Engineers embedded a modified melittin pore-forming peptide into an anti-PD-L1 antibody scaffold, creating a hybrid that remains inactive during transport and releases the tumor-killing peptide only when cleaved by a cancer-associated enzyme.
Latent-to-active tumor switchThe cytotoxic melittin peptide remains biologically dormant within the antibody scaffold and only becomes active when cleaved by matriptase — an enzyme overexpressed in carcinomas — enabling tumor-specific killing.
What the researchers found
A single-chain IgG (scIgG) was engineered from atezolizumab (anti-PD-L1) with flexible linkers embedding the melittin sequence flanked by matriptase cleavage sites. Initial constructs with native melittin were too cytotoxic during production, leading to development of Pmod2-2, a melittin variant that retained potent pore-forming ability while being compatible with antibody fusion.
The resulting scIgG-Pmod2-2 hybrid preserved the Fab (antigen binding) and Fc (immune signaling) functionalities of atezolizumab, displayed favorable pharmacokinetics, and released the active cytotoxic peptide specifically in response to matriptase — an enzyme overexpressed in carcinomas.
Why it matters
This work addresses one of the biggest challenges in peptide therapeutics: how to deliver a potent but toxic peptide selectively to tumors while keeping it inactive elsewhere. By combining an antibody's targeting precision with a pore-forming peptide's killing power and enzyme-activated release, this approach could enable a new class of cancer therapeutics that are both more effective and safer than either component alone.
How the study worked
Researchers used protein engineering to create single-chain IgG constructs based on atezolizumab with melittin sequences embedded in flexible linkers. Matriptase-cleavable flanking regions were added for tumor-specific activation. After initial constructs failed due to cytotoxicity during expression in producer cells, they designed Pmod2-2, a modified melittin variant. The hybrid was characterized for antibody functionality (Fab and Fc activity), pharmacokinetics, and protease-responsive peptide release.
What this study cannot tell us
This is an engineering and proof-of-concept study — anti-tumor efficacy in animal models is not reported in the abstract. The matriptase cleavage specificity in complex in vivo environments needs validation, as other proteases could potentially cause off-target release. The pharmacokinetics of the full hybrid may differ from standard antibodies due to the embedded peptide. Manufacturing scalability of this complex construct was not addressed.
How to read the evidence
This is a proof-of-concept protein engineering study published in Communications Biology (Nature portfolio). It demonstrates successful construct design and characterization, but lacks anti-tumor efficacy data from animal models, placing it at an early development stage.
When this study was published
Published in 2025, this is very recent work at the cutting edge of antibody-peptide hybrid engineering for cancer therapy.
The bigger picture
This study represents a convergence of three major therapeutic modalities: immune checkpoint antibodies (atezolizumab), cytotoxic peptides (melittin), and protease-activated prodrug design. The concept of hiding toxic payloads within antibody scaffolds for tumor-specific release could be broadly applied — not just with pore-forming peptides, but potentially with other cytotoxic or immunomodulatory peptides that are too toxic for systemic delivery.
Questions still open
- Does the scIgG-Pmod2-2 hybrid show superior anti-tumor efficacy compared to atezolizumab alone in animal cancer models?
- How specific is the matriptase-mediated activation in vivo — could other tissue proteases cause premature peptide release?
- Could this scaffold approach be adapted for other toxic therapeutic peptides beyond melittin variants?
Common questions
How does this antibody-peptide hybrid kill cancer cells?
Why can't pore-forming peptides be used directly as cancer drugs?
Read the original research
Engineering a single-chain immunoglobulin scaffold loaded with a latent-releasable cytotoxic pore-forming peptide.
Communications biology, 8(1), 1665
Citation
Morillo, Izaskun; Zulaica, Joao; R Caballero, Asier; Auzmendi-Iriarte, Jaione; Largo, Eneko; Apellaniz, Beatriz; Carracedo, Arkaitz; Piva, Marco; Nieva, José L; Rujas, Edurne. (2025). Engineering a single-chain immunoglobulin scaffold loaded with a latent-releasable cytotoxic pore-forming peptide.. Communications biology, 8(1), 1665. https://doi.org/10.1038/s42003-025-09066-9