BASIC SCIENCE

Nanomedicine: Nanotechnology, Biology, and Medicine11 (2015) 391–400

Original Article

Monocyte mediated brain targeting delivery of macromolecular drug forthe therapy of depression

Jing Qin, PhDa, Xu Yang, MDa,c,d, Ruo-Xi Zhang, PhDb, Yi-Xiao Luo, PhDb, Jia-Li Li, PhDb,Jia Hou, MDa,c,e, Chun Zhang, MDa, Yong-Ji Li, PhDc, Jie Shi, PhDb, Lin Lu, MD, PhDb,

Jian-Xin Wang, PhDa,⁎, Wei-Li Zhu, PhDb,⁎⁎aDepartment of Pharmaceutics, School of Pharmacy, Fudan University; Key Laboratory of Smart Drug Delivery, Ministry of Education, Shanghai, China

bNational Institute on Drug Dependence, Peking University, Beijing, ChinacDepartment of Pharmaceutics, School of Pharmacy, Heilongjiang University of Chinese Medicine, Haerbin, Heilongjiang Province, China

dDepartment of Pharmacy, The Fifth People's Hospital of Shanghai, Fudan University, Shanghai, ChinaeDepartment of Pharmacy, Municipal Hospital, Ministry of Healthcare, Weihai, Shandong Province, China

Received 1 February 2014; accepted 22 September 2014

Abstract

Leukocytes can cross intact blood-brain barrier under healthy conditions and in many neurological diseases, including psychiatric diseases. In presentstudy, a cyclic RGD (cRGD) peptidewith high affinity for integrin receptors of leukocyteswas used tomodify liposomes. The cRGD-modified liposomes(cRGDL) showed high affinity for monocytes in vitro and in vivo and co-migrated across in vitro BBBmodel with THP-1. The trefoil factor 3 (TFF3), amacromolecular drug, was rapidly and persistently delivered to brain for at least 12 h when loaded into cRGDL while 2.8-fold increase in drugconcentration in basolateral amygdala regions related to depression was observed. A systemic administration of cRGDL-TFF3 mimicked antidepressant-like effect of direct intra-basolateral amygdala administration of TFF3 solution in rats subjected to chronic mild stress. The effective dual-brain targetingdelivery resulting from the combination and co-migration of cRGDL with leukocyte cross BBBmay be a promising strategy for targeted brain delivery.

From the Clinical Editor: In an effort to treat depression, brain targeted delivery via monocyte-cRGD liposome complexes capable of crossing theintact BBB was performed in this study in a murine model. Similar approaches may be helpful in the treatment of other neuropsychiatric conditions.© 2015 Elsevier Inc. All rights reserved.

Key words: cRGD; brain-targeting drug delivery system; monocyte; depression; trefoil factor 3

nanomedjournal.com

Depressive disorder reportedly affects approximately 16% of thepopulation, leading to amajor burden on health.1 Peptide and proteindrugs have shown great promise for the treatment of variousneuropsychiatric diseases. However, despite their potential, a majorchallenge is the delivery of peptide and protein drugs across theblood-brain barrier (BBB),2 which has been considered as the mostimportant barrier that impedes drug transport into the brain via theblood circulation.3,4 The BBB can exclude from the brain 100% of

Abbreviations: cRGD, cyclic RGD; cRGDL, cRGD-modified liposomes; TFF3,brain barrier; i.p, intraperitoneal; HBMEC, human brain microvessel endothelial cell; Pphosphatidylcholine; EE, entrapment efficiency; UPLC, ultra-performance liquid chrosuccinimidyl ester; BLA, basolateral amygdala; OB, olfactory bulbectomy; CMS, chmolecule; ICAM, intercellular adhesion molecule.

Fundings: This work was supported by the National Natural Science Foundati(2013CB932502), Asia 3 Foresight Program (No. 81361140344) Funded by NationalProgram of Higher Education (No. 20110071120042), School of Pharmacy of FudanFudan University, and the Ministry of Education National Science and Technology M

Competing interests: The authors declare that they do not have any conflict⁎Correspondence to: J.Wang, Department of Pharmaceutics, School of Pharmacy, Fud

⁎⁎Correspondence to: W.L. Zhu, National Institute on Drug Dependence, PekE-mail addresses: jxwang@fudan.edu.cn (J.-X. Wang), zhu_wl@bjmu.edu.

http://dx.doi.org/10.1016/j.nano.2014.09.0121549-9634/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Qin J, et al, Monocyte mediated brain targeting delivNBM 2015;11:391-400, http://dx.doi.org/10.1016/j.nano.2014.09.012

large-molecule neurotherapeutics and more than 98% of all small-molecule drugs.4 Many existing peptide and protein drugs arerendered ineffective in the treatment of these clinical disordersbecause of an inability to effectively deliver and sustain themwithinthe brain.2 Thus, the need for a brain-targeting drug delivery systemis important to effectively treat brain-related diseases.

