Drug discovery has long grappled with a fundamental tension in high-throughput screening: the more biologically relevant your model, the harder it is to scale. Phenotypic assays in primary disease-relevant cells offer rich biological context, but capturing meaningful, target-specific readouts from these complex systems at screening scale has remained a significant challenge. In contrast, simpler, more scalable systems are easier to deploy but sacrifice the biological fidelity that makes hits meaningful. A recent study by Samowitz et al. in Nature Communications describes an interesting approach to resolve this tension, the Endo-GeneScreen (EGS) platform. A high-throughput screening system designed to enable scalable detection of endogenous protein levels within disease-modeling cellular contexts.
A Well-Chosen Proof of Concept
The authors selected Syngap1 as their proof-of-concept target to develop and demonstrate this approach. De novo mutations in this gene that lead to haploinsufficiency are among the most common genetic causes of sporadic neurodevelopmental disorders, including intellectual disability, autism, and epilepsy. Small molecules that boost SynGAP protein levels back toward wildtype would address the root cause of these disorders rather than managing downstream symptoms. Importantly, Syngap1 function is closely tied to cortical excitatory neurons. Well-validated in vitro and in vivo models for these neurons already exist, creating an integrated system for both discovering new compounds and validating them in the same biological context. That continuity is an important step toward improving the translational relevance of lead molecules coming out of the screen.
The choice to pursue small molecule modulators of Syngap1 protein levels was also deliberate. Small molecules can be iteratively improved through medicinal chemistry: optimizing brain penetrance, selectivity, and dosing flexibility over successive rounds of synthesis and testing. That iterative potential is central to what makes a phenotypic hit a valuable starting point.
Building a Screening-Compatible Model Without Sacrificing Biology
To enable screening at this scale within a biologically relevant context, the authors built their platform around endogenous protein detection in primary cortical neurons. Using standard CRISPR methods, they knocked the small 11-amino acid HiBiT tag into the endogenous Syngap1 locus, creating a Syngap1-HiBiT knock-in mouse strain. Because the tag is inserted at the endogenous locus rather than expressed from a transgene, protein levels reflect native regulation rather than artificial overexpression. During the screen, the HiBiT complementation partner LgBiT is added to the cells during the detection step, reconstituting active NanoBiT® luciferase and generating a luminescent signal directly proportional to endogenous SynGAP protein levels.
The authors took the model a step further by incorporating a second firefly luciferase (Fluc) reporter. By crossing the HiBiT knock-in mouse with a commercially available transgenic mouse line constitutively expressing Fluc from a ubiquitous promoter, they created an “in-mouse” Dual Luciferase Reporter (DLR) assay. The Nano-Glo® HiBiT Dual-Luciferase® Reporter System was used to sequentially detect first the Fluc signal, reporting on global changes in total protein and cellular toxicity, followed by the HiBiT signal reporting on SynGAP protein levels. This built-in counterscreen is what allows the platform to distinguish compounds that specifically boost SynGAP from those that simply increase all proteins, as well as identifying cytotoxic compounds. Per-well normalization using the median Fluc and HiBiT signals from negative controls further enabled a hit detection algorithm that reduced false positives without sacrificing sensitivity. The assay system leverages the simple add-mix-read workflow, low background, and wide dynamic range that bioluminescent reporter systems deliver at scale.
From Screen to Validated Hits
Two independent screens totaling more than 100,000 compounds identified over 40 validated small molecules capable of boosting endogenous SynGAP protein in haploinsufficient neurons. The lead compound, SR-1815, was characterized extensively as a proof-of-concept for the full platform workflow. Beyond the primary DLR assay, the team confirmed SR-1815’s activity using an orthogonal Dot Blot protein assay employing knockout-validated, isoform-specific antibodies against SynGAP, confirming that SR-1815 raised all isoforms proportionally. SR-1815 also demonstrated functional activity, rescuing elevated excitatory synapse strength and neuronal hyperexcitability in haploinsufficient neurons in a genotype-specific manner, consistent with restoring SynGAP’s role in synaptic regulation.
Early Results and the Road Ahead
SR-1815 currently faces real pharmacological challenges, notably poor brain penetrance due to P-glycoprotein-mediated efflux and a short half-life, which limits in vivo evaluation until optimized analogs are available. This is expected at this stage, and it is precisely the kind of problem that iterative medicinal chemistry is designed to solve. Interestingly, target deconvolution studies described in a companion publication identified SR-1815 as a multikinase inhibitor that regulates SynGAP levels through alternative splicing, with some targets also implicated in cancer biology. This is an instructive reminder that phenotypic hits can open biological doors well beyond their original indication, and a strong argument for investing in mechanism-of-action work alongside the medicinal chemistry program.
The ~40 additional validated SynGAP-boosting compounds from EGS are now entering early preclinical evaluation, each representing a distinct chemical scaffold and potential therapeutic starting point.
Conclusion
The Endo-GeneScreen platform is a convincing demonstration that biological relevance and screening scale are not mutually exclusive, but achieving both requires the right technology at the foundation. By combining endogenous HiBiT tagging with the sensitivity and dynamic range of bioluminescence and embedding that within a dual-reporter system that handles selectivity and toxicity in the same assay read, the authors built a platform designed to find the right hits in the right context from the start. Because the HiBiT tag can be knocked into any gene of interest using standard CRISPR methods, and the resulting knock-in mouse line bred with the existing Fluc line, the same infrastructure is immediately extensible to other disease targets, making EGS a modular starting point for physiologically relevant screening well beyond Syngap1. The hits that emerge from this kind of approach carry the credibility of having been found in the right cellular context, jumpstarting next-step translational research efforts.
To learn more about HiBiT tagging technology for building physiologically relevant model systems, visit: HiBiT Protein Tagging System.
