Evaluating the effect of minimal TIMP variants on protecting and transport across the rat brain microvascular cells (RBMEC) Authors: Elham Taheri, Maryam Raeeszadeh-Sarmazdeh Published: 2025 ABSTRACT AbstractTissue inhibitors of metalloproteinases (TIMPs), endogenous inhibitors of matrix metalloproteinases (MMPs), can be tailored to regulate MMP activity and mitigate the disruptive effects of specific MMPs when dysregulated in diseases. MMPs, especially MMP-9, are major contributors to the degradation of extracellular matrix components, leading to BBB disruption in neurological disorders. The upregulation of MMPs undermines blood-brain barrier (BBB) integrity and drives neuroinflammation. Engineering minimal protein variants offers enhanced modularity, tissue penetration, and BBB permeability. Minimal TIMP variants were engineered, aiming to improve their therapeutic reach across both sides of the BBB, particularly when delivery to the brain is essential. In this study, we assessed the protective effects of mTC1 and mTC3 on BBB integrity using an in vitro model of rat brain microvascular endothelial cells (RBMECs). Barrier function was evaluated following treatment with recombinant MMP-9, either alone or co-treated with native TIMP-1, TIMP-3, or the engineered minimal variants. MMP-9 induced a dose-dependent increase in BBB permeability, reflected by a decrease in trans-endothelial electrical resistance (TEER) and increased paracellular transport of fluorescent tracers. Co-treatment with TIMP-1, TIMP-3, mTC1, or mTC3 significantly attenuated MMP-9-mediated disruption of tight junctions of RBMECs, preserving TEER values and reducing permeability. Immunofluorescence staining for tight junction proteins, ZO-1 and occludin, further validated the preservation of endothelial integrity in the presence of wild-type human TIMPs and engineered TIMP variants. These findings underscore the potential of engineered minimal TIMPs as molecular tools to stabilize the BBB and support their future application in mechanistic studies focused on BBB protection.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-30643-9.Keywords: Blood-brain-barrier (BBB), Matrix metalloproteinase (MMP), Tissue inhibitor of metalloproteinase (TIMP), Alzheimer’s disease (AD)Subject terms: Enzymes, Proteases, Proteins, Structural biology, Chemical biology, Drug discovery, Molecular medicine Introduction Matrix metalloproteinases (MMPs) are zinc-dependent endopeptidases that play a pivotal role in maintaining and remodeling tissue architecture by degrading various components of the extracellular matrix (ECM) and basement membrane. These enzymes are essential for physiological processes such as embryogenesis, wound healing, and angiogenesis1. In the central nervous system, MMPs are particularly important in regulating the integrity of the blood-brain barrier (BBB), as they can degrade tight junction proteins and basement membrane components, leading to increased BBB permeability. While tightly controlled MMP activity is necessary for normal brain function and repair, dysregulation can contribute to pathological conditions such as neuroinflammation2, multiple sclerosis (MS), stroke3, Alzheimer’s disease (AD)4, and brain tumors5. By altering ECM composition and BBB structure, MMPs influence cellular signaling, immune cell infiltration, and neuronal environment stability, underscoring their dual role in both CNS health and disease6–8. They participate in many physiological and pathological processes in the brain and at the BBB9,10. The BBB is composed of endothelial cells, tight junction proteins, claudin-1, claudin-5, occludin, and zonula occludins-1 (ZO-1), and a basement membrane11–15. In neurological disorders like AD, Parkinson’s disease (PD), MS, and traumatic brain injury (TBI), inflammatory signals like TNF-α and IL-1β trigger the overexpression of MMPs, particularly MMP-2 and MMP-9 which degrade essential components of the BBB, including tight junction proteins and basement membrane proteins like laminin, fibronectin16–18 (Fig. 1). This increases BBB permeability, allowing toxic molecules and immune cells to infiltrate the brain and intensify neuroinflammation, oxidative stress, and neuronal death, thus accelerating disease progression19. In Alzheimer’s disease (AD), while MMP-9 can degrade amyloid-β plaques, its excessive activity paradoxically worsens amyloid accumulation and neurotoxicity by further compromising BBB integrity20,21. Similarly, in MS, MMP-9 enhances leukocyte migration across the damaged BBB, leading to chronic inflammation, demyelination, and neuronal damage22. Targeting MMP activity with inhibitors has emerged as a potential therapeutic strategy to preserve BBB function and slow the progression of these diseases23. Molecular components of the endothelial cells in BBB. MMP activity leads to the breakdown of tight junction proteins and extracellular matrix components, contributing to increased BBB permeability. Changes in barrier function are indicated by increased cell permeability and structural damage, highlighting the role of MMPs in compromising BBB integrity. MMP activity is tightly regulated through the