Trefoil factor 3 (TFF3), a member of the trefoil factor family, isproduced by intestinal goblet cells and used for gastrointestinal

trefoil factor 3; cRGDL-TFF3, cRGD liposomes loaded with TFF3; BBB, blood-L, plain liposomes; PL-TTF3, plain liposomes loaded with TFF3; SPC, soybeanmatography; DLS, dynamic light scattering; CFSE, carboxyfluorescein diacetateronic mild stress; CNS, central nervous system; VCAM, vascular cell adhesion

on of China (No. 81102386), National Key Basic Research Program of ChinaNatural Science Foundation of China, Specialized Research Fund for the DoctoralUniversity, the Open Project Program of the Key Lab of Smart Drug Delivery ofajor Project (No. 2012ZX09103301 to Prof. Jie Shi).s of interest.anUniversity, Smart DrugDelivery,Ministry of Education, Shanghai 201203, China.ing University, Beijing 100191, China.cn (W.-L. Zhu).

ery of macromolecular drug for the therapy of depression. Nanomedicine:

392 J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

disease therapy.5,6 TFF3 injectionwas recently found to induce Fos-positive cells in magnocellular oxytocin neurons in the hypothal-amus, prompting further research on the role of TFF3 in the centralnervous system and related disorders.7 Our recent studies usedseveral behavioral models of depression and suggested that systemicadministration (intraperitoneal, i.p.) of high doses of TFF3 producedantidepressant-like effects in both mice and rats.8 However, thisantidepressant-like effect was not observed in depression modelswhen the dosage was reduced to 0.05 mg/kg. Therefore, a highconcentration in brain regions related to depression appears to benecessary for the practical application of TFF3.

The neuroinflammation occur under many neurological diseasesincluding depression.4,9,10 A typical feature of neuroinflammationresponse is the recruitment of leukocytes (mainly monocytes andneutrophils) to the lesions in brain. Perivascularmacrophages, whichreside on the parenchymal side of endothelia cells close to astrocyteendfeet, originally derive from circulating phagocytes, such asmonocytes, and have shown a remarkable capability to cross anintact BBB, with 80% turnover in 3 months.11 Thus, one strategy todeliver drugs to the brain under pathological conditions is to exploitthese inflammatory cells as targeted delivery systems.

The RGD (Arg-Gly-Asp) peptide can combine with integrinreceptors that are expressed on the surface of leukocytes(including monocytes and neutrophils).12 Liposomes modifiedwith RGD, therefore, may possibly be developed for selectiveand preferential presentation to blood monocytes/neutrophils.Subsequently, liposomes can be taken into the brain in responseto the recruitment of inflammation processes. In our previousstudy, obvious brain-targeting drug delivery was obtained usingRGD liposome.13

In the present study, a cyclic RGD (cRGD, Arg-Gly-Asp-D-Phe-Lys) peptide with higher affinity for integrin receptorscomparing with RGD14 was used as a ligand to be coupled withliposomes. Furthermore, the possible mechanisms of brain-targeted delivery were investigated in vitro and in vivo using ahuman acute monocytic leukemia cell line (THP-1), human brainmicrovessel endothelial cell (HBMEC) and inflammation rats.Then, the pharmacological effects of cRGDL loaded with TFF3(cRGL-TFF3) were studied. The aim of the present study was toverify the brain-targeting effect and potential mechanism ofcRGDL and study whether cRGDL enhance the behavioralresponse to TFF3.

Methods

Animals and cells

Male Sprague-Dawley rats (200-220 g) or male nude mice(20 ± 2 g) were housed under a constant temperature (23 ± 2 °C)and a 12 h/12 h light/dark cycle with free access to food andwater. All of the animal experiments were performed incompliance with the National Institutes of Health Guide for theCare and Use of Laboratory Animals, and the procedures wereapproved by the Peking University Animal Use Committee(LA2012/21) or by the Ethics Committee of Fudan University(SYXK-2010-0099).

THP-1 cell line was obtained from the Institute of Biochemistryand Cell Biology (Shanghai, China) and cultured as previously

described.15 HBMEC was from ScienCell (Carlsbad, CA, USA)and cultured on upper chamber of Transwell insert of purifiedcollagen with endothelial cell medium.16,17

Liposome preparation

Plain liposomes (PL) loaded with TFF3 (PL-TTF3) wereformulated with soybean phosphatidylcholine (SPC), DSPE-PEG(PEG 2000), DSPG (Merck, Schaffhausen, Switzerland) andcholesterol and were prepared using a modified reverse-phaseevaporation method described previously.18 Briefly, A mixture ofphospholipids and cholesterol,19 including DSPE-PEG 2mol% oftotal lipid, was dissolved in 15 ml chloroform and 5 ml phosphatebuffer (pH 7.4) containing TFF3 (10:1 lipid:drug molar ratio,Beijing Yong Kang Jia Xin Science and Technology Develop-ment, Beijing, China). The resulting two-phase system wassubjected to bath-type sonication and then was placed in a rotaryevaporator under reduced pressure for at least 12 h until ahomogeneous suspension of PL-TFF3 was obtained.

For the preparation of cRGDL-TFF3, 2mol% cRGD-PEG-DSPE (GL Biochem, Shanghai, China) was used in the lipidformulation instead of DSPE-PEG.13

For the preparation of fluorescence PL or cRGDL, coumarin6, TFF3-Cy5 or TFF3-Cy7 was loaded into liposomes to obtainC6-PL, C6-cRGDL; PL-TFF3-Cy5, cRGDL-TFF3-Cy5; PL-TFF3-Cy7 and cRGDL-TFF3-Cy7. C6-PL non-DSPG and C6-cRGDL non-DSPG present the lipid formulation without theaddition of DSPG. For the preparation of liposomes of variousparticle sizes, different ultrasonic times were used. Finally, C6-loaded liposomes of different particle size (300, 200, 100, and50 nm) were obtained for further study.

Liposome characterization evaluation

The entrapment efficiency (EE) was determined by sizeexclusion chromatography and ultra-performance liquid chro-matography (UPLC) method. The equation for calculating EEwas the following:

EE ¼ Winterior=Wtotal � 100%

Winterior represents the intraliposomal content of TFF3, andWtotal

represents the total content in the liposomal suspension whenTriton-100 was added to the suspension. The mean diametersand zeta potential of PL-TFF3 and cRGDL-TFF3 weredetermined by dynamic light scattering (DLS) using a ZetaSizerNano series 3600 (Malvern, GmbH, Herrenberg, Germany).

