Brain-Penetrating Bioconjugate M13-CTX-PEP-1 for Targeted Glioblastoma Therapy In Silico

Video Transcript:

Glioblastoma is the most common malignant brain tumor in adults. Current treatments for GBM are impeded by the Blood Brain Barrier, a semi-permeable membrane surrounding the brain and responsible for molecular transport.

Because this barrier is highly selective, it blocks around 98% of small molecules from ever accessing the brain, rendering chemotherapeutic drugs largely ineffective.

This study aims to combat this issue by developing a BBB-permeable, highly selective drug.

Bacteriophage M13 was used as a delivery platform for the molecules PEP-1 and Chlorotoxin. Chlorotoxin is a peptide with a compact structure that allows for the easy passage of M13 through the blood brain barrier.

PEP-1 is a synthetic peptide which has shown strong targeting capabilities for cancer protein Il13Ra2.

Once the drug-like complex reaches cancer protein Il13Ra2, PEP-1 binds to the protein, triggering internalization of the bound complex. The complex can then be degraded or migrated back to the surface, where it will be internalized by more attacking phages.

The loss of Il13 causes major inhibition in GBM cell metastasis, leading to cell death.

Introduction:

Glioblastoma (GBM) is the most common malignant brain tumor in adults, accounting for up to 50% of all cases (Fig 1.0) and presenting with a 5-year survival rate of roughly 7% (Pillay, 2017). GBM is characterized by the rapid proliferation and aggressive metastasis of astrocytes, cells responsible for maintaining homeostasis and structural integrity in the brain.

Current chemotherapeutic drugs are often impeded by the Blood-Brain Barrier (BBB), a semi-permeable layer of endothelial and glial cells protecting the brain. Tight gap junctions and limited transport systems in the BBB create a highly selective environment, preventing 98% of small molecule drugs from crossing into the central nervous system (Dynarowicz et al., 2026).

Bacteriophage M13 (Fig 2.0) is a virus known for its ability to cross the BBB with ease given its rod-like structure, which triggers receptor-mediated transcytosis (Haqqani et al., 2024). This study aims to utilize bacteriophage M13 computationally fused with cell-penetrating peptide PEP-1 (Fig 3.0) and Leiurus quinquestriatus (deathstalker scorpion) venom–derived toxin Chlorotoxin (CTX) (Fig 4.0) for the selective targeting of Il13Rα2.

Il13Rα2 (Fig. 5.0) is a surface protein overexpressed uniquely on GBM cells and responsible for facilitating metastasis; studies have demonstrated that the binding of large molecules such as M13 causes internalization of the protein-drug complex, leading to the lysosomal degradation of Il13Rα2 and cell death (Sattiraju et al., 2017). CTX is known for its ability to breach the BBB, making it a crucial mediator in this approach.

PHASE I - M13-CTX Conjugate construction:

CTX's 3D structure was accessed via PDB (Fig 6.0); M13's pIII protein's amino acid sequence was inputted into AlphaFold's protein folding server for 3D reconstruction. Both molecules were optimized in UCSF Chimera via (i) the removal of solvents, (ii) the addition of polar-only Hydrogen atoms, (iii) Gasteiger/AMBER charges, (iv) energy minimization, and (v) backbone strengthening via calcium ions (Fig 7.0). They were then loaded onto MolModa and molecular docking was conducted a total of 5 times on the Top 3 binding sites determined by binding affinity, druggability, and solvent-accessible surface area (SASA) (Fig 8.0).

PHASE II - M13-PEP-1 Dual-functionalization validation:

PEP-1's amino sequence was derived from literature (Muñoz-Morris et al., 2007) and reconstructed via AlphaFold. Following optimization, it was bound with M13 pIII in MolModa. Binding sites occupied by CTX were excluded from the list of available sites.

PHASE III - Il13Ra2-PEP-1 targeting:

Il13Ra2 was retrieved from UniProt (Fig 9.0) and optimized via the addition of AMBER's ff14SB force field; it was then inputted into MolModa for binding to PEP-1, which was dyed purple for ease of identification.

GRAMM Docking was employed to confirm the validity of docking in each Phase and was able to replicate results with 97% accuracy measured via comparative docking scores. The accuracy of reconstructed models was also obtained via AlphaFold's residue error plot and confidence scoring key (Fig 10.0).

Drug-likeness assays:

PEP-1 was uploaded onto ToxinPred for the determination of toxicity and the discovery of 400 mutant peptides. BBB Predictor was used to model gastrointestinal absorption and BBB permeability values for PEP-1; furthermore, the drug-like compliance tests Lipinski's Rule of 5, Ghose Filter, and Veber, were conducted through SwissADME.