formation of complexes with tissue inhibitors of metalloproteinases (TIMPs), which are naturally produced proteins within the body that act as endogenous inhibitors, ensuring proper regulation of MMP activity in various physiological processes8,24. Changes in the physiological balance between MMPs and TIMPs have been related to the development of neurodegenerative diseases8,24–26. Therefore, TIMPs have great potential to inhibit MMPs and protect against BBB disruption related to MMPs23. TIMPs demonstrate a high sequence and structure homology level, with a broad range of binding and inhibition to the family of MMPs. To create effective treatments that target certain MMPs, it is critical to determine the essential TIMP patterns that inhibit MMPs. Our research group recently generated a minimal TIMP library in yeast by employing DNA shuffling of the human TIMP family to identify key MMP inhibitory regions. Screening several minimal TIMP variants against MMP-3 and MMP-9 using fluorescent-activated cell sorting (FACS) revealed that variants as short as 20 amino acids maintained or improved their binding affinity to both MMPs27. In this study, we used a cell based assay using rat brain microvascular endothelial cell (RBMEC), which has been widely used in previous studies23,28–30 to evaluate the effect of MMPs and engineered minimal TIMP variants on BBB integrity and transport across the BBB. Permeability and trans-endothelial electrical resistance (TEER) were measured over time following adding MMP-9 enzyme or in a complex with TIMP inhibitors and engineered minimal TIMP variants, as indicators of BBB disruption. Results Binding kinetics of TIMP-1 and its minimal variant The minimal TIMP variants were previously engineered to target MMP-3 and MMP-9 in a previous study27. Protein binding kinetics of the minimal TIMP variant (mTC1) compared to wild-type TIMP-1 toward binding to MMP-3 and MMP-9 catalytic domains (MMP-3 cd and MMP-9 cd) in solution were assessed by biolayer interferometry (BLI) using an Octet biosensor platform. Dose-response curves for mTC1–MMP-3 and mTC1–MMP-9 cd interactions demonstrated a concentration-dependent increase in binding signals for mTC1 across a range of protein concentrations (0.5–20 µM), indicating enhanced interaction strength (Fig. 2A and B). During the association phase, signals rapidly increased, reflecting efficient complex formation. The dissociation phase exhibited a gradual signal decline that was consistent across concentrations. Maximal binding responses were observed at 20 µM, while minimal activity was noted at 0.5 µM. Kinetic analysis yielded dissociation constants (KD) of approximately 164 nM for mTC1–MMP-3 and 3 µM for mTC1–MMP-9 cd, indicative of moderate affinity interactions. (Table 1). Parallel binding analyses were conducted for full-length TIMP-1 interactions with the same MMP targets. The results, depicted in Fig. 3A and B, similarly demonstrated a concentration-dependent binding pattern, indicative of specific and saturable interactions between TIMP-1 and the tested MMPs. Higher concentrations of TIMP-1 (ranging from 0.25 to 4 µM) exhibited faster association rates and greater binding responses, clearly reflected in the increasing sensorgram signals. During dissociation, signals consistently decreased across concentrations, implying comparable stability and off-rates for TIMP-1 complexes with both MMP-3 and MMP-9 cd. Notably, TIMP-1 demonstrated high affinity for both proteases, with calculated KD values of approximately 11 nM and 59 nM for interactions with MMP-3 and MMP-9, respectively (Table 1), consistent with previous reports31,32. Binding kinetics of mTC1 with MMP-3 and MMP-9. Biolayer interferometry (BLI) analysis of mTC1 peptide binding kinetics to MMP-3 cd and MMP-9 cd proteins was performed using the BLI Octet R8. Binding responses were recorded over time at mTC1 concentrations of 0, 0.5, 1, 2.5, 5, 10, and 20 µM (color-coded). (A) and (B) show the binding kinetics of mTC1 with MMP-3 cd and MMP-9 cd, respectively, including association and dissociation phases. The curves corresponding to different concentrations are color-coded, with higher concentrations showing faster association rates and greater binding responses. mTC1 exhibited a 2.4-fold faster association rate (kon) with MMP-9 cd compared to MMP-3 cd, indicating more rapid complex formation. Table 1Binding kinetics of TIMP-1 and the minimal variant mTC1 with MMP-3 and MMP-9 measured by biolayer interferometry (BLI). The equilibrium dissociation constant (KD), association rate constant (ka), and dissociation rate constant (kdis) were calculated using a global 1:1 binding model. KD values are reported in nM or µM as indicated. Binding kinetics of TIMP-1 and the minimal variant mTC1 with MMP-3 and MMP-9 measured by biolayer interferometry (BLI). The equilibrium dissociation constant (KD), association rate constant (ka), and dissociation rate constant (kdis) were calculated using a global 1:1 binding model. KD values are reported in nM or µM as indicated. Binding kinetics of TIMP-1 with MMP-3 and MMP-9. Biolayer interferometry (BLI) analysis of TIMP-1 binding to MMP-3 cd and MMP-9 cd was