Uptake and competition assays of liposomes on leukocytesin vitro

THP-1 cells were used to investigate the possible mechanismof the uptake of liposomes by leukocytes. C6-PL or C6-cRGDLwith different lipid concentrations (0.66, 1.32, and 1.97 μmollipid/L) was added to THP-1 cells for 1h incubation at 37 °C.After incubation, the liposome suspension was removed, and thecells were washed three times with PBS by centrifugation. Thecells were re-suspended and visualized under an IX2-RFACAfluorescent microscope (Olympus, Osaka, Japan). The fluores-cence intensity of THP-1 of C6-PL group and C6-cRGDL group

393J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

was assayed by flow cytometer (FACSCalibur, BD Biosciences,Bedford, MA, USA) and the data was analyzed statistically.

C6-PL, C6-cRGDL, C6-PL non-DSPG, C6-cRGDL non-DSPG of different particle sizes (50, 100, 200, and 300nm) wereadded to THP-1 cells for 1 h incubation at 37 °C. After incubation,the stained cells were analyzed by flow cytometer.

For the competition assay, free cRGD peptide at aconcentration of 500 μg/ml was added to the wells in advanceor simultaneously with liposomes (C6-PL or C6-cRGDL) for 1 hincubation at 37 °C. After incubation, the free cRGD peptide andnon-combinative liposomes were removed. The stained THP-1was washed three times by centrifugation and analyzed using aflow cytometer. C6-PL or C6-cRGDL was also incubated withTHP-1 cells at 4 °C or 37 °C separately. THP-1 cells alsounderwent the same treatment and analyzed by flow cytometer.

Brain targeting effect in vitro

HBMEC was used as a BBB model in vitro to study themechanism of cRGD liposome crossing blood vessels. Astandard chemotactic factor, 10-7 M fMLP was placed in thelower chamber under the HBMEC monolayers to establish achemotactic gradient for monocyte migration across BBBin vitro.17 THP-1 was stained with cell tracer (carboxyfluoresceindiacetate, succinimidyl ester, CFSE. Invitrogen, Carlsbad, CA,USA). Various liposomes (TFF3-Cy5 solution, PL-TFF3-Cy5 orcRGDL-TFF3-Cy5) and stained THP-1 were added into upperchamber above confluent HBMECmonolayers and incubated for3h. After incubation, the suspension in the lower chamber wasvisualized by confocal scanning laser microscopy (Leica TCSSP2, Germany) and then subjected to ultrasonication fordetermining Cy5 fluorescence with a microplate reader (ThermoMultiskan MK3, Thermo Scientific, Waltham, MA, USA).

Brain targeting effect in vivo

Nine male nude mice were received intrastriatal microinjec-tions of 10 U of human recombinant IL-1β (Sigma St. Louis,MO, USA) to produce brain inflammation10 and divided intothree groups, followed by intraperitoneal administration of theTFF3-Cy7 solution, PL-TFF3-Cy7 and cRGDL-TFF3-Cy7suspensions (equivalent to 50μg/mlTFF3-Cy7). Fluorescentscans were performed 1, 3, 6, and 12 h after administration ofliposomes using an in vivo image system (Caliper IVISSpectrum, Caliper Life Science, Hopkinton, MA USA).

Brain distribution assay

Nine rats were treated with olfactory bulbectomy (OB)procedure as described in supplementary materials and wereallowed 14-days recovery. Then, the OB rats were divided intothree groups and were administrated of TFF3 solution, PL-TFF3or cRGL-TFF3 suspensions (equivalent to 0.05 mg/kg TFF3, i.p.) for 7 days, respectively. After 7-days administration, rats wassacrificed and the tissues from basolateral amygdala (BLA)regions were collected and homogenized with an electricaldisperser after being lysed with RIPA lysis buffer (BeyotimeBiotechnology, Haimen, Jiangsu Province, China) for 30 min.The homogenate was subjected to 10,000 × g centrifugation at4°C for 20 min and pre-separated by mini-C18 column for UPLC

assay. All of the above procedures were performed under lowtemperature (0-4°C). The TFF3 concentrations of all of thesamples were determined using UPLC method.

Combination of liposomes with leukocytes in vivo

Nine male Sprague-Dawley rats were divided into threegroups and received an intrastriatal microinjection of 100 U ofhuman recombinant IL-1β to produce brain inflammation,followed by intraperitoneal administration of the TFF3-Cy5solution, PL-TFF3-Cy5 and cRGDL-TFF3-Cy5 suspensions.Leukocytes were isolated from anti-coagulated peripheral bloodfrom rats by centrifugation in a gradient of OptiPrep™ (AXIS-SHIELD PoC AS, Norway) at 0.5, 3, and 6 h afteradministration. The fluorescence of monocyte/neutrophil wasassayed by flow cytometer analysis.

Behavioral evaluation

Forced swim testThe rats were placed in a 25 cm diameter × 65 cm height

plastic cylinder that was filled to a depth of 30 cm with 23-25°Cwater for 15 min. Twenty-four hours later, the rats received asingle dose of the TFF3 solution (0.05 mg/kg, i.p.) and PL-TFF3and cRGDL-TFF3 suspensions (equivalent to 0.05 mg/kg TFF3,i.p.) before being tested in this cylinder as previouslydescribed.20 Immobility was defined as the minimum movementrequired to passively keep the animal’s head above the waterwithout other motions. Swimming was related to the rats’ activebehavior to escape from the water. Climbing was defined asupward-directed movements of the forepaws against the wall.The results are expressed as the time (in seconds) that the animalsspent immobile during the 5 min test.

