{"id":2689,"date":"2026-08-25T10:00:00","date_gmt":"2026-08-25T14:00:00","guid":{"rendered":"https:\/\/www.insilens.com\/?p=2689"},"modified":"2026-08-25T19:45:02","modified_gmt":"2026-08-25T23:45:02","slug":"evobind-designs-macrocyclic-molecular-glues-from-sequence-alone","status":"publish","type":"post","link":"https:\/\/www.insilens.com\/?p=2689","title":{"rendered":"EvoBind Designs Macrocyclic Molecular Glues From Sequence Alone"},"content":{"rendered":"

Consortium:<\/strong> EvoBind-Multimer Academic Team · Modality:<\/strong> Sequence-First Macrocyclic Molecular Glues · Targets:<\/strong> VHL-KRAS, VHL-BRD4 · Date:<\/strong> August 21, 2026 (preprint)<\/p>\n

\"20260825_EvoBind_Multimer_Consortium_Technology_and_Modalities\"<\/p>\n

Summary<\/h4>\n

EvoBind-multimer is a sequence-first generative framework for designing macrocyclic peptides that bridge an E3 ligase to a chosen target without a known ligand or pre-existing protein-protein interface. The authors report VHL-recruiting designs for KRAS and BRD4 with live-cell proximity, proteasome-dependent target modulation and activity in patient-derived neuroblastoma tumoroids. The central translational warning is context dependence: an identical macrocycle behaved as a degrader in one model and a VHL-dependent stabilizing “LOCKTAC” in another.<\/p>\n

What Happened<\/h4>\n

The framework repurposes a single-chain AlphaFold2 representation to co-model two proteins and a cyclic peptide bridge. Designs were selected computationally and tested with NanoBRET proximity assays, target-abundance readouts, proteasome inhibition and downstream signaling measurements; the work is a preprint and has not undergone peer review. Macrocycles targeting VHL-KRAS and VHL-BRD4 formed ternary complexes in cells, and proteasome inhibition increased ternary-complex signal for active compounds, supporting Cullin-RING-ligase involvement. Patient-derived tumoroid experiments showed biological activity but also divergent processing of the same ternary complex across models. The disclosure does not establish systemic exposure, cell permeability across tissues, oral bioavailability, pharmacokinetics, tolerability, selectivity across the proteome, in-vivo tumor delivery, manufacturing yield or a development candidate.<\/p>\n

Deep Analysis<\/h4>\n

Mechanistic novelty lies in designing an induced interface from sequence rather than optimizing a discovered glue; if generalizable, the approach could expand addressable targets and E3 ligases. Yet ternary-complex formation is not equivalent to productive degradation, since ubiquitin geometry, lysine accessibility, complex residence time, proteasomal routing and cellular state all determine fate. The degrader-versus-stabilizer switch is more than noise: it may enable programmable target sequestration, but it also makes efficacy direction difficult to predict from structure alone and creates a safety risk when target stabilization would be harmful. Prospective rules that predict degradation, sequestration or no effect are essential before this becomes a reliable design platform.<\/p>\n

Translation requires a full chain from computational hit rate to medicinal chemistry, permeability, exposure, tumor distribution, pharmacodynamics, off-target proteomics, resistance, immunogenicity and scalable macrocycle synthesis. Patient-derived tumoroids strengthen disease relevance but do not reproduce whole-body pharmacokinetics, immune interactions or dose-limiting toxicity.<\/p>\n

Company and Product Background<\/h4>\n

EvoBind-multimer was developed by an academic consortium led by researchers in computational protein design and cancer biology. Molecular glues are monovalent compounds that stabilize a new protein interface, often recruiting an E3 ligase to alter target abundance or function. VHL is a Cullin-RING E3-ligase substrate receptor frequently used in targeted protein degradation; KRAS and BRD4 are oncology targets with distinct structures and cellular roles. Macrocyclic peptides can cover larger interfaces than conventional small molecules, but intracellular access, metabolic stability and manufacturability remain recurring constraints.<\/p>\n

Signal Extraction<\/h4>\n\n\n\n\n\n\n\n
Factor<\/th>\nAssessment<\/th>\n<\/tr>\n
Design Scope<\/td>\nSequence-based generation without a known ligand or holo interface<\/td>\n<\/tr>\n
Cell Evidence<\/td>\nKRAS\/BRD4 proximity and proteasome-dependent target modulation<\/td>\n<\/tr>\n
Patient Models<\/td>\nActivity in patient-derived neuroblastoma tumoroids (ex vivo only)<\/td>\n<\/tr>\n
Context Dependence<\/td>\nSame macrocycle degraded in one model, stabilized in another<\/td>\n<\/tr>\n
Development Readiness<\/td>\nComputational\/experimental proof of concept; no candidate, PK, or CMC data<\/td>\n<\/tr>\n<\/table>\n

Reading the Signal<\/h4>\n

Bull case:<\/strong> Sequence-first interface design could convert induced-proximity discovery from serendipity into a programmable modality, with cross-target cellular validation supporting generalizability.<\/p>\n

Bear case:<\/strong> The number of tested targets and compounds is small, context-dependent outcomes undermine predictability, and no in-vivo, PK, or selectivity data exist to support translation beyond a hit-generation tool.<\/p>\n

InSilens Take<\/h4>\n

This work demonstrates a credible route to de novo macrocyclic induced proximity, not a treatment-ready product. The observed switch between degradation and stabilization is both the platform’s most interesting capability and its most important control problem, and resolving it will determine whether EvoBind becomes a reliable design engine or remains a research curiosity.<\/p>\n

Signal Assessment<\/h4>\n

Importance:<\/strong> 4\/5 · Direction:<\/strong> Uncertain · Confidence:<\/strong> Moderate-High on facts, Low-Moderate on interpretation<\/p>\n","protected":false},"excerpt":{"rendered":"

EvoBind-multimer is a sequence-first generative framework for designing macrocyclic peptides that bridge an E3 ligase to a chosen target without a known ligand or pre-existing protein-protein interface. The authors report VHL-recruiting designs for KRAS…<\/p>\n","protected":false},"author":1,"featured_media":2696,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[434,435,380],"class_list":["post-2689","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-technology-modalities","tag-evobind","tag-macrocyclic-peptides","tag-molecular-glue"],"blocksy_meta":[],"_links":{"self":[{"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/posts\/2689","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.insilens.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=2689"}],"version-history":[{"count":1,"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/posts\/2689\/revisions"}],"predecessor-version":[{"id":2701,"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/posts\/2689\/revisions\/2701"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.insilens.com\/index.php?rest_route=\/wp\/v2\/media\/2696"}],"wp:attachment":[{"href":"https:\/\/www.insilens.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=2689"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.insilens.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=2689"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.insilens.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=2689"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}