ww.sciencedirect.com

a s i a n j o u rn a l o f p h a rma c e u t i c a l s c i e n c e s 8 ( 2 0 1 3 ) 2 2 8e2 3 3

Available online at w

journal homepage: ht tp: / /ees.elsevier .com/ajps/defaul t .asp

Original Research Paper

The distribution and cell uptake of ApoA1 modified lipidcarriers of siRNA in mouse liver in vivo

Yan Li, Mengjie Rui, Hailing Tang, Yuhong Xu*

Shanghai Jiao Tong University, No. 800, Dongchuan Road, Shanghai 200240, China

a r t i c l e i n f o

Article history:

Received 17 June 2013

Received in revised form

13 July 2013

Accepted 20 August 2013

Keywords:

siRNA

Targeting

High density lipoprotein

Immunofluorescence staining

Liposome

* Corresponding author. Tel.: þ86 21 34204739E-mail address: yhxu@sjtu.edu.cn (Y. Xu).

Peer review under responsibility of Shenyan

Production and hosting by El

1818-0876/$ e see front matter ª 2013 Shenyhttp://dx.doi.org/10.1016/j.ajps.2013.09.004

a b s t r a c t

Fluorescence labeled small interfering RNAs (siRNAs) were loaded into lipopolyplexes

modified with ApoA1 (named as rHDL) and administered by intravenous injection. The

biodistribution with time of these lipopolyplexes inside the liver and among various cell

types was followed using tissue sections by Confocal fluorescence microscopy. At about

0.5 h after tail vein injection at a dose of 0.408 mg/kg, very few fluorescence signals were

found in the liver. But then the signals could be seen to accumulate inside hepatocytes as

discrete spots and diffused signals at around 2e4 h after injection. Such a distribution and

uptake pattern was significantly different from what were observed using the commercial

agent Invivofectamine� 2.0 or DOTAP lipoplexes as the carriers. The differences indicated

different mechanisms concerning the in vivo behavior of these carriers. The rHDL carrier

systemwe developed was able to deliver siRNA specifically into hepatocytes while avoiding

the uptake by REM cells especially the Kupffer cells. With it’s low toxicity and off target

effect, it may be suitable to be developed as a hepatocyte targeting delivery system for

siRNA.

ª 2013 Shenyang Pharmaceutical University. Production and hosting by Elsevier B.V. All

rights reserved.

1. Introduction average molecular weight of 13KD. A main obstacle for their

Small interfering RNAs (siRNAs) are 21e23 nucleotide double-

strand RNA segments that could suppress gene expression by

activating RNA-induced silencing complex (RISC) and subse-

quently cleaving targeted mRNA [1,2]. It has been widely re-

ported that siRNAs have significant therapeutically potentials.

However, siRNAs are highly negatively charged with an

, þ86 13501604118 (mobil

g Pharmaceutical Univer

sevier

ang Pharmaceutical Univ

application is that siRNAs could be awfully unstable and

inefficient when entering target cells to exert the therapeutic

effects. Therefore there is an urgent need for siRNA delivery

systems to protect and direct siRNA molecules into the

diseased site and targeted cells [3e6].

There have a lot of efforts made in pursuing of safe and

efficient siRNA delivery systems. Delivery systems based on

e); fax: þ86 21 34204738.

sity

ersity. Production and hosting by Elsevier B.V. All rights reserved.

a s i a n j o u rn a l o f p h a rm a c e u t i c a l s c i e n c e s 8 ( 2 0 1 3 ) 2 2 8e2 3 3 229

synthetic materials such as lipids or polymers are considered

more versatile and practical, but the efficiencies had been in

general unsatisfactory. Recently there have been some pro-

gresses made in screening novel lipid structures for siRNA

delivery to the liver [7e10]. But there have been very limited

study investigating the detailed liver uptake mechanism

involved. Many studies had reported that after systemic

administration, most siRNA carriers could deposit massively

into the liver and sometimes the spleen [11e13]. But the liver

is a big and vital organ consisted of many cell types, including

parenchymal and non-parenchymal cells. In order to address

diseases involving the liver parenchymal, it would be impor-

tant to study the distribution pattern of different siRNA car-

riers among different cell types, and to optimize targeted

delivery to hepatocytes while avoiding the mononuclear

phagocyte cells.