Locomotor activity assay in olfactory bulbectomized rats

The rats were anesthetized with sodium pentobarbital (50mg/kg,i.p.), and bilateral OB was performed as described in supplementarymaterials. After OB procedure and 14-days recovery, the rats weretreated with the TFF3 solution, PL-TFF3 or cRGDL-TFF3suspensions (equivalent to 0.05 mg/kg TFF3, i.p.) for 7 days beforelocomotor activity (seen in supplementary materials)21 wasassessed.

Antidepressant-like effect evaluated in rats subjected to chronicmild stress

The chronic mild stress (CMS) paradigm of 28-day was basedon our earlier study (seen in supplementary materials).22 Controlrats that received normal handling were not subjected to anystress in the housing room. After the 28-day stress paradigm,TFF3 solution (1 ng/side) was intracranially microinjected using10 μl Hamilton syringes (Hamilton, Reno, NV, USA) that wereconnected via polyethylene-50 tubing to 30-gauge injectors(Plastics One, Roanoke, VA, USA).23 A total volume of 0.5 μlwas infused into each side over 1 min, and the injection syringewas left in place for an additional 1 min to allow for diffusion.PL-TFF3 and cRGDL-TFF3 were intraperitoneally injected(equivalent to 0.05 mg/kg TFF3).

After administration, the rats were subjected to the forcedswim test and sucrose preference test24 (seen in supplementary

Figure 1. The influence of liposome characterization on cellular uptake on THP-1. (A) Cellular uptake of C6-PL and C6-cRGDL examined by fluorescentmicroscopy and flow cytometry (B). Scale bar =50μm. (C) Cellular uptake of C6-PL and C6-cRGDL of different particle sizes with or without the addition ofDSPG (1h incubation with THP-1 cells at 37°C) examined by flow cytometry.

394 J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

materials). At the end of sucrose preference test, sucrose andwater consumption (in milliliters) was measured, and sucrosepreference (%) was calculated as the ratio of sucrose consump-tion to sucrose plus water consumption.

Results

Characteristics of PL-TFF3 and cRGDL-TFF3

The EE of PL-TFF3 and cRGDL-TFF3 were 25.3% and27.6%, respectively, containing about 63 μg/ml TFF3. Theaverage size and zeta potential for PL-TFF3/cRGDL-TFF3 were129.4/133.0 nm and -22.6/-21.8 mv, respectively. No obviousdifferences in EE, the average size, and the zeta potential wereobserved between the two types of liposomes.

Characterization of liposomes uptake by THP-1 cells

As shown in Figure 1, A-B, the cellular uptake of C6-cRGDLwas concentration-dependent and significantly higher than C6-PL when the concentration of the lipid ranged from 0.66 to 1.97μmol/L. The liposomes characterized by average particle sizes of100 to 200 nm were more suitable for effective uptake by THP-1cells, comparing with those of less than 50 nm average particlesizes. Furthermore, the addition of the negative lipid (DSPG)increased the uptake of both PL and cRGDL by THP-1 cells. Themodification of cRGD on liposomes significantly increased theuptake (Figure 1, C). Therefore, cRGDL with 100-200 nmparticle size and the addition of DSPG exhibited the highest

uptake by THP-1 cells. It indicated that liposomes of about 130nm particle size and -20 mv zeta potential obtained in this studywere suitable for effect uptake by leukocytes.

Competition assay

The addition of free cRGD peptide in advance significantlyinhibited the uptake of C6-cRGDL by THP-1 cells, whereas ithad no obvious influence when the free peptide was added toTHP-1 cells simultaneously with C6-cRGDL (Figure 2, A). Itsuggests that the same binding site was used for free cRGD andcRGDL while the binding ability of cRGDL was not influencedby free cRGD peptide. Additionally, the cellular uptake ofcRGDL at 4°C was much lower than that at 37°C (Figure 2, B).The addition of free peptide and different temperatures had noeffect on the uptake of PL. This indicates that the modification ofcRGD on liposomes significantly enhanced the combination ofliposomes with THP-1 cells, and this combination was in anactive manner.

Brain targeting effect in vitro

HBMECwas a well-established BBBmodel in vitro17 and wasused to study the co-migration of THP-1 and liposomes crossingblood vessel. The Cy5 fluorescence of cRGDL-TFF3-Cy5 wasstrongest among three groups while almost no fluorescence wasdetermined for TFF3-Cy5 (Figure 2, C). The fluorescence signalof cRGDL-TFF3-Cy5 (red) overlapped with that of THP-1 (green,Figure 2,D, white arrow) while the fluorescence of PL-TFF3-Cy5

Figure 2. The combination mechanism and co-migration across BBB of cRGDL with THP-1 in vitro. Competition assay with the addition of free cRGD peptideto THP-1 cells in advance or simultaneously with liposomes (A) and the influence of incubation temperature on the uptake of liposomes by THP-1 cells (B). Thefluorescence intensity of TFF3-Cy5, PL-TFF3-Cy5 and cRGL-TFF3-Cy5 in the lower chamber (C) and the co-location of migrated THP-1 (CFSE, Green) andvarious liposomes (Cy5, Red) (D). The fluorescence signal of cRGDL overlapped with that of THP-1 (white arrow); the fluorescence of PL-TFF3-Cy5 wasobserved at distance from THP-1 (white arrow); The fluorescence of Cy5 wasn't observed for TFF3-Cy5.

395J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

was observed at distance fromTHP-1 (Figure 2,D, white arrow). Itsuggested that more cRGDL could cross HBMEC resulting fromthe co-migration with THP-1 compared with PL. It is difficult forTFF3 to cross HBMEC. Hence, cRGDL greatly enhanced abilityof crossing BBB of TFF3, which resulting from the combination ofcRGDL with THP-1.

