for drug-induced cardiac risks with good translation to other established preclinical in vitro and in vivo assays, confirming the application of this novel scoring system in different stem cell-CM lines for early cardiac hazard identification.

Designing Soluble PROTACs: Strategies and Preliminary Guidelines.

Solubility optimization is a crucial step to obtaining oral PROTACs. Here we measured the thermodynamic solubilities (log S) of 21 commercial PROTACs. Next, we measured BRlogD and log kwIAM (lipophilicity), EPSA, and Δ log kwIAM (polarity) and showed that lipophilicity plays a major role in governing log S, but a contribution of polarity cannot be neglected. Two-/three-dimensional descriptors calculated on conformers arising from conformational sampling and steered molecular dynamics failed in modeling solubility. Infographic tools were used to identify a privileged region of soluble PROTACs in a chemical space defined by BRlogD, log kwIAM and topological polar surface area, while machine learning provided a log S classification model. Finally, for three pairs of PROTACs we measured the solubility, lipophilicity, and polarity of the building blocks and identified the limits of estimating PROTAC solubility from the synthetic components. Overall, this paper provides promising guidelines for optimizing PROTAC solubility in early drug discovery programs.

Pose Classification Using Three-Dimensional Atomic Structure-Based Neural Networks Applied to Ion Channel-Ligand Docking.

The identification of promising lead compounds showing pharmacological activities toward a biological target is essential in early stage drug discovery. With the recent increase in available small-molecule databases, virtual high-throughput screening using physics-based molecular docking has emerged as an essential tool in assisting fast and cost-efficient lead discovery and optimization. However, the best scored docking poses are often suboptimal, resulting in incorrect screening and chemical property calculation. We address the pose classification problem by leveraging data-driven machine learning approaches to identify correct docking poses from AutoDock Vina and Glide screens. To enable effective classification of docking poses, we present two convolutional neural network approaches: a three-dimensional convolutional neural network (3D-CNN) and an attention-based point cloud network (PCN) trained on the PDBbind refined set. We demonstrate the effectiveness of our proposed classifiers on multiple evaluation data sets including the standard PDBbind CASF-2016 benchmark data set and various compound libraries with structurally different protein targets including an ion channel data set extracted from Protein Data Bank (PDB) and an in-house KCa3.1 inhibitor data set. Our experiments show that excluding false positive docking poses using the proposed classifiers improves virtual high-throughput screening to identify novel molecules against each target protein compared to the initial screen based on the docking scores.

Optimization of a locomotion-based zebrafish seizure model

BackgroundLocomotor assays in zebrafish have emerged as a screening test in early drug discovery for antiseizure compounds. However, parameters differ considerably between published studies, which may explain some discrepant results with (candidate) antiseizure medications. New methodWe optimized a locomotor-based seizure assay in zebrafish with pentylenetetrazol (PTZ) as the pharmacological proconvulsant to generate a therapeutic window in which proconvulsant-treated zebrafish larvae could be discriminated from a non-treated control. To generate a reliable control, exposure time and concentration of valproate (VPA, anticonvulsant) was optimized. ResultsWells with one or three larvae show a similar PTZ dose-dependent increase in locomotion with less variability in motility for the latter. Zebrafish immersed in 10 mM PTZ showed a significant increase in movement with a sustained effect, without any indication of toxicity. Animals treated with 3 mM VPA showed the strongest reduction of PTZ-induced movement without toxicity. The decrease in PTZ-induced locomotion was greater after 18 h versus 2 h. Comparison with existing method(s)For the larval zebrafish PTZ-induced seizure model, varying experimental parameters have been reported in literature. Our results show that PTZ is often used at toxic concentrations, and we provide instead reliable conditions to quantify convulsant behaviour using an infrared-beam motility assay. ConclusionsWe recommend using three zebrafish larvae per well to quantify locomotion in 96-multiwell plates. Larvae should preferably be exposed to 10 mM PTZ for 1 h, consisting of 30 min acclimation and 30 min subsequent recording. As positive control for anticonvulsant activity, we recommend exposure to 3 mM VPA for 18 h before administration of PTZ.