performed using the Octet biosensor instrument. The binding responses were recorded as a function of time at different concentrations of TIMP-1. (A) shows the binding kinetics of TIMP-1 with MMP-3 cd, demonstrating a concentration-dependent increase in binding signal during the association phase, followed by a dissociation phase, indicating a stable protein complex. (B) shows the binding kinetics of TIMP-1 with MMP-9 cd, exhibiting similar concentration-dependent binding profiles. The curves corresponding to different concentrations are color-coded, with higher concentrations showing faster association rates and greater binding responses. TIMP-1 is associated with MMP-9 cd approximately 2.3-fold faster than with MMP-3 cd, as reflected by the higher ka. When comparing minimal TIMP variant mTC1 to wild-type TIMP-1, mTC1 demonstrated good binding to both MMP-3 cd and MMP-9 cd, which was particularly comparable to TIMP-1 in the case of MMP-3 cd. Kinetic parameters revealed that mTC1 exhibited higher association rate constants (ka), indicative of faster complex formation, as well as higher dissociation rate constants (kdis), relative to TIMP-1, which is consistent with the smaller size and less interactive surface area of the peptide compared to the full-length human TIMP-1 protein with larger interactive surface and multiple interacting loops. This kinetic profile contributes to the moderate overall binding affinity observed for mTC1. Effect of MMP-9, TIMPs, and TIMP variants on TEER of RBMEC Transendothlial electrical resistance (TEER) is a widely used quantitative measure of BBB integrity and maturation. In RBMEC cultures, attainment of monolayer confluency resulted in TEER values exceeding 100 Ω·cm², a threshold commonly associated with functional BBB formation and the establishment of intact endothelial tight junctions. The effects of the addition of MMP-9 cd protein alone and in complex with wild-type or engineered TIMP variants on the integrity of tight junctions in RBMEC monolayers forming the BBB were evaluated using TEER measurements. Treatment with 0.5 µM MMP-9 cd protein resulted in a moderate decrease in TEER over 2 h. Co-treatment with wild-type TIMP protein or minimal TIMP peptides at the same concentration partially restored TEER values compared to MMP-9 cd alone, showing a protective effect (Fig. 4). In contrast, treatment with 1.5 µM MMP-9 cd (Fig. 5A) led to a stronger reduction in TEER, indicating more disruption of the barrier. However, co-treatment with all TIMPs and minimal TIMPs still improved TEER compared to MMP-9 cd alone, with mTC1 and mTC3 showing similar recovery to TIMP-1 and TIMP-3. Measuring TEER values of MMP-9 cd and its protein complex with TIMPs and minimal TIMPs on RBMEC. The TEER value of RBMEC after treating with 0.5 µM MMP-9 cd alone or in complex with TIMP-3, TIMP-1, mTC1, mTC3 was compared at 60-min intervals for 2 h. Statistical analyses were performed using a one-way ANOVA test (p-values: *< 0.1). Measurement of TEER values following treatment with MMP-9 cd and its complexes with TIMPs and minimal TIMP variants in RBMEC monolayers. (A) TEER measurements of RBMECs treated with 1.5 µM MMP-9 cd alone or in complex with TIMP-1, TIMP-3, mTC1, and mTC3, along with FITC-dextran, recorded at 60-minute intervals over 2 h. (B) TEER measurements of RBMECs treated with 0.5 µM and 1.5 µM MMP-9 cd alone or in complex with 0.5 µM and 1.5 µM TIMP-3 or mTC1, along with Alexa Fluor™ 488 cadaverine, showing concentration-dependent protective effects of TIMP-3 and mTC1 in maintaining barrier integrity. Statistical analysis was performed using one-way ANOVA (p-values: *< 0.1). To further validate these findings, TEER measurements were performed with the Alexa Fluor™ 488 cadaverine permeability assay. Consistent with the FITC-dextran results, treatment with MMP-9 cd enzyme resulted in a dose-dependent reduction in TEER value, confirming degradation of RBMEC tight junctions. Co-treatment with MMP-9 cd in complex with TIMP-3 protein or mTC1 variant effectively restored TEER values toward baseline. This demonstrated a protective effect of TIMP and its variant on endothelial integrity via inhibition of MMP-9 cd (Fig. 5B). Together, these results reinforce previous evidence that increased TEER correlates with strengthened tight junctions and decreased paracellular permeability, highlighting the importance of targeting MMP-9 activity using either natural or engineered TIMP variants to preserve RBMEC tight junction to maintain BBB homeostasis23. Effect of MMPs, TIMPs, and minimal TIMP variants on RBMEC permeability The effects of MMP-9 catalytic domain (MMP-9 cd), wild-type human TIMP protein, and engineered minimal TIMP variants on endothelial tight junction integrity were assessed using permeability assays. Rat brain microvascular endothelial cells (RBMECs) were cultured on 24-well inserts until they reached approximately 90% confluency, as indicated by stable TEER measurements averaging 150 Ω × cm². To evaluate the influence on endothelial barrier function, the cells were treated with MMP-9 cd alone or in complex with either natural TIMPs (TIMP-1, TIMP-3) or engineered minimal TIMP