Note: For identification purposes, the following dyes were applied: M13 - Green; CTX - Blue; PEP-1 - Purple; Il13Ra2 - Red.

M13-CTX Phase I binding:

A total of 35 binding "hot spot" pockets were detected on M13's minor coat protein pIII's structure via MolModa (Fig 11.0). The Top 3 binding spots identified via the pocket selection criteria were P_1 (highest predicted affinity), P_6 (highest druggability) and P_15 (highest SASA). The number of ligand conformational poses obtained were low given CTX's limited flexibility (16 pose average per pocket); however, the high docking scores (avg. -12) indicated M13-CTX's strong potential as a stable drug delivery platform (P_6 had the "highest" score of -12.8, followed by P_15 at -12.53 and P_1 at -12.51).

Note: binding score values are inversely proportional to their strength.

M13-PEP-1 Phase II binding:

The Top 3 binding sites identified were P_2, P_16, and P_21, with P_2 showing the strongest docking score of -10.52, the most conformational poses at 21, and high pocket score stability as opposed to P_16, which provided a range of values (-10 to -5) and P_21, which showed moderately strong scores averaging -7. P_2's strong binding values demonstrate M13-CTX-PEP-1's ability to function as a bioconjugate drug (Fig 12.0, 13.0).

Il13Ra2-PEP-1 Phase III binding:

Only two binding pockets were considered for docking (P_1 and P_8) given the exclusion of druggability in the selection criteria. P_1 demonstrated fluctuations in docking scores between trials, with the most stable trial providing a docking score of -7 across all 5 runs; however, P_6 (Fig 14.0) provided stable binding values with the highest recorded score of -16.01, suggesting its strong targeting capabilities for Il13Ra2.

PEP-1 drug-likeness testing:

SwissADME testing indicated moderate-high BBB permeability and high gastrointestinal absorption in PEP-1 (Fig 15.0); furthermore, PEP-1 passed the Lipinski's Rule of 5 and Veber drug compliance tests with one noted violation in the Ghose filter test (molecular weight > 480 Da). However, this is hypothesized to be unproblematic given pIII's carrying capacity of 30 kDa.

Failure Analysis:

MolModa's docking parameters had to be manually adjusted to accommodate for PEP-1's large number of rotatable bonds (>15), increasing runtime from 15 minutes to 45 minutes per docking trial. Furthermore, due to the size of M13 pIII, only moderate energy minimization could be performed (i.e. configuring 100 steps of steepest descent during optimization to 70 and decreasing the number of Monte Carlo iterations by 15%). Il13Ra2's outward "arm-like" projections were folded with low confidence given the varying angles at which they can exist; the lack of binding pockets on these arms, however, suggests molecular docking in Phase III remained unaffected.

Conclusions:

M13-CTX-PEP-1 (Fig 16.0) has demonstrated strong potential as a bioconjugate drug for the systemic phage-mediated targeting of Il13Ra2 in glioblastoma cells. CTX and PEP-1’s high affinity towards M13 validate their creation as a drug; PEP-1’s high HIA, flexibility, and BBB permeability, paired with its validation by the Ghose filter, Veber, and Lipinski’s Rule of 5 drug-like compliance tests enables its use as an effective BBB-penetrating peptide.

Furthermore, PEP-1’s very strong affinity for Il13Ra2 and the encouraging number of ligand conformational poses generated (23 average) between them indicates strong targeting capabilities.

However, in vitro research must be conducted in order to definitively test the targeting capabilities of PEP-1 to Il13Ra2; as well, lab techniques must be employed to test the validity of M13-CTX-PEP-1’s assembly. Future in silico research should explore the use of PEP-1’s mutant peptides as alternatives or additional conjugated peptides in therapy design. Standardization in molecular optimization and computational phage design must also be conducted in order to increase methodological rigor and external validity.

Cognizant of these current limitations and future endeavors, M13-CTX-PEP-1 holds strong potential in the systemic inhibition of Il13Ra2 function in GBM cells with minimal predicted cytotoxic effects and high BBB-penetration (Fig 17.0), opening a new avenue in targeted GBM treatment.

Note: A library of 400 identified mutant PEP-1 peptides was published on the research data repository Zenodo including the following associated physicochemical properties:

Mutation Position

SVM Score

Toxicity Prediction

Hydrophobicity

Steric hindrance

Sidebulk

Molecular weight

Amphipathicity

Future Improvements:

Modelling the GBM tumor microenvironment (Fig 18.0) in silico and in vitro for off-target effects and drug selectivity testing:

Cytokines, cell surface receptors, and signaling pathways 3D reconstruction.