In this study, we used an ApoA1 modified lipopolyplex

system as a model system to study the detailed distribution

and uptake patterns of fluorescence labeled siRNA in vivo

after iv administration. It’s prepared by complexing siRNA

with short PEI and then coating with EPC/Chol/ApoA1. It was

named rHDL-siRNA delivery system because the particle

morphology and surface properties resembled an HDL parti-

cle. We have shown in a previous study that such a carrier

system is almost non-toxic and can interact specifically with

hepatocytes and deliver siRNA into the cells [14]. In this study

we are very interested to examine its biodistribution after iv

injection and interaction with liver tissue and hepatocytes

in vivo. In addition, we alsowould like to compare such a rHDL

system with other delivery systems made from the classic

cationic lipid DOTAP and the commercial reagent

Invivofectamine�2.0.

2. Materials and methods

2.1. Materials

Branched polyethylenimine (PEI, MW 1800) and sodium

cholate were purchased from Alfa Aesar (MA, USA). Egg

phosphatidylcholine (EPC) and cholesterol were purchased

from NOF (Tokyo, Japan). DOTAP was purchased from Avanti

Polar Lipids (Alabama, USA). Recombinant human apolipo-

proteinA-Iwas expressed in Escherichia coli andpurified byHis-

Trapnickel affinity chromatographyas describedbyRyan et al.

[15]. Other materials were purchased from SigmaeAldrich

(St Louis, MO, USA).

Both fluorescence labeled and unlabeled siRNAs were

synthesized by Genepharm Co. (China). The siRNA targeting

the luciferase gene contained the anti-sense strand

50eUCGAAGUACUCAGCGUAAGdTdT-3’and sense strand 50-CUUACGCUGAGUACUUCGAdTdT-30. Cy5 dye was labeled at

the 50-end of siLuc siRNA.

2.2. Preparation of rHDLs-siRNA

rHDLs-siRNA was prepared as described by Rui et al. [14]

Briefly, 60 mg of siRNA was dissolved in 214 ml of RNase-free

water. 16 ml of Branched PEI (1800 Da) (10mg/ml)was dissolved

in 284 ml RNase-free water. Then the siRNA solution was

added to PEI solutions resulting in PEI/siRNA complexes. Li-

posomes with the lipid composition of EPC:cholesterol (20:1,

molar ratio) at a total lipid concentration of approximately

100 mM, were prepared using the thin film hydration method.

Then 130 ml of cholate solution (200 mg/ml in Tris buffer) and

1500 ml of apoA-I protein solution (7 mg/ml) were added to

liposome solutions to form mixed micelles containing lipid,

cholate and apoA1. The PEI/siRNA complex solution 600 ml was

then vortex-mixed with 1500 ml of the mixed micelles for

5 min. The final volume was adjusted to 6000 ml in 10 mM Tris

buffer. The mixture was incubated overnight at 4 �C, and then

dialyzed against 2 L of Tris buffer, changing 3 times over 2

days.

2.3. Preparation of Invivofectamine� 2.0-siRNA

The Invivofectamine�-siRNA complexes were prepared based

on the protocol provided by invitrogen. Briefly, siRNA solution

was diluted using the Complexation Buffer to 1.5 mg/ml and

added directly to the Invivofectamine� 2.0 reagent and vor-

texed immediately for 2e3 s. The mixture was incubated for

30min at 50 �C and then loaded into the Float-A-Lyzer� device

and dialyzed for 2 h in 1 L of PBS, pH 7.4, with gentle agitation.

2.4. Preparation of DOTAP-siRNA lipoplexes

DOTAP chloroform solution was dried into a lipid film and re-

suspended in 1 ml PBS. The DOTAP liposome solution was

mixedwith Cy5-siRNA at the N/P ratio of 8:1 and incubated for

half an hour before injection.

2.5. Physico-chemical characterization of the carriers

The siRNA loading efficiencies were evaluated by gel electro-

phoresis. The mean particles size and zeta potential of the

three carrier particles were measured by dynamic light scat-

tering using Zetasizer Nano-ZS90 (Malvern, UK).

2.6. Animal studies

The animal study protocol was approved by the institute’s

animal care and use committee. Female ICR mice (20e30 g)

were obtained from Shanghai B&K laboratory animal Corp.