variants (mTC1, mTC3). Treatment with MMP-9 cd alone caused a dose-dependent increase in permeability, aligning with previous findings that link MMP-9 activity to BBB disruption33,36,35. Treatment with 0.5 µM MMP-9 cd led to a six-fold increase in permeability relative to baseline conditions. Specifically, 0.5 µM MMP-9 cd induced a six-fold increase in permeability compared to baseline (Fig. 6). Furthermore, to mitigate this barrier disruption, both natural and engineered inhibitors were tested for their protective effects. At a concentration of 0.5 µM, TIMP-1 and TIMP-3, when complexed with MMP-9 cd, reduced the increase in permeability to approximately 2.5-fold and 3-fold, respectively, compared to vehicle-treated controls (Fig. 6). At higher concentrations of the enzyme, MMP-9 cd (1.5 µM), the wild-type TIMP proteins still showed some level of elevated permeability, 3-fold with TIMP-1 and 4-fold with TIMP-3. However, the minimal TIMP variants mTC1 and mTC3 demonstrated more consistent efficacy. When used at 1.5 µM in complex with MMP-9 cd, both variants reduced permeability significantly, achieving a 3-fold decrease relative to MMP-9 cd treatment alone and restoring barrier integrity closer to baseline levels (Fig. 7A). Effect of MMP-9 and its complex with TIMPS and minimal TIMPs on the cellular tight junctions. MMP-9 cd disrupts the tight junctions of RBMEC in a dose-dependent manner based on an average of 2 different wells from 6 different groups of RBMEC and in the presence of 0.5 µM of both TIMP-3, TIMP-1, mTC1 and mTC3 and MMP-9 cd, permeability was decreased over one and half of baseline. Ordinary one-way ANOVA was used to determine statistical significance (p-value: *< 0.1, **< 0.01). (Experiments were performed in 24-well Transwells;Vc, t = 600 µl × Sb, t/Sa). Effect of MMP-9 cd and its complexes with TIMPs and minimal TIMP variants on BBB permeability. (A) Permeability of RBMEC monolayers to 10 kDa FITC-dextran following treatment with MMP-9 cd alone or in complex with TIMP-1, TIMP-3, mTC1, and mTC3. MMP-9 cd disrupted endothelial tight junctions in a dose-dependent manner, whereas co-treatment with TIMPs and minimal TIMPs markedly reduced permeability, restoring values to near-baseline levels. (B) Permeability of RBMEC monolayers to Alexa Fluor™ 488 cadaverine, showing a similar dose-dependent increase in permeability with MMP-9 cd and significant protection in the presence of TIMP-3 and mTC1. Ordinary one-way ANOVA was used to determine statistical significance (p-value: *< 0.1, **< 0.01 ). Experiment was performed in 24-well Transwells;Vc, t = 600 µl × Sb, t/Sa). To further assess barrier permeability to small molecules, Alexa Fluor™ 488- conjugated cadaverine was used as a alternative tracer in parallel experiments. Consistent with FITC-dextran results, MMP-9 cd treatment increased cadaverine permeability in a dose-dependent manner, resulting in approximately 1.5-fold and 3-fold increases at 0.5 µM and 1.5 µM enzyme concentrations, respectively. Co-treatment with TIMP-3 or mTC1 effectively reduced this increase, restoring permeability toward baseline levels (Fig. 7B). These complementary results confirm that both natural and engineered TIMP inhibitors help preserve endothelial barrier integrity against MMP-9–induced disruption and highlights the potential of minimal TIMP variants as promising candidates for therapeutic interventions targeting MMP-mediated BBB disruption. Effect of MMP-9, TIMPs, and TIMP variants on tight junctions in RBMEC To investigate the effects of MMP-9 cd and the engineered minimal TIMP variant (mTC1) on the RBMEC tight junctions (TJs), the cells were cultured on fibronectin-coated plates treated with either 1.5 µM MMP-9 cd protein alone or 1.5 µM MMP-9 cd protein pre-incubated with an equimolar amount of mTC1 peptide to form an enzyme/inhibitor complex before addition to the cells. Following treatment with MMP-9 cd and MMP-9 cd/mTC1, immunofluorescent staining for the tight junction proteins ZO-1 and occludin was performed, and samples were visualized using fluorescence microscopy. Control samples, RBMECs without treatment, demonstrated normal localization of ZO-1 (Fig. 8A) and occludin (Fig. 8B) at cell junctions, indicative of intact tight junctions. In contrast, cells treated with 1.5 µM MMP-9 cd enzyme exhibited disrupted localization patterns of both ZO-1 and occludin, signaling compromised tight junction integrity. Notably, co-treatment with mTC1 effectively mitigated the MMP-9-induced disruption, stabilizing ZO-1 and occludin distribution and preserving tight junction integrity (Fig. 8A and B). These results confirm the protective role of minimal TIMP variant, mTC1, against MMP-9-induced tight junction disruption in RBMECs. Effect of MMP-9 alone and with mTC1 on tight junction. The effect of MMP-9 and mTC1 on tight junction of RBMEC cell line was confirmed by immunocytochemistry (ICC). ICC images of RBMEC were made by staining with primary (A) ZO-1 antibody, or (B) occludin antibody followed by secondary antibody anti-rabbit Alexa Fluor 488 and taken by ZOE fluorescent microscope at 40x magnification. The scale bar is 40 µm. Cellular Uptake, Delivery, and