In vivo brain-targeting effect and brain distribution

TFF3-Cy7, PL-TFF3-Cy7, and cRGDL-TFF3-Cy7 were usedto investigate the in vivo brain-targeting effect of cRGDL innude mice treated with IL-1β.10 Evident fluorescence was

rapidly observed in the cRGDL-TFF3-Cy7 group at 30 min, witha peak at 6 h after administration, and the fluorescence lasted for12 h (Figure 3, A). The fluorescent signal of the PL-TFF3-Cy7group was much lower than of cRGDL-TFF3-Cy7 group andhardly observed after 3 h. In the TFF3-Cy7 group, almost nofluorescence signal could be observed after administration.

Furthermore, brain distribution results showed that theconcentration of cRGDL-TFF3 in BLA of brain was 2.8-foldcompared with that of TFF3 and 1.73-fold compared with that ofPL-TFF3 (Figure 3, B). It meant that cRGL can enhance the drugconcentration both in brain and in brain subregion relative to thedepression. Hence, a dual-brain targeting effect was observed.

Figure 3. Brain targeting effect and its mechanism investigation. (A) In vivo fluorescence imaging of IL-1β treated nude mice. (B-C) Flow cytometry assay ofmonocytes and neutrophils isolated from plasma from IL-1β-treated rats at different times after TFF3-Cy5, PL-TFF3-Cy5, and cRGDL-TFF3-Cy5 administration.

396 J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

Combination of liposomes with leukocytes in vivo

For targeting mechanism study, monocytes and neutrophilswere isolated to determine the fluorescence intensity of TFF3-Cy5.For monocytes, as shown in Figure 3, C-D, cRGDL-TFF3-Cy5group had the strongest fluorescence among the three groups, andthe fluorescence intensity wasmaintained at a high level within 6 h.The fluorescence of PL-TFF3-Cy5 group was much lowercompared with cRGDL-TFF3-Cy5 and rapidly decline after 3 h.For neutrophils, cRGDL-TFF3-Cy5 and PL-TFF3-Cy5 did notshow strong combination. TFF3-Cy5 combined with neitherneutrophils nor monocytes.

cRGDL enhanced the antidepressant-like effects of TFF3 in theforced swim test

A single low dose of TFF3 (0.05 mg/kg, i.p.) did not changeimmobility, swimming time, or climbing time compared withsaline-treated rats (Figure 4, A-C). This is consistent with ourprevious behavioral finding that a dose of TFF3 lower than 0.1mg/kg did not produce antidepressant-like effects in rats.8 Asingle acute injection of PL-TFF3 (equivalent to 0.05 mg/kgTFF3, i.p.) 30 min before the forced swim test decreasedimmobility time (one-way analysis of variance [ANOVA],

P b 0.05), increased swimming time (P b 0.05), and increasedclimbing time (P b 0.05;). Additionally, cRGDL-TFF3 at a doseequivalent to 0.05 mg/kg TFF3 also significantly decreasedimmobility time (P b 0.001), increased swimming time(P b 0.01), and increased climbing time (P b 0.01) in the forcedswim test. Notably, rats treated with cRGDL-TFF3 exhibited amuch lower immobility time than rats treated with PL-TFF3(P b 0.05; Figure 4, A). These findings suggest that acutecRGDL-TFF3 administration produced a more effective antide-pressant response than PL-TFF3. Acute TFF3, PL-TFF, andcRGDL-TFF3 did not alter normal locomotion in the 5 min test(Figure 4, D), indicating that the decreases in immobility werenot associated with general increases in locomotor activity.

Antidepressant-like effects of cRGDL-TFF3 in olfactorybulbectomized rats

We further evaluated the behavioral effects of cRGDL-TFF3treatment in OB rats to investigate the potential antidepressant-like action of cRGDL-TFF3. The behavioral data of rats exposedto the OB model demonstrated a higher level of locomotion,reflected by total distance travelled compared with the sham-saline group (P b 0.01; Figure 5). However, the hyperactivity

Figure 4. cRGDL-TFF3 produced significant antidepressant-like effects in the forced swim test in rats. Acute TFF3 administration did not affect immobility,swimming time, or climbing time. Both PL-TFF3 and cRGDL-TFF3 significantly decreased immobility time (A), increased swimming time (B), and increasedclimbing time (C). (D) Acute TFF3, PL-TFF3, and cRGDL-TFF3 did not alter normal locomotion in the 5 min test. The data are expressed as mean ± SEM (n =8 per group). *P b 0.05, **P b 0.01, compared with saline-treated group.

397J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

induced by the OB procedure was significantly attenuated bytreatment with PL-TFF3 (0.05 mg/kg, i.p.; P b 0.05) andcRGDL-TFF3 (0.05 mg/kg, i.p.; P b 0.01) for 7 days. Thedifference between the PL-TFF3 and cRGDL-TFF3 groups wasnot significant but showed a trend toward a decrease in the totaldistance travelled. No significant difference was found betweenthe TFF3 solution (0.05 mg/kg, i.p.) and OB-saline rats.

cRGDL-TFF3 produced similar antidepressant-like effects asTFF3 microinjection into the basolateral amygdala

We further explored the possibility that cRGDL-TFF3produces significant antidepressant-like effects by directlyinfusing TFF3 solution into BLA. Microinjection of TFF3directly into the BLA decreased immobility time in the forcedswim test in rats, indicating a marked antidepressant-like effect.8

