gH625-liposomes as tool for pituitary adenylate cyclase-activating polypeptide brain delivery

The blood-brain barrier (BBB) regulates the traffic of molecules into the central nervous system (CNS) and also limits the drug delivery. Due to their flexible properties, liposomes are an attractive tool to deliver drugs across the BBB. We previously characterized gH625, a peptide derived from Herpes simplex virus 1. The present study investigates the efficiency of liposomes functionalized on their surface with gH625 to promote the brain uptake of neuroprotective peptide PACAP (pituitary adenylate cyclase-activating polypeptide). Using a rat in vitro BBB model, we showed that the liposomes preparations were non-toxic for the endothelial cells, as assessed by analysis of tight junction protein ZO1 organization and barrier integrity. Next, we found that gH625 improves the transfer of liposomes across endothelial cell monolayers, resulting in both low cellular uptake and increased transport of PACAP. Finally, in vivo results demonstrated that gH625 ameliorates the efficiency of liposomes to deliver PACAP to the mouse brain after intravenous administration. gH625-liposomes improve both PACAP reaching and crossing the BBB, as showed by the higher number of brain cells labelled with PACAP. gH625-liposomes represent a promising strategy to deliver therapeutic agents to CNS and to provide an effective imaging and diagnostic tool for the brain.


Results
Characterization of liposomes. Liposomes loaded with PACAP-Rho and functionalized on their surface with gH625 ( Fig. 1) were characterized. The sequences of synthesized peptides are reported in Table 1. The hydrodynamic diameters (DH) and polydispersity index (PDI) of all liposomes were measured using dynamic light scattering (DLS). Three independent experiments were performed for each sample and each measurement was performed at least in triplicate. All liposome solutions present a monomodal distribution with a polydispersity index (PDI) <0.2 indicating a narrow and homogenous size distribution ( Table 2).

Uptake of liposomes by brain microvascular endothelial cells (BMECs).
Following 60 min incubation on live brain microvascular endothelial cells (BMECs) of Lipo and gH625-Lipo loaded at 5 µM PACAP-Rho, the photomicrographs showed a very low presence of PACAP-Rho in both experimental classes ( Fig. 2A). Note www.nature.com/scientificreports www.nature.com/scientificreports/ that at 1 µM, PACAP-Rho was not detectable within BMECs (not shown). The representative fluorescence images at 5 µM of PACAP-Rho showed that the non-functionalized liposomes undergo more BMEC uptake compared to the gH625-functionalized liposomes ( Fig. 2A).
To observe the effect of liposomes on the cellular organization of BMEC monolayers, we assessed immunofluorescence of the tight junction protein ZO1. Confocal microscopy analysis showed that 60 min incubation on live BMEC monolayers at the highest concentration of 5 µM PACAP-Rho, Lipo and gH625-Lipo caused no qualitative change in ZO1 distribution and presumably no toxicity (Fig. 2B). Note that for both experimental classes, PACAP-Rho fluorescence in BMECs is lower than that in Fig. 2A, presumably due to the membrane permeabilization step with Triton X100 surfactant for ZO1 immunostaining that might disintegrate part of liposomes.
Barrier integrity and transport of liposomes across BMEC monolayers. Before we performed the transport experiments, the cellular toxicity of liposomes was further assessed by monitoring the barrier integrity of the BBB in vitro model. Lipo and gH625-Lipo loaded at 5 µM PACAP were co-incubated with LY for 60 min at 37 °C on BMECs monolayers. Results showed that the LY permeability coefficient values (Pe) were on average 0.24 ± 0.03 × 10 −3 cm/min for both liposome preparations, and similar to the control without liposomes that is below the threshold of LY paracellular leakage. These results showed that neither Lipo alone nor gH625-Lipo loaded at 5 µM PACAP altered the barrier integrity of the BMEC monolayers (Fig. 3A).