Kinetic profiling and functional characterization of 8-phenylxanthine derivatives as A2B adenosine receptor antagonists

A2B adenosine receptor (A2BAR) antagonists have therapeutic potential in inflammation-related diseases such as asthma, chronic obstructive pulmonary disease and cancer. However, no drug is currently clinically approved, creating a demand for research on novel antagonists. Over the last decade, the study of target binding kinetics, along with affinity and potency, has been proven valuable in early drug discovery stages, as it is associated with improved in vivo drug efficacy and safety. In this study, we report the synthesis and biological evaluation of a series of xanthine derivatives as A2BAR antagonists, including an isothiocyanate derivative designed to bind covalently to the receptor. All 28 final compounds were assessed in radioligand binding experiments, to evaluate their affinity and for those qualifying, kinetic binding parameters. Both structure-affinity and structure-kinetic relationships were derived, providing a clear relationship between affinity and dissociation rate constants. Two structurally similar compounds, 17 and 18, were further evaluated in a label-free assay due to their divergent kinetic profiles. An extended cellular response was associated with long A2BAR residence times. This link between a ligand’s A2BAR residence time and its functional effect highlights the importance of binding kinetics as a selection parameter in the early stages of drug discovery.

Optimization of Cyclic Peptide Property Using Chromatographic Capacity Factor on Permeability of Passive Cell Membrane and Human Induced Pluripotent Stem Cell-Derived Intestinal Membrane

Cyclic peptides have attracted increasing attention as a privileged class of molecules addressing undruggable targets. Cell permeability of cyclic peptides has remained a challenging issue owing to their molecular properties. Various efficiency metrics have emerged to assess this issue. Among them, the lipophilic permeability efficiency (LPE) metric is the difference between an experimental 1,9-decadiene-water partition coefficient at pH 7.4 (log Ddec/w) and calculated octanol/water partition coefficients (ALogP). This metric provides insight into how structural changes affect permeability. Here, we demonstrate the chromatographic capacity factor (log k’) of cyclic peptides using reversed-phase liquid chromatography as an alternative to log Ddec/w, which enables efficient and reliable experimental lipophilicity for the adoption of LPE in early drug discovery. The log k’ indicates the passive membrane permeability of cyclic peptides and can be used to optimize passive membrane permeability in combination with other parameters. In addition, intestinal membrane permeability of cyclic peptides on human induced pluripotent stem cell-derived intestinal epithelial cells was achieved with log k’ and high passive membrane permeability, although cyclic peptides are P-glycoprotein substrates. These approaches could facilitate optimization of properties of cyclic peptides for oral administration and contribute to the successful discovery and development of cyclic peptides.

RETRACTED: Hellinen et al. Drug Flux across RPE Cell Models: The Hunt for an Appropriate Outer Blood–Retinal Barrier Model for Use in Early Drug Discovery. Pharmaceutics 2020, 12, 176

The published article [...].

The transformational role of GPU computing and deep learning in drug discovery

Deep learning has disrupted nearly every field of research, including those of direct importance to drug discovery, such as medicinal chemistry and pharmacology. This revolution has largely been attributed to the unprecedented advances in highly parallelizable graphics processing units (GPUs) and the development of GPU-enabled algorithms. In this Review, we present a comprehensive overview of historical trends and recent advances in GPU algorithms and discuss their immediate impact on the discovery of new drugs and drug targets. We also cover the state-of-the-art of deep learning architectures that have found practical applications in both early drug discovery and consequent hit-to-lead optimization stages, including the acceleration of molecular docking, the evaluation of off-target effects and the prediction of pharmacological properties. We conclude by discussing the impacts of GPU acceleration and deep learning models on the global democratization of the field of drug discovery that may lead to efficient exploration of the ever-expanding chemical universe to accelerate the discovery of novel medicines.