Therefore, we investigated whether systemic administration ofcRGDL-TFF3 produces an antidepressant response that issimilar to intra-BLA TFF3 infusion. The CMS model ofdepression is widely used to assess the possible antidepressant-like effects of drugs. In the present study, we found that ratssubjected to CMS for 28 days spent more time immobile in theforced swim test compared with control rats (P b 0.01). Intra-BLA infusion of TFF3 (1 ng/side; P b 0.05) and intraperitonealcRGDL-TFF3 administration (0.05 mg/kg; P b 0.05) reversedthe increase in immobility time induced by CMS compared withsaline-treated rats (Figure 6, B). We also found that acute intra-BLA TFF3 infusion (P b 0.01) and intraperitoneal cRGDL-TFF3 administration (P b 0.01) increased sucrose preference(Figure 6, C), which was reduced by CMS (P b 0.001), but didnot change total water intake (Figure 6, D). A single acute PL-TFF3 administration did not affect immobility time in ratsexposed to CMS or alter sucrose preference. These findings

suggest that systemic cRGDL-TFF3 administration mimickedthe antidepressant-like effect of intra-BLA TFF3 administra-tion, in which the depressive-like behavior induced by CMSwas reversed.

Discussion

The BBB has been considered as the most important barrierthat impedes macromolecular drug transport into the brain via theblood circulation. The need for a brain-targeting delivery systemfor macromolecular drugs is apparent because both peptide andprotein drugs show great promise for the treatment of variousneurodegenerative diseases. Accumulating evidence indicatesthat leukocytes (monocytes and neutrophils) can cross an intactBBB under healthy conditions and in many neurologicaldiseases, including psychiatric diseases.9,10,25 The interplaybetween the immune and nervous systems plays a pivotal role inthe pathophysiology of depression.26,27

THP-1 cells have become one of the most widely used celllines to investigate the functional regulation of monocytes.28

In the present study, THP-1 cells were used to investigate theuptake mechanism of liposomes by monocytes in vitro. Theresults showed that the addition of a negative lipid (DSPG) and100-200 nm particle size were more suitable for liposomes tocombine with THP-1 cells (Figure 1). These results are consistentwith a previous study.19 Hence, negatively charged liposomeswitha 100-200 nm particle size were prepared in this study.

Furthermore, the competition assay results (Figure 2, A-B)showed that the modification of cRGD was the main contributorto the binding of cRGDL to THP-1 cells. The addition of freecRGD peptide in advance inhibited the uptake of cRGDL by

Figure 5. Behavioral effects of cRGDL-TFF3 for 7 days on locomotion in OBrats. The data are expressed as mean ± SEM. n = 8. *P b 0.05, **P b 0.01,compared with OB group; ##P b 0.01, compared with sham-saline group.

398 J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

THP-1 cells. However, the binding activity of cRGD onliposomes did not decline. Free cRGD peptide did notsignificantly influence the combination efficiency of cRGDLwith THP-1 cells when simultaneously added to THP-1 cellswith cRGDL.

HBMEC was a well-established BBB model in vitro to studythe mechanisms by which leukocytes interacted with microvas-culature endothelial cells and enter the central nervous system(CNS). It could selectively express vascular cell adhesionmolecule (VCAM)-1 and intercellular adhesion molecule(ICAM)-1 which induce the migration of leukocytes crossingBBB.29 As show in Figure 2, C-D, TFF3-Cy5 could not crossHBMEC monolayer by itself; a considerable amount of cRGDL-TFF3-Cy5 could cross HBMEC monolayer mainly resultingfrom the co-migration with THP-1. Compared with cRGDL-TFF3-Cy5, less PL-TFF3-Cy5 was determined in the lowerchamber and no PL-TFF3-Cy5 was observed inside THP-1 cell.It suggested that liposome couldn’t effectively cross BBBwithout combination of cRGD peptide with THP-1. Takentogether, it is the combination of cRGD peptide with THP-1 thatinduced co-migration across BBB model.

In the present study, nude mice were treated with 10U humanrecombinant IL-1β to produce an inflammatory effect10 andassess the effect of brain-targeting liposomes. The cRGDLexhibited rapid brain-targeted delivery within 30 min afteradministration, and this effect reached a peak at 6 h and lasted for12 h (Figure 3, A). The concentration of TFF3 in BLA related todepression30,31 was increased 2.8 times by this targeting effect.

Furthermore, this targeting effect was believed as closelyrelated to the combination characteristics of cRGDL withmonocytes in vivo. Although we did not obtain combinationdata after 6 h (Figure 3, C) because of the necessity to collectexcessive blood, persistently high combination efficiency withmonocytes was observed within 6 h for cRGDL. Much lowercombination efficiency was observed in neutrophils for both PL-TFF3-Cy5 and cRGDL-TFF3-Cy5 compared with monocytes(Figure 3, D). Therefore, we believe that the combination ofcRGDL with monocytes is the main contributor to the effectivebrain targeting observed in the present study. This is inaccordance with the previous study.19 For plain liposomes,PL-TFF3-Cy5 combined with monocytes rapidly but declined to

a very low level after 3 h (Figure 3, C-D). For TFF3, almost nofluorescence was obtained in TFF3-Cy5 group which suggestedthat it was very difficult for TFF3, a macromolecular drug, tocross the BBB under depression.

TFF3 is a member of the trefoil factor family and produced byintestinal goblet cells and a tralmucous cells. It is very resistant todegradation in vivo.5,6 Previous pharmacokinetic studies sug-gested that systemically administered TFF3 is rapidly eliminatedfrom the circulation, with higher stability of TFF3 in thegastrointestinal tract,32 thus demonstrating possible therapeuticuse in gastrointestinal diseases. TFF3 induced Fos-positive cellsin magnocellular oxytocin neurons in the hypothalamus,prompting further research on the role of TFF3 in the centralnervous system and related disorders.7 Our previous studysuggested that systemic intraperitoneal administration of high-dose TFF3 (0.1 mg/kg but not 0.05 mg/kg, i.p.) producedantidepressant-like effects in several rodent models ofdepression.8 Therefore, a high concentration in the brain appearsto be necessary for the practical application of TFF3.

