Matrix metalloproteinase-2 (MMP-2) is a key enzyme involved in extracellular matrix remodeling and is implicated in several pathological conditions, including cancer and neurodegenerative diseases. Peptide-based inhibitors such as APP-IP represent a promising strategy for selectively targeting MMP-2 while limiting off-target effects. In this study, we used molecular docking, molecular dynamics simulations, and MM/PBSA calculations to evaluate how targeted mutations in APP-IP affect binding affinity and selectivity. Mutations were introduced at three positions (Y3, G4, and N5) of the peptide. Screening against MMP-2 identified two variants, Gly4 to Leu (G4L) and Asn5 to Thr (N5T), with predicted improved binding affinity relative to the native peptide, while mutations at position Y3 did not significantly alter affinity but were retained due to the known importance of this position for selectivity in accommodating S1´ pocket. To assess off-target effects, selected variants were further tested against MMP-9, a close homolog of MMP-2, and MMP-7, a more structurally distinct family member. The N5T variant showed reduced binding to MMP-9 while maintaining enhanced affinity for MMP-2, indicating improved selectivity trends. In contrast, G4L did not significantly alter binding to either off-target proteases. Mutations at position Y3 (Y3F) preserved affinity for MMP-2 while consistently reducing binding to both MMP-9 and MMP-7, highlighting the role of the S1′ pocket in modulating selectivity. Simulations with MMP-7 confirmed intrinsically weaker binding of APP-IP and its variants, consistent with experimental observations, and showed no enhancement of off-target affinity. Overall, this structure-based approach enabled the identification of promising candidate variants with favorable predicted selectivity profiles, with N5T, G4L, and Y3F emerging as candidates for further experimental validation. Importantly, among all variant, N5T is particularly notable, as in our comparative prediction report, it not only improved affinity in MMP-2 but also enhanced selectivity trends by reducing binding to MMP-9 while maintaining a favorable profile in MMP-7. Also at position Y3, although the overall binding affinity was largely retained, the observed reduction in off-target binding against both MMP-9 and MMP-7 underscores the importance of the S1′ pocket, suggesting that subtle modifications here can modulate specificity. This research offers insights into the rational design of peptide-based inhibitors for MMP-2 and emphasizes the value of extending computational selectivity testing across both close and divergent MMPs. Future studies could further investigate these findings through combined mutation analyses, residue-level structural assessments, and the application of more rigorous free energy methods such as QM/MM-PBSA. However, these findings represent computationally predicted comparative trends and require experimental validation to confirm their biological relevance.

Computational evaluation of APP-IP peptide variants: assessing affinity and selectivity toward MMP-2

期刊论文
发表日期：

2026年5月26日

查看全文

Samira Mirkhalaf,
Majid Taghdir,
Seyed Shahriar Arab

摘要

Matrix metalloproteinase-2 (MMP-2) is a key enzyme involved in extracellular matrix remodeling and is implicated in several pathological conditions, including cancer and neurodegenerative diseases. Peptide-based inhibitors such as APP-IP represent a promising strategy for selectively targeting MMP-2 while limiting off-target effects. In this study, we used molecular docking, molecular dynamics simulations, and MM/PBSA calculations to evaluate how targeted mutations in APP-IP affect binding affinity and selectivity. Mutations were introduced at three positions (Y³, G⁴, and N⁵) of the peptide. Screening against MMP-2 identified two variants, Gly⁴ to Leu (G4L) and Asn⁵ to Thr (N5T), with predicted improved binding affinity relative to the native peptide, while mutations at position Y³ did not significantly alter affinity but were retained due to the known importance of this position for selectivity in accommodating S1´ pocket. To assess off-target effects, selected variants were further tested against MMP-9, a close homolog of MMP-2, and MMP-7, a more structurally distinct family member. The N5T variant showed reduced binding to MMP-9 while maintaining enhanced affinity for MMP-2, indicating improved selectivity trends. In contrast, G4L did not significantly alter binding to either off-target proteases. Mutations at position Y³ (Y3F) preserved affinity for MMP-2 while consistently reducing binding to both MMP-9 and MMP-7, highlighting the role of the S1′ pocket in modulating selectivity. Simulations with MMP-7 confirmed intrinsically weaker binding of APP-IP and its variants, consistent with experimental observations, and showed no enhancement of off-target affinity. Overall, this structure-based approach enabled the identification of promising candidate variants with favorable predicted selectivity profiles, with N5T, G4L, and Y3F emerging as candidates for further experimental validation. Importantly, among all variant, N5T is particularly notable, as in our comparative prediction report, it not only improved affinity in MMP-2 but also enhanced selectivity trends by reducing binding to MMP-9 while maintaining a favorable profile in MMP-7. Also at position Y³, although the overall binding affinity was largely retained, the observed reduction in off-target binding against both MMP-9 and MMP-7 underscores the importance of the S1′ pocket, suggesting that subtle modifications here can modulate specificity. This research offers insights into the rational design of peptide-based inhibitors for MMP-2 and emphasizes the value of extending computational selectivity testing across both close and divergent MMPs. Future studies could further investigate these findings through combined mutation analyses, residue-level structural assessments, and the application of more rigorous free energy methods such as QM/MM-PBSA. However, these findings represent computationally predicted comparative trends and require experimental validation to confirm their biological relevance.