CIP-3

What is CIP-3?

CIP-3 is an AI-designed cyclic peptide CD28 receptor antagonist reported by Kuncewicz, Gabr, and colleagues at Weill Cornell Medicine (with collaborators at Macao Polytechnic University and the University of Gdansk) in Advanced Science in May 2026. CD28 is a costimulatory receptor on T cells that, together with TCR engagement, drives T cell activation. CD28-CD80/86 antagonism is the mechanism of abatacept (Orencia, FDA-approved 2005 for rheumatoid arthritis and other autoimmune indications) — a CTLA4-Fc fusion protein. CIP-3 achieves similar functional immune attenuation through a much smaller cyclic peptide format, with nanomolar CD28 binding affinity, disruption of the CD28-CD80/86 interaction, rapidly reversible suppression of T-cell activation, and dose-dependent efficacy in a T-cell transfer colitis model comparable to a benchmark anti-CD28 antibody. The implications of small-peptide chemistry include potentially oral bioavailability and lower immunogenicity than protein-based CTLA4-Fc agents. As of mid-2026, CIP-3 is preclinical — the molecule is a lead compound demonstrating that AI-designed cyclic peptides can produce clinically-relevant immune-checkpoint antagonism, rather than an advanced clinical candidate.

What CIP-3 Is Investigated For

CIP-3 is a preclinical-to-early-clinical CD28 cyclic peptide antagonist demonstrating the application of AI-designed peptide therapeutics to immunology. Mechanistically the target (CD28 costimulation blockade) has been clinically validated by abatacept (CTLA4-Fc) in rheumatoid arthritis, psoriatic arthritis, JIA, and transplant rejection. The compact cyclic peptide format aims to retain CD28 antagonism while gaining potential advantages of small peptide chemistry (oral bioavailability, lower immunogenicity, simpler manufacturing). The honest caveats: CIP-3 is at a substantially earlier development stage than approved CTLA4-Fc-based agents, AI-designed peptide therapeutics have generally not advanced to FDA approval as of mid-2026, and the translation from in silico design through preclinical to clinical efficacy in heterogeneous autoimmune patient populations has historically been difficult. The molecule is more notable as a proof-of-concept for AI peptide design in immunology than as a near-term clinical candidate.

History & Discovery

AI-designed peptide therapeutics emerged as a research field through the late 2010s and early 2020s, with multiple academic and commercial programs (including foundational de novo protein and peptide design work at the Baker lab / University of Washington's Institute for Protein Design and downstream commercial spinoffs) producing methodology that translated into disease-target applications. The application to CD28 antagonism reflects both the clinical relevance of the target (validated by abatacept's commercial success and broad use in rheumatoid arthritis, psoriatic arthritis, JIA, and transplant rejection) and the technical fit of the problem (a defined protein-protein interaction amenable to cyclic peptide design). CIP-3 was first reported by Kuncewicz, Upadhyay, Zhu, Duan, and Gabr in Advanced Science in May 2026 (PMID 42216689) — an AI-designed cyclic peptide that binds CD28 with nanomolar affinity, disrupts the CD28-CD80/86 interaction, rapidly reversibly suppresses T-cell activation, and shows dose-dependent efficacy in a T-cell transfer colitis model comparable to a benchmark anti-CD28 antibody. The Gabr lab at Weill Cornell Medicine's Molecular Imaging Innovations Institute (Department of Radiology) led the work, in collaboration with the Macao Polytechnic University group of Duan and the Kuncewicz group at the University of Gdansk. As of mid-2026, CIP-3 remains preclinical — the paper represents a methodology demonstration for AI-designed cyclic peptide immune-checkpoint antagonism rather than an advanced clinical candidate. Whether CIP-3 specifically or its successors reach FDA approval will depend on translation through Phase 1/2/3 trials — a path that no AI-designed therapeutic cyclic peptide has yet completed.

How It Works

Your immune T cells need two signals to activate fully — one through the T cell receptor (recognizing the threat) and one through CD28 (a costimulatory 'go' signal). Blocking the CD28 signal calms the immune response, which is useful in autoimmune diseases where the immune system attacks the body's own tissues. CIP-3 is a small cyclic peptide designed by AI to do this — similar in mechanism to abatacept but in a much smaller molecule that could potentially be taken orally.

CD28 is a costimulatory receptor on T cells that binds CD80 (B7-1) and CD86 (B7-2) on antigen-presenting cells. CD28 engagement provides the second signal (alongside TCR-MHC engagement) required for full T cell activation, IL-2 production, and clonal expansion. Without CD28 costimulation, T cells receiving TCR signal alone enter an anergic state. Abatacept and the second-generation belatacept (more selective for B7-1) achieve immunosuppression by competitively binding CD80/86 (their endogenous ligands), preventing endogenous CD28 from engaging them. CIP-3 takes a different antagonist strategy — directly binding CD28 itself (rather than the CD80/86 ligands) to prevent the costimulatory interaction. The cyclic peptide design constrains the peptide into the receptor-binding conformation, achieving sufficient binding affinity for clinical antagonism with a much smaller molecular size than the protein-based CTLA4-Fc agents. The AI design approach uses computational protein design to generate cyclic peptides with optimized binding affinity, selectivity, and biophysical properties for the CD28 receptor target. Multiple cycles of computational design, screening, and biochemical/cellular validation produce candidate molecules; CIP-3 is one such optimized candidate.

Evidence Snapshot

Research Gaps & Open Questions

What the current literature has not yet settled about CIP-3:

• 01Phase 1 clinical data — pending if/when the molecule advances to Phase 1.
• 02Translation of AI-designed peptide therapeutics to FDA approval — no AI-designed peptide has yet completed regulatory approval as of mid-2026.
• 03Comparative efficacy versus abatacept and belatacept in head-to-head data.
• 04Route of administration optimization — whether oral bioavailability is achievable.
• 05Long-term immunosuppressive safety profile.
• 06Specific autoimmune indication where CIP-3 would be positioned remains to be established.

Forms & Administration

Route to be established — subcutaneous injection most likely for early clinical work; oral bioavailability is a longer-term goal of cyclic peptide design.

Dosing & Protocols

The ranges below reflect protocols commonly discussed in the literature and by clinicians — not a prescription. Actual dosing for any individual should be determined by a qualified healthcare provider who knows the patient.

CIP-3 is preclinical/early clinical and not approved for any indication.

Timeline of Effects

Common Questions

Who CIP-3 Is NOT For

Drug & Supplement Interactions

Specific drug interactions for CIP-3 will be characterized in clinical development. The CD28 antagonist mechanism implies general considerations for concurrent immunosuppressive therapy (combination may be additive in immune attenuation, increasing infection risk) and live vaccines (typically avoided during immune attenuation).

Safety Profile

Legal Status

Regulatory status changes over time. Verify current local rules with a qualified professional.

Myths & Misconceptions