A single injection of cRGDL-TFF3 had an antidepressant-likeeffect that was similar to direct intra-BLA TFF3 infusion, inwhich depressive-like behavior induced by CMS was reversed.In addition, an obvious enhanced concentration of TFF3 in BLAof cRGD-TFF3 group was determined in OB rats. Hence, webelieved that the cRGDL for TFF3 delivery (i.e., a novel brain-targeting drug delivery system) promoted transport through theBBB into brain regions, possibly including the medial BLA,which is a major brain region associated with emotional andpsychiatric disorders.30,31 The increased antidepressant-like ef-fects of cRGDL-TFF3 were associated with the enhanced centraldistribution of TFF3. However, PL-TFF3 did not exert antide-pressant-like effects like direct intra-BLA TFF3 infusion,indicating that the modification of cRGD peptide on liposomesmight be the crucial factor for effective brain-targeting delivery tospecific subregion in the brain. The possible mechanism mayinvolve the high affinity of monocytes for cRGD peptide in vivo.

The present data suggest that TFF3 at a low dose (0.05 mg/kg)produces no antidepressant-like actions by itself but exertsenhanced behavioral responses when loaded into cRGDL.Consistent antidepressant-like effects of cRGDL-TFF3 wereobserved in both the forced swim test and olfactory bulbectomizedrats. Altogether, these data suggest that cRGDLmay be a valuabledrug delivery system for the treatment of depression by allowingBBB transport. Furthermore, Numerous studies have implicatedTFF3 in several neuropsychiatric diseases, including drug abuse,33

anxiety and depression.8 Therefore, the effective dual-braintargeting delivery of the cRGDL-monocyte combination may bea promising strategy for treatment for not only depression but alsoaddiction and anxiety.

In conclusion, we present the cRGDL characterized by anegatively charged membrane and 100-200 nm particle sizewhich exhibiting an excellent dual-brain-targeting effect becauseof the high binding efficiency of the cRGD ligand withmonocytes in rats subjected to animal models of depression.cRGDL-TFF3 produced an enhanced antidepressant response ofTFF3 by increasing the brain distribution of TFF3. Furtherresearch on the safety and efficacy of cRGD coupled withliposomes as an effective brain-targeting drug delivery system

Figure 6. Effect of acute cRGDL-TFF3 administration on depressive-like behavior induced by chronic mild stress in rats. (A) Behavioral test procedure. (B) Bothintra-BLA infusion of TFF3 (1 ng/side) and cRGDL-TFF3 (0.05 mg/kg, i.p.) administration reversed the increase in immobility time induced by CMS, whereasacute PL-TFF3 did not affect immobility. (C) Acute intra-BLA TFF3 and cRGDL-TFF3 (i.p.) administration increased sucrose preference, with no change intotal water intake (D) in the 1 h test during the dark phase in rats subjected to CMS. The data are expressed as mean ± SEM (n = 8 per group). *P b 0.05,**P b 0.01, compared with CMS-saline group; #P b 0.05, ##P b 0.01, compared with saline-control group. CMS, chronic mild stress; FST, forced swim test;SPT, sucrose preference test; TFF3, trefoil factor 3.

399J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

may aid the development of cRGDL-TFF3 for the treatment ofdepression-related psychiatric diseases.

Authors' contributions

QJ, WJX, YX, ZC, LYJ and JH prepared the liposomes andassessed the brain targeting effect. ZRX, ZWL, LYX, and LJLperformed the behavioral tests. QJ, WJX, ZWL, SJ, and LLparticipated in the design of the study and performed thestatistical analysis. QJ and ZWL conceived the study andparticipated in drafting the manuscript. All of the authors readand approved the final manuscript.

Appendix A. Supplementary data

Supplementary data to this article can be found online athttp://dx.doi.org/10.1016/j.nano.2014.09.012.

References

1. Kessler RC, Berglund P, Demler O, Jin R, Koretz D,Merikangas KR, et al.The epidemiology of major depressive disorder: results from the NationalComorbidity Survey Replication (NCS-R). JAMA 2003;289:3095-105.

2. Brasnjevic I, Steinbusch HW, Schmitz C, Martinez-Martinez P. EuropeanNanoBioPharmaceutics Research I. Delivery of peptide and protein drugsover the blood-brain barrier. Prog Neurobiol 2009;87:212-51.

3. Chen Y, Liu L. Modern methods for delivery of drugs across the blood-brain barrier. Adv Drug Deliv Rev 2012;64:640-65.

4. Neuwelt E, Abbott NJ, Abrey L, Banks WA, Blakley B, Davis T, et al.Strategies to advance translational research into brain barriers. LancetNeurol 2008;7:84-96.

5. Thim L, Woldike HF, Nielsen PF, Christensen M, Lynch-Devaney K,Podolsky DK. Characterization of human and rat intestinal trefoil factorproduced in yeast. Biochemistry 1995;34:4757-64.

6. Poulsen SS, Thulesen J, Nexo E, Thim L. Distribution and metabolism ofintravenously administered trefoil factor 2/porcine spasmolytic poly-peptide in the rat. Gut 1998;43:240-7.

7. Derbyshire A, Ludwig M. TFF3 induced Fos protein expression in themagnocellular oxytocin neurons of the hypothalamus. Peptides2004;25:833-8.

8. Shi HS, Zhu WL, Liu JF, Luo YX, Si JJ, Wang SJ, et al. PI3K/Aktsignaling pathway in the basolateral amygdala mediates the rapidantidepressant-like effects of trefoil factor 3. Neuropsychopharmacology2012;37:2671-83.