A guide to enzyme kinetics in early drug discovery.

Drugs interact with their target of interest to bring about the desired phenotypic outcome that results in disease alleviation. Traditionally, most lead optimization exercises were driven by affinity measures (like IC50 ) to inform structure-activity relationship (SAR)-guided medicinal chemistry. However, an IC50 value is a thermodynamic estimate measured under equilibrium conditions that can vary as a function of substrate concentration and/or time (the latter especially for nonequilibrium modalities). Further, like other thermodynamic estimates, it is a state-function that is indifferent to the path traversed from the initial state to the final state. This can be a cause for concern in drug discovery given the predominance of nonequilibrium interactions and the open thermodynamic nature of the human system. Under such situations, employing rates along with equilibrium constants (or IC50 values) would be far more relevant to capture the time evolution of the small molecule's interaction with the target of interest. These rates are generally typified by the rate of association, rate of dissociation and the residence time of the small molecule on the target (target occupancy). These parameters, when combined with the concept of target vulnerability, therapeutic window, pharmacokinetic profile of the small molecule, estimates of endogenous ligand and target turnover, will shed critical insights into the kinetics and dynamics of a small molecule's interaction with the protein, and allow realistic modelling of the system to enable optimizations and dosing decisions. With that aim, this guide will attempt to introduce the traditional role of mechanistic enzymology within drug discovery and emphasize the importance of kinetics in guiding SAR-based optimizations. It will also present initial ideas on how kinetic investigation should be positioned relative to the temporal span of a drug-discovery pipeline to leverage maximal utility from the investment in time and effort.

Practical Application of Rodent Transporter Knockout Models to Assess Brain Penetration in Drug Discovery.

Compound X is a drug candidate for the treatment of neurodegenerative diseases. Its brain distribution was evaluated as part of the lead identification and optimization activities undertaken in early drug discovery. The brain distribution of compound X was studied in genetic transporter knockout rodent models, in vivo models with a chemical inhibitor, and in vitro transporter cell systems. Compound X was found to be a substrate for human Breast Cancer-Resistance Protein (BCRP) in vitro (efflux ratio 8.1) and rodent Bcrp in vivo (Kp, uuKO/Kp, uuWT = 0.15/0.057 = 2.7, p< 0.05) but not a substrate for human P-glycoprotein (P-gp) in vitro (efflux ratio 1.0) nor rodent P-gp in vivo (Kp, uuKO/Kp, uuWT = 0.056/ 0.051 = 1.1, p> 0.05). When both transporters were knocked out in vivo, Kp, uu increased to 0.51±0.02. A similar pattern observed across compounds with related chemistry corroborating the structure-activity relationship. While in vitro assays showed compound X to be a substrate for human BCRP and not P-gp, in vivo studies indicated a synergistic effect between rodent efflux transporters. However, this only accounted for ~50% of restricted BBB-transport, suggesting involvement of other efflux transporters. Considering Kp, uu as a key criterion for assessing the technical quality of CNS candidates before progression into clinical development, it is important to identify relevant screening assays for a better understanding of low Kp, uu and brain distribution in pre-clinical models for translation to humans.

Label-free and homogeneous detection of flap endonuclease 1 by ligation-promoted hyperbranched rolling circle amplification platform

The structure-specific endonuclease FEN1 participates in various genome maintenance pathways in eukaryotes and is associated with different human diseases. Herein, we demonstrate label-free and homogeneous detection of FEN1 based on ligation-promoted hyperbranched rolling circle amplification (HRCA). This assay can be performed isothermally with the involvement of primers 1 and 2 and a circular DNA substrate with a 5′-flap. When FEN1 is present, it cleaves 5′-flap of circular DNA substrate to obtain a circular padlock probe with the assistance of Taq DNA ligase. The circular padlock probe can serve as a template to initiate HRCA in the presence of primers 1 and 2 and Vent (exo−) DNA polymerase. The obtained dsDNA fragments can produce an enhanced fluorescence signal with SYBR Green I as indicator. This method displays good specificity and high sensitivity, and it can be employed to screen FEN1 inhibitors and quantitatively detect FEN1 activity in human cancer cells, with potential applications in early diagnosis and drug discovery.