For transport studies, gH625-Lipo and Lipo were incubated at 5 µM and 1 µM PACAP-Rho on BMEC monolayers. After incubation, the amount of dye was determined in the cell lysate, and in the luminal and abluminal compartments and compared to the initial amount of dye incubated. Standard curves were performed for each class of liposomes and for free PACAP-Rho and used to obtain accurate concentrations of PACAP-Rho in all compartments of the in vitro BBB model. Results showed that the amount of PACAP-Rho in the luminal compartment after 60 min decreased compared to the initial concentration of PACAP-Rho (Fig. 3B). The amount of PACAP-Rho in BMECs was very low for both liposomes but significantly higher for Lipo than for gH625-Lipo (Fig. 3C). At lower concentration (1 µM) the amount of PACAP-Rho in the abluminal compartment was very low (around 0.41 µM), without significantly difference between the two classes of liposomes. Liposome transport across BMEC monolayers increased with concentration but was statistically significant only for the gH625-Lipo class. At higher concentration (5 µM), the amount of PACAP-Rho that crossed the monolayer was higher for gH625-Lipo than unfunctionalized liposomes (0.93 and 0.64 µM, respectively) (Fig. 3D). The experiments of transport at different times, at 5 µM concentration, show that after each time of incubation the amount of PACAP-Rho in the luminal compartment decreased compared to the initial concentration of PACAP-Rho (Fig. 4A). PACAP-Rho in BMECs was very low for both liposomes at each time point, with significant differences only at 60 min (Fig. 4B). The data from the abluminal compartment indicated that liposome transport was time-dependent. The amount of PACAP-Rho that crossed BMEC monolayers was higher for gH625-Lipo than Lipo also in the first 30 min of incubation (0.72 and 0.60, respectively) (Fig. 4C). After 120 min, the amount of PACAP-Rho detected in the abluminal compartment was higher compared to the 30 min time point for both liposomes but without significant difference between gH625-Lipo and Lipo (Fig. 4C). Data analysis on the abluminal compartment at 5 µM showed that the amount of PACAP-Rho which crosses the BMEC monolayers after 30 min of incubation is 19.4% higher when transported by gH625-Lipo than Lipo (Fig. 5A) and 44.3% higher after 60 min (Fig. 5B). Furthermore, PACAP-Rho transported across BMEC monolayers by gH625-Lipo increased by 29% from 30 min to 60 min and by 36.5% from 60 min to 120 min (Fig. 5C).
Brain distribution of PACAP-Rho in mice after i.v. administration of liposomes. For light sheet fluorescence microscopy analysis, the total volume (µm 3 ) of brains and the total number of cells analysed was reported in Table 3. The results showed an increase of labelled cells within the region of interest (ROI) (Fig. 6). In particular, the administration of Lipo led to release of PACAP-Rho in 1.34% of cells (Fig. 7A); the percentage of cells labelled with PACAP-Rho in the brain of gH625-Lipo treated animals was doubled (Fig. 7A). Fluorescence measured in control brains was considered as background. Integrated density was calculated to detect the amount of fluorescence within the ROI (Fig. 7B). 3D brain volume reconstruction of control, Lipo and gH625-Lipo experimental groups are showed in Supplementary Material section (Movies 1-3).  Table 2. Liposomes size and zeta potential analysis.
The analysis of PACAP-Rho distribution in mice brain BBB vessels is showed in Fig. 8. The inbound fluorescence, which means fluorescence signal within the blood vessels, is significantly higher in gH625-Lipo samples than in control and Lipo class both as mean and integrated density values (Fig. 8A,C). The outbound fluorescence, which means fluorescence outside blood vessels, is significantly higher in gH625-Lipo samples than in control and Lipo class as integrated density values show ( Fig. 8B) but only slightly higher as mean (Fig. 8D).

Discussion
The BBB is a highly selective vascular system; this is beneficial for neural cells that are protected from toxic molecules and pathogens but renders brain drug administration difficult or ineffective. Overcoming this limitation is mandatory and different strategies are evaluated to ameliorate brain drug delivery. Here we present results obtained with liposomes functionalized with a membranotropic CPP (gH625) to deliver the neuroprotective PACAP peptide across the BBB, both in vitro and in vivo.
Our in vitro studies demonstrated that gH625 functionalized liposomes are better nanocarriers for BBB drug delivery than liposomes. The uptake studies first showed the ability of gH625 to favour low retention of PACAP in the BMECs compartment compared to unfunctionalized liposomes, likely changing the mechanism of liposomes uptake in the BMECs. It is of note that gH625 liposomes appeared devoid of toxic effects as indicated by ZO1 www.nature.com/scientificreports www.nature.com/scientificreports/ immunostaining that shows no alteration of cellular organization of BMEC monolayers; this latter result is also supported by the barrier integrity assay, suggesting that gH625 does not impact on the liposome safety profile. gH625 ameliorates the transport of liposomes across BMEC monolayers and increases significantly the amount of PACAP that crosses these monolayers. This occurs in a time dependant manner, suggesting that no receptor is involved in the transport process. Further, we demonstrated that gH625 allows liposomes to transport higher amounts of PACAP beyond the BMEC monolayers at every time point considered.
