Exploration of the Tunability of BRD4 Degradation by DCAF16 Trans-labelling Covalent Glues.

Small molecules that can induce protein degradation by inducing proximity between a desired target and an E3 ligase have the potential to greatly expand the number of proteins that can be manipulated pharmacologically. Current strategies for targeted protein degradation are mostly limited in their target scope to proteins with preexisting ligands. Alternate modalities such as molecular glues, as exemplified by the glutarimide class of ligands for the CUL4CRBN ligase, have been mostly discovered serendipitously. We recently reported a trans-labelling covalent glue mechanism which we named 'Template-assisted covalent modification', where an electrophile decorated small molecule binder of BRD4 was effectively delivered to a cysteine residue on an E3 ligase DCAF16 as a consequence of a BRD4-DCAF16 protein-protein interaction. Herein, we report our medicinal chemistry efforts to evaluate how various electrophilic modifications to the BRD4 binder, JQ1, affect DCAF16 trans-labeling and subsequent BRD4 degradation efficiency. We discovered a decent correlation between the ability of the electrophilic small molecule to induce ternary complex formation between BRD4 and DCAF16 with its ability to induce BRD4 degradation. Moreover, we show that a more solvent-exposed warhead presentation is optimal for DCAF16 recruitment and subsequent BRD4 degradation. Unlike the sensitivity of CUL4CRBN glue degraders to chemical modifications, the diversity of covalent attachments in this class of BRD4 glue degraders suggests a high tolerance and tunability for the BRD4-DCAF16 interaction. This offers a potential new avenue for a rational design of covalent glue degraders by introducing covalent warheads to known binders.

Template-assisted covalent modification of DCAF16 underlies activity of BRD4 molecular glue degraders.

Small molecules that induce protein-protein interactions to exert proximity-driven pharmacology such as targeted protein degradation are a powerful class of therapeutics1-3. Molecular glues are of particular interest given their favorable size and chemical properties and represent the only clinically approved degrader drugs4-6. The discovery and development of molecular glues for novel targets, however, remains challenging. Covalent strategies could in principle facilitate molecular glue discovery by stabilizing the neo-protein interfaces. Here, we present structural and mechanistic studies that define a trans-labeling covalent molecular glue mechanism, which we term "template-assisted covalent modification". We found that a novel series of BRD4 molecular glue degraders act by recruiting the CUL4DCAF16 ligase to the second bromodomain of BRD4 (BRD4BD2). BRD4BD2, in complex with DCAF16, serves as a structural template to facilitate covalent modification of DCAF16, which stabilizes the BRD4-degrader-DCAF16 ternary complex formation and facilitates BRD4 degradation. A 2.2 Å cryo-electron microscopy structure of the ternary complex demonstrates that DCAF16 and BRD4BD2 have pre-existing structural complementarity which optimally orients the reactive moiety of the degrader for DCAF16Cys58 covalent modification. Systematic mutagenesis of both DCAF16 and BRD4BD2 revealed that the loop conformation around BRD4His437, rather than specific side chains, is critical for stable interaction with DCAF16 and BD2 selectivity. Together our work establishes "template-assisted covalent modification" as a mechanism for covalent molecular glues, which opens a new path to proximity driven pharmacology.

Using a Deep Learning Algorithm and Integrated Gradients Explanation to Assist Grading for Diabetic Retinopathy.