Ltd. One group received the rHDLs-siRNA formulation at the

dose of 0.408 mg/kg by iv injection. They were sacrificed at

different time points (0.5, 2, 4, 24 h after the injection) for the

collection of liver for the immunohistochemistry study. The

Invivofectamine�2.0-siRNA group received the iv injection at

0.547 mg/kg (body weight) dose. They were sacrificed at 0.5, 2,

4, 10 h for the collection of liver. The DOTAP-siRNA group

received the iv injection at 0.338mg/kg (bodyweight) dose and

they were sacrificed at 0.5, 2, 10 h for the collection of liver. In

addition, we also had a control group of mice received PBS

injections via the tail vein and liver were collected at 2 h post

dosing.

Before the collection of liver tissues, mice (n ¼ 3) were

deeply anesthetized with Chloral hydrate (40 mg/kg i.p.), and

perfused through the heart at a rate of 5ml/min using 20ml of

0.9% saline followed by 20 ml 4% paraformaldehyde. The liver

samples were then taken out and divided into four different

Fig. 1 e The siRNA loading efficiency of rHDLs-siRNA,

Invivofectamine2.0-siRNA and DOTAP-siRNA was

analyzed by gel electrophoresis. All samples were run on

1% agarose gel and stained with GelRed dye.

a s i a n j o u rn a l o f p h a rma c e u t i c a l s c i e n c e s 8 ( 2 0 1 3 ) 2 2 8e2 3 3230

lobes (caudate, left lateral, right lateral, and medial lobes). We

usually use tissues from the left lateral lobe in our standard

procedure.

2.7. Immunofluorescence study

The mouse liver tissue were sectioned and stained using

various antibodies to reveal the histology features and locate

the fluorescence labeled siRNA. The antibodies used included

anti-CD68 from AbD SeroTec (Raleigh, NC), anti-collagen IV

fromMillipore (Billerica,MA). In addition, the fluorescence dye

DAPI was used to stain the nuclei, and Phalloidin (Invitrogen,

Carlsbad, CA) was used to outline hepatocyte cell membrane

on the tissue sections. Specifically, 10 mm-thick frozen tissue

sections were cut using Cryostat (Leica CM1850), mounted on

Charged Slides (Premiere 9308W), and stored in a �80 �Cfreezer. Before staining, slides were allowed to air dry at room

temperature for 30 min. Tissue sections were washed with

1xPBS (PH 7.4, 3� 5min), and incubated overnight at 4 �C in 5%

serum from the species in which the secondary antibody was

generated. Afterward, the sections were stained with the

primary antibodies at the concentration of 10 mg/ml for 16 h at

Table 1 e Characterization of rHDLs-siRNA-Cy5, Invivofectami

Test preparations Sizea (nm) Po

rHDLs-siRNA-Cy5 142.3 � 1.852

Invivo2.0-siRNA- Cy5 100.8 � 3.304

DOTAP-siRNA-Cy5 194.07 � 5.22

a Samples were diluted 1:50 with Tris buffer (pH8.0) or PBS buffer (PH7.4b The polydispersity index of rHDLs-siRNA, Invivofectamine2.0-siRNA an

deviation of an underlying Gaussian size distribution.c Samples were diluted 1:50 with Tris buffer (pH8.0), PBS buffer (PH7.4) f

4 �C washed again in PBS, and labeled with secondary anti-

bodies (Donkey anti-rat or rabbit IgG) conjugated with Alexa

488 (Invitrogen, Carlsbad, CA). Finally, all sections were

stained with DAPI to locate the nuclei, andmounted using the

aqueous mounting medium Prolong (Vector). The slides were

scanned using Confocal Microscope and images were

captured under three channels: DAPI for nuclei, Alex488 for

phalloidin-stained cell membrane or CD68, and Cy5 for siRNA.

3. Results and discussion

3.1. Preparation and characterization of rHDLs-siRNA,Invivofectamine�2.0-siRNA and DOTAP-siRNA complexes

rHDLs-siRNA, Invivofectamine�2.0-siRNA and DOTAP-siRNA

complexes were prepared as described. They were formu-

lated based on the optimization studies reported previously

[14,16,17] or as recommended by the supplier. The siRNA

encapsulation/complexation ratios in three formulationswere

all close to 100%because free siRNAbandwas no longer visible

in any of the formulations, as shown in Fig. 1. The size distri-

butions and Zeta potentials of the three types of complexes

were listed in Table 1. They all form small particles withmean

diameter at about 100 nm. The surface potentials were quite

different. The rHDLs-siRNA and Invivofectamine�2.0-siRNA

particles were both slightly negative charged. The DOAP-

siRNA particles were highly positive charged.