The in vivo experiments in mice showed that the gH625-liposomal formulation loaded with PACAP improved both PACAP reaching and crossing the BBB, with a higher number of PACAP positive brain cells compared to non-functionalized liposomes. This result demonstrates that gH625 improves brain delivery of liposomes. www.nature.com/scientificreports www.nature.com/scientificreports/ The light sheet fluorescence microscopy analysis showed that gH625 liposomes allowed PACAP-Rho to reach the BBB more efficiently than non-functionalized liposomes; gH625 liposome showed a 2-fold increase in the number of labelled parenchyma cells. Although we did not perform any cito-typing of these cells, the morphology of the labeled cells strongly resembles that of neurons. To our knowledge, this is the first report of effective brain cell labeling beyond the BBB, following in vivo administration of a functionalized nanovector.
We also showed that gH625 liposomes ameliorate the PACAP-Rho reaching of the BBB (about 4-fold) compared to non-functionalized liposomes. Furthermore, gH625 liposomes increase PACAP-Rho capability to overstep the BBB (about 3-fold), compared to non-functionalized liposomes.
Hence, our analysis of the PACAP-Rho distribution in the BBB brain vessels and brain parenchyma, demonstrates that gH625 consistently helps PACAP-Rho to reach and cross the BBB and to enter neuron-like cells.
Although the exact molecular mechanism for the entry of gH625 remains to be established, the results of this study show for the first time that a membranotropic peptide derived from Herpes simplex virus type 1 can improve the transport of liposomes across the BBB endothelium, both in vitro and in vivo in mice.

Materials and Methods
Materials. Fmoc-protected amino acid derivatives, coupling reagents, and Rink amide p-methylbenzhy-  Peptide synthesis. The gH625Cys and PACAP27 peptides were synthesized using a standard solid-phase Fmoc (9-fluorenylmethoxycarbonyl) method as previously reported 67 and were obtained with good yields (30-40%). PACAP27 was prepared with and without Rhodamine labeled (5(6)-Carboxytetramethylrhodamine N-succinimidyl ester). Labeling was performed on resin-bound peptides as previously reported 68 . Peptides were fully deprotected and cleaved from the resin with trifluoroacetic acid. The crude peptides were precipitated with icecold ethyl ether, filtered, dissolved in water, lyophilized, and purified by preparative reverse-phase HPLC. The samples were eluted with a solvent mixture of H 2 O and 0.1% trifluoroacetic acid (solvent A) and CH 3 CN and 0.1% trifluoroacetic acid (solvent B). A linear gradient of 20-80% solvent B over 20 min at a flow rate of 20 ml/  www.nature.com/scientificreports www.nature.com/scientificreports/ min was employed. The collected fractions were lyophilized to dryness and analyzed by mass spectrometry. For the synthesis of DSPE-PEG2000-gH625, DSPE-PEG2000-Mal (1 eq) was reacted with gH625-Cys (1 eq) in DMF containing triethylamine (5 eq) for 24 h. After the completion of the reaction, the solvent was evaporated, and the residue was redissolved in water, lyophilized and analyzed by RP-HPLC. Products were analyzed by MALDI-TOF mass spectrometry (MALDI-TOF MS). The peptides were stored at −20 °C until used.   (Galdiero et al., 2005) 67 . Lipid concentrations of liposome suspensions were determined by phosphate analysis 69 Briefly, various amounts of lipids and when necessary DSPE-PEG2000-gH625 and/or PACAP-Rho were dissolved in chloroform. The solvent was then removed with a nitrogen gas stream and the sample was lyophilized overnight. The lipid film was suspended in buffer by vortexing to produce LUVs, freeze-thawed eight times and then extruded 10 times through polycarbonate membranes with 0.1 µm diameter pores (Northern Lipids). PACAP-Rho encapsulated in liposomes was quantified by ultracentrifugation followed by HPLC analysis. The hydrodynamic diameters (DH) and polydispersity index (PDI) of PACAP-Rho loaded liposomes (Lipo) and PACAP-Rho loaded gH625-liposomes (gH625-Lipo) were measured using dynamic light scattering (DLS) (Malvern Zetasizer Nano ZS, Malven, UK). The analysis was performed with He-Ne laser 4 mW operating at 633 nm at scattering angle fixed at 173 °C and at 25 °C.