transcytosis of TIMP variant and CPP-Conjugated TIMP variant To evaluate the recycling and transcytosis efficiencies of CPP, mTC1 peptides, and their combination (CPP and mTC1), a fluorescence-based assay using FAM-labeled peptides was conducted on confluent RBMEC cells that formed tight junctions (Fig. 9). The recycling efficiency (as determined by fluorescence intensity in the apical chamber) showed clear differences among the groups after 4 h. mTC1 alone displayed moderate recycling efficiency (~ 30%), whereas CPP alone exhibited a significantly lower recycling rate (~ 10%). Remarkably, the CPP-mTC1 combination resulted in a substantial increase in recycling efficiency, exceeding 60%, indicative of a synergistic effect. Similarly, transcytosis efficiency, measured by fluorescence intensity in the basolateral chamber after 4 h, followed a comparable trend. Both mTC1 and CPP individually presented low transcytosis levels, each below 20%. However, their combination markedly enhanced transcytosis, reaching approximately 70%. This highlights the positive role of CPP in facilitating the transcellular transport of mTC1 across endothelial barriers. Collectively, these findings suggest that conjugation of CPP significantly augments mTC1 transport efficiency across the blood-brain barrier model, underscoring the potential of CPP as an effective delivery facilitator for therapeutic molecules. Minimal TIMP penetration through BBB. After RBMEC cells formed BBB, Transport of FAM-labeled peptides was evaluated by fluorescence measurement after incubation and compare to initial fluorescent ratio. FAM-labeled mTC1, FAM-labeled mTC1-CPP conjugated, FAM-labeled CPP were added to apical chamber and samples were taken from basolateral chamber for 30 min interval and measured using the Biotek Synergy 2 microplate reader (excitation: 485/20 nm, emission: 528/20 nm, gain: 84). Discussion This study investigated the efficacy of minimal engineered TIMP variants modulating MMP-9 activity and preserving BBB integrity. Using a cell-based model of RBMECs, we evaluated the effects of the catalytic domain of MMP-9 (MMP-9 cd) on BBB permeability and tight junction structure and assessed the protective capabilities of both natural TIMPs (TIMP-1, TIMP-3) and minimal TIMP variants (mTC1, mTC3). Consistent with prior research, MMP-9 cd treatment induced a dose-dependent increase in BBB permeability, highlighting its potent role in barrier disruption23. A concentration of 0.5 µM MMP-9 cd led to a six-fold increase in permeability compared to baseline, underscoring the enzyme’s capacity to compromise endothelial tight junctions. Co-treatment with TIMP-1, TIMP-3, or engineered minimal TIMPs significantly attenuated this effect. Specifically, TIMP-1 and TIMP-3 at 0.5 µM reduced permeability to approximately 2.5- and 3-fold, respectively, while higher concentrations (1.5 µM) resulted in even greater protection. Notably, the minimal TIMP variants mTC1 and mTC3 demonstrated comparable efficacy, restoring permeability to levels similar to those achieved by full-length TIMPs, indicating their strong potential as simplified yet effective MMP-9 inhibitors. To further characterize barrier integrity, Alexa Fluor™ 488-cadaverine, a low–molecular-weight tracer, was used to show permeability of the lower molecular weight. Consistent with macromolecular permeability results, MMP-9 cd significantly increased cadaverine permeability in a concentration-dependent manner, confirming enhanced paracellular diffusion through the disrupted barrier. Co-treatment with TIMP-3 and mTC1 effectively reduced cadaverine flux and restored permeability values toward baseline, demonstrating that both inhibitors maintained the barrier against leakage of small molecules. Complementary TEER measurements further supported these findings. MMP-9 cd treatment led to a pronounced reduction in TEER, indicative of disrupted tight junction integrity. Conversely, co-treatment with TIMP-1, TIMP-3, or mTC1 restored TEER values, reflecting preservation of the endothelial barrier. Immunofluorescence staining for the tight junction proteins ZO-1 and occludin revealed marked disorganization and delocalization following MMP-9 exposure, consistent with structural breakdown of cell-cell junctions. However, co-application of mTC1 preserved the localization and continuity of ZO-1 and occludin at the cell borders, further confirming the protective role of minimal TIMPs in maintaining tight junction integrity. Collectively, these results demonstrate that engineered minimal TIMPs, particularly mTC1, not only match the inhibitory effectiveness of full-length TIMPs but also offer a streamlined and potentially more versatile alternative for therapeutic development. Beyond their inhibitory activity, the transport properties of mTC1 were assessed across the RBMECs, particularly in combination with a cell-penetrating peptide (CPP). Using a fluorescence-based assay, it was shown that mTC1 alone exhibited moderate recycling and low transcytosis efficiency, while CPP alone performed poorly in both metrics. Remarkably, the combination of CPP and mTC1 significantly enhanced both recycling (to over 