PURPOSE: To understand the impact of deep learning diabetic retinopathy (DR) algorithms on physician readers in computer-assisted settings. DESIGN: Evaluation of diagnostic technology. PARTICIPANTS: One thousand seven hundred ninety-six retinal fundus images from 1612 diabetic patients. METHODS: Ten ophthalmologists (5 general ophthalmologists, 4 retina specialists, 1 retina fellow) read images for DR severity based on the International Clinical Diabetic Retinopathy disease severity scale in each of 3 conditions: unassisted, grades only, or grades plus heatmap. Grades-only assistance comprised a histogram of DR predictions (grades) from a trained deep-learning model. For grades plus heatmap, we additionally showed explanatory heatmaps. MAIN OUTCOME MEASURES: For each experiment arm, we computed sensitivity and specificity of each reader and the algorithm for different levels of DR severity against an adjudicated reference standard. We also measured accuracy (exact 5-class level agreement and Cohen's quadratically weighted κ), reader-reported confidence (5-point Likert scale), and grading time. RESULTS: Readers graded more accurately with model assistance than without for the grades-only condition (P < 0.001). Grades plus heatmaps improved accuracy for patients with DR (P < 0.001), but reduced accuracy for patients without DR (P = 0.006). Both forms of assistance increased readers' sensitivity moderate-or-worse DR: unassisted: mean, 79.4% [95% confidence interval (CI), 72.3%-86.5%]; grades only: mean, 87.5% [95% CI, 85.1%-89.9%]; grades plus heatmap: mean, 88.7% [95% CI, 84.9%-92.5%] without a corresponding drop in specificity (unassisted: mean, 96.6% [95% CI, 95.9%-97.4%]; grades only: mean, 96.1% [95% CI, 95.5%-96.7%]; grades plus heatmap: mean, 95.5% [95% CI, 94.8%-96.1%]). Algorithmic assistance increased the accuracy of retina specialists above that of the unassisted reader or model alone; and increased grading confidence and grading time across all readers. For most cases, grades plus heatmap was only as effective as grades only. Over the course of the experiment, grading time decreased across all conditions, although most sharply for grades plus heatmap. CONCLUSIONS: Deep learning algorithms can improve the accuracy of, and confidence in, DR diagnosis in an assisted read setting. They also may increase grading time, although these effects may be ameliorated with experience.

Real time alert system: a disease management system leveraging health information exchange.

BACKGROUND: Rates of preventive and disease management services can be improved by providing automated alerts and reminders to primary care providers (PCPs) using of health information technology (HIT) tools. METHODS: Using Adaptive Turnaround Documents (ATAD), an existing Health Information Exchange (HIE) infrastructure and office fax machines, we developed a Real Time Alert (RTA) system. RTA is a computerized decision support system (CDSS) that is able to deliver alerts to PCPs statewide for recommended services around the time of the patient visit. RTA is also able to capture structured clinical data from providers using existing fax technology. In this study, we evaluate RTA's performance for alerting PCPs when their patients with asthma have an emergency room visit anywhere in the state. RESULTS: Our results show that RTA was successfully able to deliver "just in time" patient-relevant alerts to PCPs across the state. Furthermore, of those ATADs faxed back and automatically interpreted by the RTA system, 35% reported finding the provided information helpful. The PCPs who reported finding information helpful also reported making a phone call, sending a letter or seeing the patient for follow up care. CONCLUSIONS: We have successfully demonstrated the feasibility of electronically exchanging important patient related information with the PCPs statewide. This is despite a lack of a link with their electronic health records. We have shown that using our ATAD technology, a PCP can be notified quickly of an important event such as a patient's asthma related emergency room admission so further follow up can happen in near real time.

Targeted screening for pediatric conditions with the CHICA system.

OBJECTIVE: The Child Health Improvement through Computer Automation (CHICA) system is a decision-support and electronic-medical-record system for pediatric health maintenance and disease management. The purpose of this study was to explore CHICA's ability to screen patients for disorders that have validated screening criteria--specifically tuberculosis (TB) and iron-deficiency anemia. DESIGN: Children between 0 and 11 years were randomized by the CHICA system. In the intervention group, parents were asked about TB and iron-deficiency risk, and physicians received a tailored prompt. In the control group, no screens were performed, and the physician received a generic prompt about these disorders. RESULTS: 1123 participants were randomized to the control group and 1116 participants to the intervention group. Significantly more people reported positive risk factors for iron-deficiency anemia in the intervention group (17.5% vs 3.1%, OR 6.6, 95% CI 4.5 to 9.5). In general, far fewer parents reported risk factors for TB than for iron-deficiency anemia. Again, there were significantly higher detection rates of positive risk factors in the intervention group (1.8% vs 0.8%, OR 2.3, 95% CI 1.0 to 5.0). LIMITATIONS: It is possible that there may be more positive screens without improving outcomes. However, the guidelines are based on studies that have evaluated the questions the authors used as sensitive and specific, and there is no reason to believe that parents misunderstood them. CONCLUSIONS: Many screening tests are risk-based, not universal, leaving physicians to determine who should have a further workup. This can be a time-consuming process. The authors demonstrated that the CHICA system performs well in assessing risk automatically for TB and iron-deficiency anemia.