3.2. Histological and immunohistochemical analysis ofthe liver tissues

In order to study the detailed intracellular distribution pat-

terns of the siRNA carrier, we first attempted to characterize

the major cell types using histology and immunohistochem-

ical methods. As shown in Fig. 2, mouse liver stained with a

standard hematoxylin and eosin (H&E) procedure reveal pat-

terns of cellular organization similar to what has been re-

ported [18e21].Majority of the cells can be identified as

hepatocytes, sinusoids (which contain endothelial linings),

Kupffer Cells [18e21].We applied immunofluorescent staining

of collagen IV to outline the sinusoidal endothelial cells.

Kupffer cells were identified using CD68 staining. Phalloidin

was used to outline cell membranes. It’s not specific to any

specific type, but hepatocytes can be easily identified

morphologically with features such as large single or double

nuclei with approximately 7 mm in diameter.

ne2.0-siRNA-Cy5 and DOTAP-siRNA-Cy5.

lydispersity indexb (PdI) Zeta potentialc (mV)

0.200 � 0.017 �4.99 � 1.13

0.289 � 0.031 �5.70 � 1.20

0.267 � 0.006 18.57 � 2.17

) for DLS measurement.

d DOTAP-siRNA correspond to the square of the normalized standard

or zeta potential measurement.

Fig. 2 e H&E staining and immunofluorescence staining Photomicrographs of mouse liver tissue. Bright field images

showing central venule (A, B) and Kupffer cells (C, D). (E) Nuclei outlined with DAPI. (F) Phalloidin outlined liver cells (green)

(G) CD68-stained Kupffer cells. (H) Overlay images for phalloidin outlined liver cells (green), CD68-stained Kupffer cells and

nuclei (DAPI, blue).

a s i a n j o u rn a l o f p h a rm a c e u t i c a l s c i e n c e s 8 ( 2 0 1 3 ) 2 2 8e2 3 3 231

3.3. Distribution of cy5 labeled rHDL-siRNA particles inthe liver and their uptake by hepatocytes

The intracellular distribution and uptakemechanismof rHDLs

in the liver was investigated by confocal microscopy. Cy5

labeled rHDL-siRNAwas injected via tail vein of ICRmice at the

dose of 0.408 mg/kg body weight. Tissue sections from livers

collectedat different timepoints after injectionwere analyzed.

As shown in Fig. 3, the amount of visible red cy5 signals were

quite rare at 0.5 h after injection (Figs. 2E and 3A) but they

increased to a significant number after 2 h (Figs. 2F and 3B).

Interestingly, these positive signals are clearly all inside

Fig. 3 e rHDLs-siRNA-Cy5 biodistribution in mouse liver at 0.5 h

and counterstained with DAPI (blue). (AeH) Overlaid images of C

(DAPI blue) for the time points. Cy5esiRNA distributed in the liv

24 h, and (H) PBS control.

hepatocytes, indicating a specific uptake mechanism by he-

patocytes. The signals became diffused and eventually died

downafter 4 h and 24h (Figs. 2D and 3C, G). These observations

agreed very well with our earlier report describing the hepa-

tocyte uptake of rHDLs in vitro. The ApoA1 and surface prop-

erties resembling the natural HDL particles enabled a specific

uptake mechanism of the rHDL-siRNA particles by hepato-

cytes. They were taken up into closed compartments (endo-

some or early lysosome) and then released into the cytoplasm

probably based on the PEI photon sponge effect. However,

compared with the in vitro data [14], the amount of fluores-

cence signals inside hepatocytesweremuchweaker. Thismay

, 2 h, 4 h, 24 h Sections were stained with phalloidin (green)

y5 (red) and phalloidin outlined liver cells (green) and nuclei

er is time dependent. (A, E) 30 min, (B, F) 2 h, (C, G) 4 h, (D)

Fig. 4 e Invivofectamine2.0-siRNAbiodistribution in mouse liver 0.5, 2 h, 4 h, 10 h post injection Sections were stained with

phalloidin (green) and counterstained with DAPI (blue). (AeH) Overlaid images of Cy5 (red) and phalloidin outlined liver cells

(green) and nuclei (DAPI blue) for the time points. (A, E) 30 min, (B, F) 2 h, (C, G) 4 h, (D) 10 h, and (H) PBS control.