9. Beumer W, Gibney SM, Drexhage RC, Pont-Lezica L, Doorduin J, KleinHC, et al. The immune theory of psychiatric diseases: a key role for activatedmicroglia and circulating monocytes. J Leukoc Biol 2012;92:959-75.

10. Perry VH, Anthony DC, Bolton SJ, Brown HC. The blood-brain barrierand the inflammatory response. Mol Med Today 1997:335-41.

11. Vass K, Hickey WF, Schmidt RE, Lassmann H. Bone marrow-derivedelements in the peripheral nervous system. An immunohistochemicaland ultrastructural investigation in chimeric rats. Lab Invest1993;69:275-82.

400 J. Qin et al / Nanomedicine: Nanotechnology, Biology, and Medicine 11 (2015) 391–400

12. Saiki I, Koike C, Obata A, Fujii H, Murata J, Kiso M, et al. Functionalrole of sialyl Lewis X and fibronectin-derived RGDS peptide analogueon tumor-cell arrest in lungs followed by extravasation. 1996;15:833-9.

13. Qin J, Chen D, Hu H, Cui Q, Qiao M, Chen B. Surface modification ofRGD-liposomes for selective drug delivery to monocytes/neutrophils inbrain. Chem Pharm Bull 2007;55:1192-7.

14. Yin J, Li Z, Yang T, Wang J, Zhang X, Zhang Q. Cyclic RGDyKconjugation facilitates intracellular drug delivery of polymeric micellesto integrin-overexpressing tumor cells and neovasculature. J DrugTarget 2011;19:25-36.

15. Jonsdottir G, Ingolfsdottir IE, Thormodsson FR, Petersen PH.Endogenous aggregates of amyloidogenic cystatin C variant areremoved by THP-1 cells in vitro and induce differentiation and aproinflammatory response. Neurobiol Aging 2013;34:1389-96.

16. Kuo YC, Lu CH. Effect of human astrocytes on the characteristics ofhuman brain-microvascular endothelial cells in the blood-brain barrier.Colloids Surf B: Biointerfaces 2011;86:225-31.

17. Wong D, Prameya R, Dorovini-Zis K. Adhesion and migration ofpolymorphonuclear leukocytes across humanbrainmicrovessel endothelialcells are differentially regulated by endothelial cell adhesionmolecules andmodulate monolayer permeability. J Neuroimmunol 2007;184:136-48.

18. Sada E, Katoh S, Terashima M, Tsukiyama K. Entrapment of an lon-dependent enzyme into reverse-phase evaporation vesicles. BiotechnolBioeng 1998;32:826-30.

19. Afergan E, Epstein H, Dahan R, Koroukhov N, Rohekar K, Danenberg HD,et al.Delivery of serotonin to the brain bymonocytes following phagocytosisof liposomes. J Control Release 2008;132:84-90.

20. Porsolt RD, Anton G, Blavet N, Jalfre M. Behavioural despair in rats: anew model sensitive to antidepressant treatments. Eur J Pharmacol1978;47:379-91.

21. Xu CM, Wang J, Wu P, Zhu WL, Li QQ, Xue YX, et al. Glycogensynthase kinase 3beta in the nucleus accumbens core mediates cocaine-induced behavioral sensitization. J Neurochem 2009;111:1357-68.

22. ZhuWL, Shi HS, Wang SJ, Xu CM, Jiang WG, Wang X, et al. IncreasedCdk5/p35 activity in the dentate gyrus mediates depressive-likebehaviour in rats. Int J Neuropsychopharmacol 2012;15:795-809.

23. Wang XY, Zhao M, Ghitza UE, Li YQ, Lu L. Stress impairsreconsolidation of drug memory via glucocorticoid receptors in thebasolateral amygdala. J Neurosci 2008;28:5602-10.

24. Willner P, Towell A, Sampson D, Sophokleous S, Muscat R. Reductionof sucrose preference by chronic unpredictable mild stress, and itsrestoration by a tricyclic antidepressant. Psychopharmacology (Berl)1987;93:358-64.

25. Perry VH, Bell MD, Brown HC, Matyszak MK. Inflammation in thenervous system. Curr Opin Neurobiol 1995;5:636-41.

26. Lichtblau N, Schmidt FM, Schumann R, Kirkby KC, Himmerich H.Cytokines as biomarkers in depressive disorder: current standing andprospects. Int Rev Psychiatry 2013;25:592-603.

27. Dantzer R. Cytokine, sickness behavior, and depression. ImmunolAllergy Clin North Am 2009;29:247-64.

28. Qin Z. The use of THP-1 cells as a model for mimicking the function andregulation of monocytes and macrophages in the vasculature. Athero-sclerosis 2012;221:2-11.

29. Stins MF, Gilles F, Kim KS. Selective expression of adhesion molecules onhuman brain microvascular endothelial cells. J Neuroimmunol 1997;76:81-90.

30. Ochsner KN, Gross JJ. The cognitive control of emotion. Trends CognSci 2005;9:242-9.

31. Opal MD, Klenotich SC, Morais M, Bessa J, Winkle J, Doukas D, et al.Serotonin 2C receptor antagonists induce fast-onset antidepressanteffects. Mol Psychiatry 2014;19:1106-14.

32. Kjellev S, Vestergaard EM, Nexo E, Thygesen P, EghojMS, Jeppesen PB,et al. Pharmacokinetics of trefoil peptides and their stability ingastrointestinal contents. Peptides 2007;28:1197-206.

33. Wu P, Shi HS, Luo YX, Zhang RX, Li JL, Shi J, et al. Neuropeptidetrefoil factor 3 attenuates naloxone-precipitated withdrawal in morphine-dependent mice. Psychopharmacology (Berl) 2014;231:4659-68.