60%) and transcytosis (to approximately 70%) across RBMEC monolayers. These results highlight a synergistic interaction, in which the CPP moiety substantially improves the intracellular trafficking and transcellular transport of mTC1. This suggests that mTC1 variants have a high BBB transport, and CPP conjugation may be an effective strategy to facilitate the delivery of minimal TIMPs or other therapeutic agents across the BBB. In conclusion, our findings show the pivotal role of MMP-9 in degradation of tight junctions in endothelial cells which can result in BBB disruption. Further, it was demonstrated that both wild-type human TIMP protein and engineered minimal TIMP variants offer promising strategies to counteract MMP-9-induced permeability changes. The ability of minimal TIMPs, such as mTC1, to preserve tight junction integrity and their compatibility with CPP-mediated delivery systems positions them as attractive candidates for therapeutic development. The enhanced transcytosis observed with the CPP–mTC1 combination further supports their potential in targeting neurological disorders where BBB disruption is a key pathological feature. Future research should focus on in vivo validation, optimization of delivery systems, and deeper investigation into the molecular mechanisms underpinning TIMP transport and activity within the central nervous system. Methods Protein expression, purification The catalytic domain of MMP-9 (MMP-9cd) and MMP-3 (MMP-3cd) proteins were expressed as recombinant proteins with an N-terminal Hisx6 tag in Rosetta™ (DE3) pLysS cells (Sigma Aldrich, MO, USA) using a pET28a vector, as previously reported23,36,37. Briefly, 0.5 mM IPTG was used to induce MMP-9cd protein expression for 4 h at 37 °C. After cell lysis using sonication, the insoluble MMP-9cd and MMP-3cd proteins was solubilized with urea in a denaturing condition. Bacterial pellets were combined with lysis buffer and incubated at 4 °C overnight. Then, the lysate was incubated for 1 h at room temperature with DNase I and sodium deoxycholate. The lysate was separated by centrifugation, and aliquots were collected for SDS-PAGE analysis. Sonication, solubilization, lysis, and centrifugation were repeated until there was no more insoluble material in the pellet fraction. Immobilized metal affinity chromatography (IMAC) was used to purify the pooled supernatants containing solubilized MMP-9 cd using Ni resin. Finally, gradient dialysis was used to refold the MMP-9 cd. The recombinant human TIMP-1 and TIMP-3 proteins were expressed using the mammalian protein expression vector pTT32. 1–2 × 10⁵ viable cells/mL of Human embryonic kidney 293-FreeStyle (HEK293-Freestyle) (Thermo Fisher Scientific, USA) were seeded into 250 ml flasks containing pre-warmed FreeStyle™ 293 Expression Medium. Flask was incubated in a 37 °C humidified atmosphere with 8% CO2 on an orbital shaker platform rotating at 125 rpm. To maintain or expand the cultures, subculture was performed as necessary. For transfection, cells were seeded at 0.5 × 10⁶ cells/mL in 1 L roller bottles with 300 mL total volume and incubated under the same conditions. After reaching a density of 1.0 × 10⁶ cells/mL and over 90% viability, 300 µg of filter-sterilized DNA was added to 30 mL sterile Dulbecco’s phosphate-buffered saline (DPBS) and vortexed for 3 s. Then, 0.9 mL of 1 mg/mL PEI solution was added, and the mixture was incubated at room temperature for 20 min before adding it to the cells. Following transfection, the cells were incubated for 48 h, and intracellular proteins were harvested by centrifugation at 3,000 x g for 5 min. For wild-type human TIMP-1 purification using chromatography, a linear gradient of buffer B (50 mM HEPES, pH 6.8, and 0.5 M NaCl) was used to purify the clarified medium over an SP-Sepharose column after it had been concentrated and dialyzed into buffer A (50 mM HEPES, pH 6.8, and 25 mM NaCl)32. The immobilized metal affinity chromatography (IMAC) with a gravity flow column based on HisPur™ Ni-NTA Resin (Thermo Fisher Scientific, USA) was used purify Hisx6-tagged wild-type TIMP-3. The engineered minimal TIMP variant peptides were purchased from Genscript with over 97% purity as a powder and prepared in working stocks as previously described27. Measuring binding kinetics using BLI-Octet R8 To investigate the binding kinetics and affinities between matrix metalloproteinases (MMP-3cd and MMP-9cd) and TIMP-1 as well as engineered minimal TIMP variants (mTC-1), a biolayer interferometry (BLI) assay was performed using an BLI Octet R8 biosensor system (Sartorius). Hisx6-tagged recombinant MMP proteins were immobilized onto HIS1K biosensor tips following the manufacturer’s protocol. After immobilization, the sensors were blocked with phosphate-buffered saline (PBS) to minimize nonspecific interactions. Binding assays were then conducted by exposing the sensors to serial dilutions of TIMP-1 (4, 2, 1, 0.5, 0.25, 0.125, 0 µM) and minimal TIMP variants (20, 10, 5, 2.5, 1, 0.5, 0 µM) in a kinetic buffer composed of Tris-NaCl-CaCl2 (TNC) at 25 °C with continuous shaking at 1000 rpm. Real-time binding curves were recorded during both the association and