Fatty Acid Biosynthesis: An Updated Review on KAS Inhibitors.

Since the early twentieth century, with the isolation of penicillin and streptomycin in the 1940s, the modern era of anti-infective drug development has gained momentum. Due to the enormous success of early drug discovery, many infectious diseases were successfully prevented and eradicated. However, this initial hope was wrongheaded, and pathogens evolved as a significant threat to human health. Drug resistance develops as a result of natural selection's relentless pressure, necessitating the identification of new drug targets and the creation of chemotherapeutics that bypass existing drug resistance mechanisms. Fatty acid biosynthesis (FAS) is a crucial metabolic mechanism for bacteria during their growth and development. Several crucial enzymes involved in this biosynthetic pathway have been identified as potential targets for new antibacterial agents. In Escherichia coli (E. coli), this pathway has been extensively investigated. The present review focuses on progress in the development of Kas A, Kas B, and Fab H inhibitors as mono-therapeutic antibiotics.

A network-based computational and experimental framework for repurposing compounds toward the treatment of non-alcoholic fatty liver disease

SummaryNon-alcoholic fatty liver disease (NAFLD) is among the most common liver pathologies, however, none approved condition-specific therapy yet exists. The present study introduces a drug repositioning (DR) approach that combines in vitro steatosis models with a network-based computational platform, constructed upon genomic data from diseased liver biopsies and compound-treated cell lines, to propose effectively repositioned therapeutic compounds. The introduced in silico approach screened 20′000 compounds, while complementary in vitro and proteomic assays were developed to test the efficacy of the 46 in silico predictions. This approach successfully identified six compounds, including the known anti-steatogenic drugs resveratrol and sirolimus. In short, gallamine triethiotide, diflorasone, fenoterol, and pralidoxime ameliorate steatosis similarly to resveratrol/sirolimus. The implementation holds great potential in reducing screening time in the early drug discovery stages and in delivering promising compounds for in vivo testing.

Quadruple Target Evaluation of Diversity-Optimized Halogen-Enriched Fragments (HEFLibs) Reveals Substantial Ligand Efficiency for AP2-Associated Protein Kinase 1 (AAK1).

Fragment-based drug discovery is one of the most utilized approaches for the identification of novel weakly binding ligands, by efficiently covering a wide chemical space with rather few compounds and by allowing more diverse binding modes to be found. This approach has led to various clinical candidates and approved drugs. Halogen bonding, on the other hand, has gained traction in molecular design and lead optimization, but could offer additional benefits in early drug discovery. Screening halogen-enriched fragments (HEFLibs) could alleviate problems associated with the late introduction of such a highly geometry dependent interaction. Usually, the binding mode is then already dominated by other strong interactions. Due to the fewer competing interactions in fragments, the halogen bond should more often act as an anchor point for the binding mode. Previously, we proposed a fragment library with a focus on diverse binding modes that involve halogens for gaining initial affinity and selectivity. Herein, we demonstrate the applicability of these HEFLibs with a small set of diverse enzymes: the histone-lysine N-methyltransferase DOT1L, the indoleamine 2,3-dioxygenase 1 (IDO1), the AP2-associated protein kinase 1 (AAK1), and the calcium/calmodulin-dependent protein kinase type 1G (CAMK1G). We were able to identify various binding fragments via STD-NMR. Using ITC to verify these initial hits, we determined affinities for many of these fragments. The best binding fragments exhibit affinities in the one-digit micromolar range and ligand efficiencies up to 0.83 for AAK1. A small set of analogs was used to study structure-affinity relationships and hereby analyze the specific importance of each polar interaction. This data clearly suggests that the halogen bond is the most important interaction of fragment 9595 with AAK1.