Claudin-5 binder enhances focused ultrasound-mediated opening in an in vitro blood-brain barrier model

Abstract

Rationale: The blood-brain barrier (BBB) while functioning as a gatekeeper of the brain, impedes cerebral drug delivery. An emerging technology to overcome this limitation is focused ultrasound (FUS). When FUS interacts with intravenously injected microbubbles (FUS^+MB), the BBB opens, transiently allowing the access of therapeutic agents into the brain. However, the ultrasound parameters need to be tightly tuned: when the acoustic pressure is too low there is no opening, and when it is too high, tissue damage can occur. We therefore asked whether barrier permeability can be increased by combining FUS^+MB with a second modality such that in a clinical setting lower acoustic pressures could be used. Methods: Given that FUS^+MB achieves BBB opening in part by disruption of tight junction (TJ) proteins such as claudin-5 of brain endothelial cells, we generated a stable MDCK (Madin-Darby Canine Kidney) II cell line (eGFP-hCldn5-MDCK II) that expresses fluorescently tagged human claudin-5. Two claudin-5 binders, the peptide mC5C2 and cCPEm (truncated form of an enterotoxin), reported previously to weaken the barrier, were synthesized and assessed for their abilities to enhance the permeability of cellular monolayers. We then performed a comparative analysis of single and combination treatments, measuring transendothelial electrical resistance (TEER) and cargo leakage, combined with confocal image analysis. Results: We successfully generated a novel cell line that formed functional monolayers as validated by an increased TEER reading and a low (< 0.2%) permeability to sodium fluorescein (376 Da). We found that the binders exerted a time- and concentration-dependent effect on barrier opening when incubated over an extended period, whereas FUS^+MB caused a rapid opening followed by recovery after 12 hours within the tested pressure range. Importantly, preincubation with cCPEm prior to FUS^+MB treatment resulted in greater barrier opening compared to either FUS^+MB or cCPEm alone as measured by reduced TEER values and an increased permeability to fluorescently labelled 40 kDa dextran (FD40). Conclusion: The data suggest that pre incubation with clinically suitable binders to TJ proteins may be a general strategy to facilitate safer and more effective ultrasound-mediated BBB opening in cellular and animal systems and potentially also for the treatment of human diseases of the brain.

Keywords: Blood-brain barrier (BBB); claudin-5; focused ultrasound; transendothelial electrical resistance (TEER).

Figure 1

eGFP-hCldn5-MDCK II cells exhibit a tight monolayer and human claudin-5 is localized to cell/cell contacts. (A) Scheme of tight junctions formed by proteins such as claudin-5, occludin, and ZO-1. Domain structure of claudin-5. (B) Representative phase contrast image of confluent parental MDCK II cells and epifluorescence images of confluent MDCK II cells expressing eGFP only (eGFP-MDCK II) or eGFP-tagged human claudin-5 (eGFP-hCldn5-MDCK II). (C) Epifluorescence images of isolated clusters of eGFP-hCldn5-MDCK II cells. White arrows indicate the absence of localization of eGFP fluorescence in areas without cell/cell contacts. (D) Time-lapse fluorescence images of eGFP-hCldn5-MDCK II cells from the time of seeding at a density of 200,000 cells/cm² (0 h) to complete formation of a confluent monolayer (30 h). (E) Expression of claudin-5, (F) occludin and (G) ZO-1, localized to cell/cell borders. Nuclei were stained with DAPI. Scale bars: 50 µm (B-C), 250 µm (D) and 20 µm (E-G).

Figure 2

eGFP-hCldn5-MDCK II cells exhibit a tight monolayer as determined by TEER and cargo leakage. (A) Western blotting with either a claudin-5 or GFP antibody reveals expression of the eGFP-hClaudin5 fusion protein in eGFP-hCldn5-MDCK II but not eGFP-MDCK II cells. Expression of the tight junction proteins ZO-1 and occludin is also shown. (B) Scheme of Transwell insert to measure TEER and permeability. (C) eGFP-hCldn5-MDCK II cells display a four-fold higher TEER than MDCK II cells. iBEC cells are included for comparison. (D) All three MDCK II cell lines show a < 0.2% permeability for sodium fluorescein (NaFl), indicating a tight BBB. (E) TEER of eGFP-hCldn5-MDCK II and MDCK II cells shown as a function of cell density and days in culture. Asterisks refer to TEER differences between eGFP-hCldn5-MDCK II and MDCK II for a given time point. TEER values are shown as Ω·cm² and results are expressed as mean ± SEM. N>10 per condition. (F) Schematic illustration of the experimental workflow. Statistical significance was determined as unpaired Student's t-test (*p<0.05, **p<0.01, ***p<0.001 and ****P<0.0001).