a s i a n j o u rn a l o f p h a rma c e u t i c a l s c i e n c e s 8 ( 2 0 1 3 ) 2 2 8e2 3 3232

be due to the dilution of rHDL particle concentration in the

blood so that there were only limited numbers available in the

liver. Therefore the siRNA efficiency (activity per dose) was

much lower in vivo than what we could obtain in vitro. But a

clearadvantageof the rHDL-siRNAformulation is itsminimum

toxicity [14], so it is possible to increase the dose to improve

efficacy. In this studywealso showed that rHDLswere takenup

only by hepatocytes with almost none in Kupffer cells and

likes. It’s also important because there had always been a

concern for off target effect of siRNAmediatedbyKupffer cells.

For comparison, we also examined the distribution and

uptake patterns of the Invivofectamine�2.0-siRNA and

Fig. 5 e DOTAP-siRNA biodistribution in mouse liver Sections w

DAPI (blue). (AeF) Overlaid images of Cy5 (red) and phalloidin ou

points. (A, D) 30 min, (B, E) 2 h, (C) 10 h, and (F) PBS control.

DOTAP/siRNA complexes. Their time course pictures were

shown in Figs. 4 and 5 respectively. The cy5 fluorescence

distribution patterns were completely different. The

Invivofectamine�2.0-siRNA particles showed up quickly in the

liver but were mostly outside of the cells binding to the cell

membrane within half hour of the injection (Figs. 4A and E).

They were then taken up into intracellular compartments at

around 2e4 h after injection (Fig. 4B, C, F and G) and gradually

diffused out after 10 h (Fig. 4D). The DOTAP-siRNA particles,

however, were seen only in sinusoids and colocalized with

CD68 labeled Kupffer cells (Fig. 5AeE). These observations

agreed with most of the literature describing the distribution

ere stained with phalloidin (green) and counterstained with

tlined liver cells (green) and nuclei (DAPI blue) for the time

a s i a n j o u rn a l o f p h a rm a c e u t i c a l s c i e n c e s 8 ( 2 0 1 3 ) 2 2 8e2 3 3 233

and uptake mechanisms of these two types of carriers [22,23].

It was proposed that Invivofectamine�2.0-siRNA particles

would interact with LDLs during circulation and bind to he-

patocytes mediated by the LDL receptor [23,24]. On the other

hand, DOTAP-siRNA complexes, being positively charged,

would interact with the opsonins and then be taken up by

Kupffer cells. They represent two distinctive pathways that

were both different from the rHDL-siRNA uptake mechanism

we described in this study.

The most logical and feasible way to introduce siRNA car-

riers into the liver cells is through intravenous injection. The

injected particles enter the liver from portal vein to sinusoids

and exit through the central veins. Particles with sizes of

about 150 nm could diffuse out of the liver sinusoidal endo-

thelial lining via the fenestrae and come into direct contact

with hepatocytes [25e27]. Specific mechanisms are required

for the particles to bind to hepatocytes and initiate the

endocytosis process.

4. Conclusion

In summary, we’ve shown that rHDL-siRNA carriers could be

used as a prospective delivery system for safe and efficient

delivery of siRNA to hepatocytes in vivo. Because the common

obstacle for current siRNA delivery strategies is the accumu-

lation in organs of RES systems, the unique properties of the

rHDL nanoparticle make it a potential vector for hepatocyte-

specific delivery of siRNA.

Acknowledgments

This study was supported by grants from the Natural Science

Foundation of China No. 30825045.

r e f e r e n c e s

[1] Jayaraman M, Ansell SM, Mui BL, et al. Maximizing thepotency of siRNA lipid nanoparticles for hepatic genesilencing in vivo. Angew Chem Int Edition2012;51(34):8529e8533.

[2] Hannon GJ, Rossi JJ. Unlocking the potential of the humangenome with RNA interference. Nature2004;431(7006):371e378.

[3] Bumcrot D, Manoharan M, Koteliansky V, et al. RNAitherapeutics: a potential new class of pharmaceutical drugs.Nat Chem Biol 2006;2(12):711e719.

[4] Whitehead KA, Langer R, Anderson DG. Knocking downbarriers: advances in siRNA delivery. Nat Rev Drug Discov2009;8(2):129e138.

[5] de Fougerolles A, Vornlocher HP, Maraganore J, et al.Interfering with disease: a progress report on siRNA-basedtherapeutics. Nat Rev Drug Discov 2007;6(6):443e453.