PEP-1 molecular docking to TME components for affinity prediction.

Constructing a phage peptide library (Fig 19.0) with mutated M13 variants for systemic ligand selection and optimization:

Degenerate oligonucleotide pool synthesis.

Kunkel mutagenesis phage vector preparation.

Phage particle amplification and purification.

E. coli-based biopanning for mutant selection.

Conducting molecular docking trials and drug-like compliance tests for identified peptides:

Lipinski's Rule of 5

Ghose Filter

Veber

BOILED-egg BBB/HIA plot

I would like to extend my gratitude to Dr. Osei at Grand River Hospital for peer-reviewing my accompanying research paper, as well as Dr. Sarah Abubakr at Buffalo General Hospital and Dr. Ahmad Hanif at Geisinger Medical Centre for their advice and encouragement. It would also be impossible to ignore the unending support from the WWSEF-CWSF team. And to all the people battling cancer, you are the real heroes.

Never stop believing.

References:

Abubakr, A. (2026). PEP-1 Mutated Peptides and their Physicochemical Properties Determined through ToxinPred [Data set]. Zenodo. https://doi.org/10.5281/zenodo.19410757

Phage display peptide library platform. Explore Phage Display Peptide Library Platform for Advanced Research - Alpha Lifetech. (2026). https://www.alpha-lifetech.com/phage-display-peptide-library-platform/

Dynarowicz, K., Aebisher, D., Tylutki, J., Kozak, N., Kawczyk-Krupka, A., & Bartusik-Aebisher, D. (2026, February 6). Small particles, big impact: Inorganic nanotechnology for glioblastoma. Molecules (Basel, Switzerland). https://pmc.ncbi.nlm.nih.gov/articles/PMC12899315/

Gupta et. al. In silico Approach for Predicting Toxicity of Peptides and Proteins. PLoS ONE 8(9):e73957. doi:10.1371/journal.pone.0073957

Haqqani, A. S., Bélanger, K., & Stanimirovic, D. B. (2024, March 12). Receptor-mediated transcytosis for brain delivery of therapeutics: Receptor classes and Criteria. Frontiers in drug delivery. https://pmc.ncbi.nlm.nih.gov/articles/PMC12363266/

Jia, Q., & Xiang, Y. (2023, September 5). Cryo-em structure of a bacteriophage M13 mini variant. Nature News. https://www.nature.com/articles/s41467-023-41151-7

Lakshmanachetty, S., & Mitra, S. S. (2022a, May 27). Mapping the tumor-infiltrating immune cells during glioblastoma progression. Nature News. https://www.nature.com/articles/s41590-022-01223-0

Mayo Foundation for Medical Education and Research. (2026, March 5). Glioblastoma. Mayo Clinic. https://www.mayoclinic.org/diseases-conditions/glioblastoma/survival-rates/gnc-20596050

Muñoz-Morris, M. A., Heitz, F., Divita, G., & Morris, M. C. (2007). The peptide carrier Pep-1 forms biologically efficient nanoparticle complexes. Biochemical and biophysical research communications, 355(4), 877–882. https://doi.org/10.1016/j.bbrc.2007.02.046

Neurosnap: Online bioinformatics tools – fast, easy, code-free. Neurosnap | Online Bioinformatics Tools – Fast, Easy, Code-Free. (n.d.). https://neurosnap.ai/

Pillay, P. (2017, July 26). Brain Tumor Statistics. Specialising in advanced treatment for Brain & Spine Disorders, Singapore Brain Center. https://singaporebrain.org/brain/brain-tumor-statistics/

Phage display peptide library platform. Explore Phage Display Peptide Library Platform for Advanced Research - Alpha Lifetech. (2026). https://www.alpha-lifetech.com/phage-display-peptide-library-platform/

Sattiraju, A., Solingapuram Sai, K. K., Xuan, A., Pandya, D. N., Almaguel, F. G., Wadas, T. J., Herpai, D. M., Debinski, W., & Mintz, A. (2017, June 27). IL13RA2 targeted alpha particle therapy against glioblastomas. Oncotarget. https://pmc.ncbi.nlm.nih.gov/articles/PMC5522122/

UCSF Chimera - a visualization system for exploratory research and analysis. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, Ferrin TE. J Comput Chem. 2004 Oct;25(13):1605-12.