dissociation phases. Kinetic parameters, including the association rate constant (ka), dissociation rate constant (kdis), and the equilibrium dissociation constant (KD), were extracted by fitting the data into a 1:1 Langmuir binding model using Octet analysis software. RBMEC culture The RBMEC cells were purchased from ScienCell Research Laboratories (ScienCell, USA). For RBMEC culture, a T-75 flask was coated with fibronectin (2 µg/cm2), prepared by adding 150 µl of 1 mg/ml fibronectin (Sigma-Aldrich, USA) to 10 ml of sterile Dulbecco’s phosphate-buffered saline (DPBS) (ThermoFisher Scientific, USA), and incubated overnight. Complete medium according to manufacturer’s recommendation (ScienCell, USA) was prepared aseptically by adding 500 ml Endothelial Cell Medium-rat (ECM-r), 5 ml 100X penicillin/streptomycin, 25 ml of calf serum, and 5 ml Endothelial Cell Growth Supplement-rat (ECGS-r). Then, cryopreserved cells were thawed in a 37 °C water bath and resuspended and transferred to the fibronectin-coated flask. The culture vessel was gently rocked for even cell distribution and incubated at 37 °C in a humidified atmosphere with 5% CO2, undisturbed for 16 h, followed by medium refreshment to remove residual DMSO and unattached cells. Once the RBMEC cells reached 90% confluency, they were passaged with brief treatment with 0.25% Trypsin/EDTA (ThermoFisher Scientific, USA), according to the manufacturer’s recommendation (ScienCell, USA). Establishment of an in vitro cell-based model based on RBMEC RBMEC cells were used to establish an in vitro cell-based model to study effect of MMP degradation of tight junction similar to BBB as described previously28–30. Once RBMEC cells were confluent, 50,000 cells/cm² of RBMECs were seeded onto the apical side of 0.4 μm pore size 24-well Transwell® polyester inserts with 0.33 cm2surface area (Corning, NY, USA). Before cell seeding, the inserts were coated with a mixture of 400 µl type IV collagen (1 mg/mL), 100 µl fibronectin (1 mg/mL) (Sigma-Aldrich, CA, USA) in 500 µl DPBS (ThermoFisher Scientific, Waltham, MA, USA) for at least 4 h at 37 °C and 5% CO2. Then, inserts containing the 100 µL complete media and cells were placed into the corresponding wells filled with 600 µL complete media and incubated at 37 °C and 5% CO₂. Measuring permeability of RBMEC cells The in vitro RBMEC model was used to evaluate BBB permeability. Once the RBMEC cells seeded into the inserts reached confluency, the media was changed to serum-free endothelial cell media (sfECM) supplemented with 0.55 nM hydrocortisone and incubated overnight. The media in both chambers was refreshed at least 1 h prior to the addition of proteins. MMP-9 cd, TIMP-1, TIMP-3, and engineered minimal TIMPs variants (mTC1, mTC3) were mixed and incubated at 37 °C for 1 h. After incubation, serum-free media containing 10 kDa fluorescein isothiocyanate (FITC) conjugated dextran (1 µM final; Sigma, USA) or Alexa Flour TM 488 cadaverine (1 µM final concentration; Thermo Fisher, USA) were added to MMP-9 cd alone or with TIMPs and minimal TIMPs mixtures. After adding protein/tracer mixtures, 60 µL samples were taken from the basolateral chamber (lower chamber), at 30-minute intervals. Samples were taken for 120 min, and then fluorescence intensity was measured using a Synergy microplate reader with excitation at 485/20 nm, emission at 528/20 nm, and a gain value was set at 84. To calculate Permeability (Pe), the first background fluorescence from tracer-free sfECM medium was subtracted from each sample. Then, the signal was corrected for sampling replacement for each timepoint after the first [Eq. 1. – Eq. 2.]. *Because no sampling occurred prior to the 15 min timepoint. Which Mb, t is the measured signal and Sb, t is the corrected signal. By dividing the corrected reading by the apical chamber reading (Sa) and multiplying by the basolateral chamber media volume (600 µL), clearance volume (Vc, t) was determined (Eq. 5.). To calculate the slope (m) of clearance volume versus time, Linear regression in GraphPad Prism was used. Permeability (Pe) was determined by subtracting the reciprocal of the blank slope (no cells) from the reciprocal of each sample slope (Eq. 6.). Which mc is the slope for a Transwell containing cells and let mb is the slope for a blank Transwell. The slope of clearance volume versus time should have units µl/min; to convert to typical units, use the area of the Transwell (0.33 cm2) Eq. 7.: Pe’s % error was calculated by dividing the standard error of the slope by the slope, then multiplying the % error by Pe. Statistical significance (p-value) was calculated using ordinary one-way ANOVA in GraphPad Prism. Transendothelial electrical resistance (TEER) measurements of RBMEC model TEER measurement was performed to confirm the integrity of the tight junction (TJ). One day after switching the media in the Transwell® chambers to serum-free media, TEER was measured using an EVOM3 resistance meter (World Precision Instruments, USA). The TEER measurement was done by placing one electrode inside the apical chamber and the other one outside in the well, following the manufacturer’s protocol. Subsequently, the