[6] Zimmermann TS, Lee ACH, Akinc A, et al. RNAi-mediatedgene silencing in non-human primates. Nature2006;441(7089):111e114.

[7] Kim DH, Rossi JJ. Strategies for silencing human diseaseusing RNA interference. Nat Rev Genet 2007;8(3):173e184.

[8] Semple SC, Akinc A, Chen J, et al. Rational design ofcationic lipids for siRNA delivery. Nat Biotechnol2010;28(2):172e176.

[9] Walsh CL, Nguyen J, Tiffany MR, et al. Synthesis,characterization, and evaluation of ionizable lysine-basedlipids for siRNA delivery. Bioconjug Chem 2012;24(1):36e43.

[10] Zhang J, Fan H, Levorse DA, et al. Interaction of cholesterol-conjugated ionizable amino lipids with biomembranes: lipidpolymorphism, structureeactivity relationship, andimplications for siRNA delivery. Langmuir2011;27(15):9473e9483.

[11] Nishiyama N, Kataoka K. Current state, achievements, andfuture prospects of polymeric micelles as nanocarriers fordrug and gene delivery. Pharmacol Ther2006;112(3):630e648.

[12] Kim TH, Jiang HL, Jere D, et al. Chemical modification ofchitosan as a gene carrier in vitro and in vivo. Prog Polym Sci2007;32(7):726e753.

[13] Thomas M, Lu JJ, Chen J, et al. Non-viral siRNA delivery to thelung. Adv Drug Deliv Rev 2007;59(2):124e133.

[14] Rui M, Tang H, Li Y, et al. Recombinant high densitylipoprotein nanoparticles for target-specific delivery ofsiRNA. Pharm Res 2012:1e12.

[15] Ryan RO, Forte TM, Oda MN. Optimized bacterial expressionof human apolipoprotein AI. Protein Expr Purif2003;27(1):98e103.

[16] Xu Y, Szoka FC. Mechanism of DNA release from cationicliposome/DNA complexes used in cell transfection.Biochemistry 1996;35(18):5616e5623.

[17] Xu Y, Hui SW, Frederik P, et al. Physicochemicalcharacterization and purification of cationic lipoplexes.Biophys J 1999;77(1):341e353.

[18] Baratta JL, Ngo A, Lopez B, et al. Cellular organization ofnormal mouse liver: a histological, quantitativeimmunocytochemical, and fine structural analysis.Histochem Cell Biol 2009;131(6):713e726.

[19] Malarkey DE, Johnson K, Ryan L, et al. New insights intofunctional aspects of liver morphology. Toxicol Pathol2005;33(1):27e34.

[20] Nagayama S, Ogawara K, Fukuoka Y, et al. Time-dependentchanges in opsonin amount associated on nanoparticlesalter their hepatic uptake characteristics. Int J Pharm2007;342(1):215e221.

[21] Shi B, Keough E, Matter A, et al. Biodistribution of smallinterfering RNA at the organ and cellular levels after lipidnanoparticle-mediated delivery. J Histochem Cytochem2011;59(8):727e740.

[22] Garbuzenko OB, Saad M, Betigeri S, et al. Intratracheal versusintravenous liposomal delivery of siRNA, antisenseoligonucleotides and anticancer drug. Pharm Res2009;26(2):382e394.

[23] Akinc A, Querbes W, De S, et al. Targeted delivery of RNAitherapeutics with endogenous and exogenous ligand-basedmechanisms. Mol Ther 2010;18(7):1357e1364.

[24] Disterer P, Al-Shawi R, Ellmerich S, et al. Exon skipping ofhepatic APOB pre-mRNA with splice-switchingoligonucleotides reduces LDL cholesterol in vivo. Mol Ther2013;21(3):602e609.

[25] Le Couteur DG, Warren A, Cogger VC, et al. Old age and thehepatic sinusoid. Anat Rec 2008;291(6):672e683.

[26] Snoeys J, Lievens J, Wisse E, et al. Species differences intransgene DNA uptake in hepatocytes after adenoviraltransfer correlate with the size of endothelial fenestrae.Gene Ther 2007;14(7):604e612.

[27] Braet F, Wisse E. Structural and functional aspects of liversinusoidal endothelial cell fenestrae: a review. Comp Hepatol2002;1(1):1.