Rat in vitro BBB model. Primary rat brain microvascular endothelial cells (BMEC) and primary astrocytes necessary for setting up a rat in vitro BBB co-culture model were obtained as reported (Molino et al. 65,66 . Briefly, primary cultures of BMECs, from 5-to 6-week-old Wistar rats, were seeded in the luminal compartment of twelve-well plate polyethylene insert filters (Merck Millipore, Billerica, MA, USA), pre-coated with collagen type IV and fibronectin (BD Biosciences, Franklin Lakes, NJ, USA) to establish the endothelial cell monolayers. Astrocyte primary cultures were prepared from the cerebral cortex of newborn Wistar rats and seeded in the bottom of the twelve-well plates to establish the co-culture with the endothelial cell monolayers in endothelial cell media (ECM) containing DMEM/F12 supplemented with 20% bovine platelet poor plasma derived serum (Alfa Aesar, Ward Hill, MA, USA), basic fibroblast growth factor (bFGF) 2 ng/mL, heparin 100 μg/ mL, gentamycin 50 μg/mL, HEPES 2.5 mM, and hydrocortisone 500 nM (Life Technologies, Carlsbad, CA, USA). Under these conditions, the BMEC monolayers differentiate, express junction-related proteins within 3 days, and remain optimally differentiated during three more days.
Immunofluorescence and confocal microscopy analysis. Uptake of liposomes by BMECs. Lipo and gH625-Lipo loaded at 1 µM and 5 µM PACAP-Rho, were incubated for 60 min at 37 °C on BMEC monolayers. After incubation, the cells were fixed for 15 min with a 4% (w/v) paraformaldehyde solution followed by cutting of the insert membranes and nuclei staining with Höechst 33258 (Invitrogen, Carlsbad, CA, USA). The membranes were washed three times with phosphate-buffered saline and then mounted on microscope slides. The analyses of slides were performed using confocal microscopy LSM 780 (Zeiss, Jena, Germany) equipped with a 63x oil-immersion objective, and ZEN Software 2012. Images were acquired using filters for rhodamine channel (λ ex 544 nm; λ em 584 nm) and Höechst channel (λ ex 350 nm; λ em 461 nm). www.nature.com/scientificreports www.nature.com/scientificreports/ ZO1 Immunofluorescence. Zonula occludens (ZO1)-1 immunofluorescence was also performed to evaluate tight junction morphology and liposome toxicity. Lipo and gH625-Lipo, loaded at 5 µM PACAP-Rho, were incubated on BMECs for 60 min at 37 °C. After incubation, the insert filters with BMECs were washed twice with phosphate-buffered saline and fixed with a 4% (w/v) paraformaldehyde solution for 15 min. Cells were then permeabilized for 10 min with PBS 0.1% Triton X100 followed by washing with PBS twice. Blocking was performed using 3% BSA for 30 min followed by PBS wash and incubation with the ZO1 antibody (Rabbit anti-ZO1 Zymed 61-7300) diluted 1:200 in a solution of 1% BSA for 60 min at room temperature. After two washes, cells were incubated with AlexaFluor 488 (1:800) secondary antibodies and Höechst (1:1000) for 30 min at room temperature, followed by three washes with PBS. Finally, cells were mounted with ProLong ® Gold Antifade Mountant (Life Technologies) and visualized under confocal microscopy (LSM 780; Carl Zeiss, Jena, Germany) equipped with a 63x oil-immersion objective by using filters for Alexa 488 channel (λ ex 544 nm; λ em 584 nm) and Höechst channel (λ ex 350 nm; λ em 461 nm).
Barrier integrity assay. In order to control the BMEC monolayer integrity of the in vitro BBB model and absence of toxicity of liposomes, transport assays were performed as previously described (Molino et al.,65,66 ). Briefly, the BMEC monolayers were gently washed with pre-warmed DMEM/F12 without phenol red and transferred to clean twelve-well plates. Lucifer yellow (LY; Sigma-Aldrich) was co-incubated with Lipo or gH625-Lipo,

Quantification of liposome uptake by and transport across BMEC monolayers.