Figure 3

Incubation with mC5C2 and GST-cCPEm reveals differences in the reduction of the absolute TEER in eGFP-hCldn5-MDCK II cells. (A-B) Schematic diagram showing mC5C2 binding of the extracellular loop 1 (ECL1) and cCPEm binding ECL2 of claudin-5. The homology model of claudin-5 was created in Swiss-Model using human claudin-9 (PDB ID 6OV2) as template. The cCPEm structure was extracted from the same PDB entry (6OV2). mC5C2 was placed in proximity to the model of claudin-5 whereas cCPEm was docked to claudin-5 for schematic purposes only. The molecular structures were generated using Maestro (Schrödinger Release 2020-4, New York, 2020). (C-D) Incubation of eGFP-hCldn5-MDCK II with mC5C2 and GST-cCPEm causes concentration- and incubation-time-dependent reductions in TEER. N=6 of each condition. Two-way ANOVA with Sidak's multiple comparison test (*p<0.05, and ***p<0.001). Asterisks refer to the difference in TEER value at the measured time points, compared to the TEER value of untreated control.

Figure 4

Focused ultrasound with microbubbles (FUS^+MB) leads to a rapid opening of the barrier followed by closure within 12 hours. (A) Schematic diagram of how ultrasound is delivered to the cells. (B) TEER measurement as a function of acoustic pressure (in MPa) before and immediately after FUS^+MB treatment. (C) Absolute TEER measurement as a function of incubation time shown for the 0.4 MPa condition. N=3-6 for each condition. Two-way ANOVA with Sidak's multiple comparisons tests (**p<0.01 and ****P<0.0001).

Figure 5

Preincubation with GST-cCPEm lowers the acoustic pressure required for focused ultrasound-mediated barrier opening. (A) Experimental scheme. (B) Absolute TEER measurement and (C) FD40 leakage of eGFP-hCldn5-MDCK II cells preincubated with 100 nM GST-cCPEm for 2 h, followed by FUS^+MB at the indicated pressures. (D) TEER measurement to assess barrier closure when GST-cCPEm was removed followed treatment. (E) Absolute TEER measurement and (F) assessment of FD40 leakage of eGFP-hCldn5-MDCK II cells preincubated with 100 nM GST-cCPEm for 12 h, followed by FUS^+MB at the indicated pressures. N≥5 from at least two independent experiments. Two-way ANOVA with Tukey's and Sidak's multiple comparisons tests (*p<0.05, **p<0.01, ***p<0.001 and ****P<0.0001).

Figure 6

Effects of GST-cCPEm and FUS^+MB alone and in combination on the junctional morphology of eGFP-hCldn5-MDCK II cells. Immunostaining for claudin-5, occludin and ZO-1 for the indicated experimental conditions. The eGFP-hCldn5 and claudin-5 staining were performed in the same sample, whereas the occludin and ZO-1 staining were performed in separate samples. Staining of cell nuclei with DAPI. Scale bar =10 µm.

Figure 7

Quantification of the redistribution of the eGFP-hCldn5 signal from the plasma membrane to the cytosol. (A) The ratio of the nearest neighbour distance in the membrane (DNN^IN) versus the nearest neighbour distance in the cytosol (DNN^OUT) across conditions. The ratio was significantly higher for the 12 h GST-cCPEm and 12 h GST-cCPEm + FUS^+MB conditions, compared to the untreated control. One-way ANOVA with Tukey's multiple comparisons tests (*p<0.05 and **p<0.01). (B) The eGFP-hCldn5 intensity ratio in the cytosol compared to that of the membrane across conditions. One-way ANOVA with Tukey's multiple comparisons tests (*p<0.05, **p<0.01 and ****P<0.0001).

Figure 8

Summary scheme of combinatorial treatment. GST-cCPEm forms a wedge which destabilizes claudin-5 and thereby the TJs, making it easier for the ultrasound pressure wave to open the barrier, with potential implications for FUS-mediated drug delivery.