media in both chambers were replaced with pre-warmed fresh media, and the chambers were returned to the incubator for 60 min. TEER was rechecked to ensure it had recovered to more than 150 Ω × cm². TEER was measured at 30-minute intervals for 2 h endothelial monolayers were then treated with MMPs and TIMPs. Resistance measurements from collagen/fibronectin-coated Transwells without cells (blank) were subtracted from those with cells to determine the final TEER value, calculated using the equation: TEER = (sample TEER value – background TEER value) × area of the insert membrane. Results are reported in Ω × cm² and reflect the change in TEER relative to the control. Immunocytochemistry of RBMEC Immunofluorescent staining of RBMEC was used to investigate the presence of TJs using an antibody against ZO-1 and occludin. RBMEC cells were grown on 8-well chamber slides coated with fibronectin (2 µg/cm2). After they reached 90% confluency, the complete media was changed to serum-free media and incubated overnight at 37 °C incubator containing 5% CO2 incubator. Cells treated with either MMP-9 protein or with minimal TIMPs in complex with MMP-9 were incubated in a 37 °C incubator with 5% CO2 for 1 h. After the incubation period, cells were washed with DPBS two times and fixed with 4% paraformaldehyde in PBS (pH 7.4) for 15 min. After fixation, the cells were permeabilized using 0.25% Triton-X100 in PBS followed by blocking with PBS containing 5% BSA in for 1 h at room temperature. For immunocytochemistry staining, the cells were incubated with primary polyclonal rabbit ZO-1 antibody (Thermo Fisher Scientific, USA, Cat. #40–2200, dilution 1:50), or polyclonal rabbit antibody against occludin (ThermoFisher Scientifics, USA, Cat. #33–1500, dilution 1:50) in DPBS containing 1% BSA overnight followed by PBS washes, and incubation with secondary antibody, Alexa Fluor 488-conjugated anti-rabbit antibody (Invitrogen, Life Technologies, Cat. #A-11008, dilution 1:1000 in 1% BSA in DPBS) for one hour in room temperature. Subsequently, the cell nucleus was stained with Hoechst stain (Invitrogen, Life Technologies) at a 1:2,000 dilution in DPBS. Cells were imaged using a fluorescence microscope (Biorad, USA), followed by washing in DPBS two times. Cellular uptake, delivery, and endocytosis of TIMP variant and CPP-conjugated TIMP variant To evaluate the recycling and transcytosis efficiencies of CPP (RKKRRQRRRdGR), mTC1, and their combination (CPP + mTC1), a fluorescence-based assay using FAM-labeled molecules was conducted on RBMEC cells. RBMEC cells were cultured on polycarbonate porous cell culture inserts (PCIs) in Transwell plates and maintained in 500 ml ECM-r, 5 ml 100X penicillin/streptomycin, 25 ml of calf serum, and 5 ml ECGS-r at 37 °C with 5% CO₂ until a confluent monolayer was formed. The transcytosis assay involved a pulse-chase protocol38 where the apical chamber was incubated with FAM-labeled mTC1, CPP, or mTC1 + CPP at a final concentration of 1.5 µM for 1 h at 37 °C (pulse phase), followed by washing with PBS to remove unbound molecules (wash phase). After each sample taking step and removing the media and washing, both apical and basolateral chambers were filled with fresh serum-free medium, and the plates were incubated for an additional 4 h at 37 °C with 5% CO₂. After this time, 60 µl samples from apical and basolateral chambers were taken, and their fluorescence intensity was measured using a fluorescence plate reader (Excitation: 488 nm, Emission: 520 nm) to calculate recycling and transcytosis of peptides, respectively. Transcytosis efficiency was calculated as the percentage of fluorescent signal detected in the basolateral chamber relative to the total fluorescence initially applied to the apical chamber during the pulse phase, representing the fraction of peptides successfully transported across the cell monolayer. Recycling efficiency was calculated based on the fluorescence recovered from the apical chamber after the chase period, indicating the proportion of internalized peptides that were recycled back to the apical surface instead of being degraded or going through transcytosis. These calculations allowed quantitative comparison of the intracellular trafficking behavior of each construct. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements We would like to thank the Cellular and Imaging Core at the University of Nevada, Reno, working under NIH P30 GM145646 grant. Author contributions M.R.-S.; conception, E.T.; performed the experiments, generated the figures, E.T., M.R.-S., data analysis, methods, E.T., writing the original draft, M.R.-S., writing, and editing the manuscript drafts. M.R.-S. funding and supervision, all authors reviewed the manuscript. Funding M. R.-S. had funding support from NIH 1R03AG070511-01A1, R35GM160474 and NIH 1S10OD032385-01A1. Data availability All data generated or analyzed during this study are included in this published article and its supplementary information files. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement All data generated or analyzed during this study are included in this published article and its supplementary information files.