To evaluate their transport across BMEC cell monolayers, Lipo and gH625-Lipo were incubated for 60 min at 37 °C at 5 µM and 1 µM PACAP-Rho in the luminal compartment of twelve-well plate polyethylene insert filters. After incubation, the medium was collected from the luminal and abluminal compartments from each group and cell monolayers rinsed twice with phosphate-buffered saline and then lysed with 500 μL/insert PBS 0.1% Triton X100 in order to release the internalized peptide. The fluorescence of medium and cell lysates was quantified by SpectraMax ® M5e (Molecular Devices, Sunnyvale, CA, USA) equipped with a fluorescence filter (λ ex 544 nm; λ em 584 nm). Furthermore, liposomes loaded with 5 µM PACAP-Rho were incubated for 30, 60 and 120 min at 37 °C to evaluate the passage across BMEC monolayers at different times. Three BBB in vitro models were used for each time condition. Fluorescence values were normalized using the equation of the standard curve for each class of liposomes. The standard curve was obtained using liposomes with different concentrations of PACAP-Rho loaded, i.e., 0.156 μM, 0.312 μM, 0.625 μM, 1.25 μM, 2.5 μM, and 5 μM. As negative controls, the DMEM/Ham's F12 without phenol red was used. Each experiment was repeated five times.

Distribution of PACAP-Rho in the cortex of mice after intravenous (i.v.) administration of liposomes.
Mice of mean 35 g were maintained in a temperature-controlled room at 24 °C, with 12-hours light/dark cycle and with free access to a standard rodent diet and water before experiments. Animals were randomly divided in 3 groups, each consisting of 5 mice. Unfuctionalized and gH625 functionalized liposomes loaded with 100 µM PACAP-Rho were intravenously injected in the tail vein (bolus of 150 μl). All liposome formulations contained the same lipid concentration and 0.9% saline solution were used as negative control. After 90 min from liposomal injection, all mice were intraperitoneally anesthesied at a letal dose of Ketamine (100 mg/kg) and Xylasine (10 mg/ kg) solution. Mice were then intracardially perfused with 0.9% saline solution following by a 4% paraformaldehyde solution. Brains were then postfixed overnight at 4 °C with 4% paraformaldehyde. Then after 3 washes in PBS 1X, brains were prepared for light sheet fluorescence microscopy analysis.
Brain clearing and light sheet fluorescence microscopy analysis. For light sheet fluorescence analysis treated mouse brains were first dissected sagittally, and each part was dissected coronally approximately to 1.0 mm from Bregma (http://mouse.brain-map.org/). Samples were stained with Höechst for 4 hours and then cleared by ScaleA2 protocol, an urea-based solution that renders mouse brains transparent 70 . ScaleA2 solution was prepared using 4 M urea, 10% (wt/vol) glycerol and 0.1% (wt/vol) Triton X-100. After two weeks, samples were mounted in the Z1 Light Sheet Fluorescence Microscope (Zeiss) in the imaging chamber. Samples were overviewed and centered by led light illumination for acquisition of coronal stacks, and illuminated by λ ex 488 nm, λ ex 405 nm and λ ex 561 nm laser sources, respectively; corresponding fluorescence signals were detected through a 20X NA = 1.0 Zeiss water immersion objective and images captured by PCO.Edge sCMOS water cooled cameras. Raw data were acquired and analyzed by ZEN 2012 software (Zeiss); Eight-bit fluorescence image stacks were calibrated using Rodbard NIH Image optical density (O.D.) curve between 0 and 2,58 and then analysed with Fiji software; 71 nuclei were counted in each volume stack using the ROI manager3D (v.3.92) and mean (the fluorescence of samples as average gray value within the sample) and integrated density (as the product of mean grey value of ROI and ROI area) evaluated on thresholded volume stacks to avoid background artifacts. To estimate the amount of PACAP-Rho that crossed the BBB, we considered the mean and the integrated density of specific ROIs: in detail, to evaluate the PACAP-Rho which reached the BBB without crossing it, the selected ROI was considered as the inside of the brain blood vessels (inbound), exclusively; then, on the non-zero thresholded volume stacks, an inversion of selected ROI was carried out to obtain the ROI where to calculate the amounts of PACAP-Rho that crossed the BBB (outbound), exclusively.