Efficient suppression of murine intracellular adhesion molecule-1 using ultrasound-responsive and mannose-modified lipoplexes inhibits acute hepatic inflammation

Authors

  • Keita Un,

    1. Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
    2. The Japan Society for the Promotion of Science, Tokyo, Japan
    Search for more papers by this author
  • Shigeru Kawakami,

    Corresponding author
    1. Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
    • Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan===

    Search for more papers by this author
  • Mitsuru Yoshida,

    1. Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Yuriko Higuchi,

    1. Institute for Innovative NanoBio Drug Discovery and Development, Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Ryo Suzuki,

    1. Department of Biopharmaceutics, School of Pharmaceutical Sciences, Teikyo University, Kanagawa, Japan
    Search for more papers by this author
  • Kazuo Maruyama,

    1. Department of Biopharmaceutics, School of Pharmaceutical Sciences, Teikyo University, Kanagawa, Japan
    Search for more papers by this author
  • Fumiyoshi Yamashita,

    1. Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
    Search for more papers by this author
  • Mitsuru Hashida

    Corresponding author
    1. Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, Kyoto, Japan
    2. Institute for Integrated Cell-Material Sciences, Kyoto University, Kyoto, Japan
    • Department of Drug Delivery Research, Graduate School of Pharmaceutical Sciences, Kyoto University, 46-29 Yoshida-shimoadachi-cho, Sakyo-ku, Kyoto 606-8501, Japan===

    Search for more papers by this author
    • fax: (81)-75-753-4575.


  • Potential conflict of interest: Nothing to report.

Abstract

Hepatitis is often associated with the overexpression of various adhesion molecules. In particular, intracellular adhesion molecule-1 (ICAM-1), which is expressed on hepatic endothelial cells (HECs) in the early stage of inflammation, is involved in serious illnesses. Therefore, ICAM-1 suppression in HECs enables the suppression of inflammatory responses. Here, we developed an ICAM-1 small interfering RNA (siRNA) transfer method using ultrasound (US)-responsive and mannose-modified liposome/ICAM-1 siRNA complexes (Man-PEG2000 bubble lipoplexes [Man-PEG2000 BLs]), and achieved efficient HEC-selective ICAM-1 siRNA delivery in combination with US exposure. Moreover, the sufficient ICAM-1 suppression effects were obtained via this ICAM-1 siRNA transfer in vitro and in vivo, and potent anti-inflammatory effects were observed in various types of inflammation, such as lipopolysaccharide, dimethylnitrosamine, carbon tetrachloride, and ischemia/reperfusion-induced inflammatory mouse models. Conclusion: HEC-selective and efficient ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure enables suppression of various types of acute hepatic inflammation. This novel siRNA delivery method may offer a valuable system for medical treatment where the targeted cells are HECs. (HEPATOLOGY 2012;56:259–269)

Hepatitis resulting from conditions such as drug-induced hepatic inflammation and ischemia/reperfusion (IR)-induced liver injury followed by surgery is a major obstacle for medical treatment.1, 2 Moreover, it was reported that chronic hepatitis progresses to cirrhosis and liver cancer3, 4; therefore, the prevention and early treatment of hepatitis are important for patients and medical professionals. Most drug-induced hepatitis is caused by nuclear factor-κB activation and proinflammatory cytokine production followed by various stimulations in medical treatments.5 In IR-induced liver injury, a large amount of reactive oxygen species produced by IR stimulation is involved in the induction of inflammatory responses.6 Although the mechanism for each inflammatory response is different, various adhesion molecules, such as vascular cell adhesion molecule (VCAM) and intracellular adhesion molecule (ICAM), are abundantly expressed on hepatic endothelial cells (HECs) in the early stage of inflammatory responses followed by various types of stimulation.7 Among these, ICAM-1 is known as a major molecule that is highly involved in the adhesion, diapedesis, and tissue infiltration of leukocytes contributing to the deterioration in inflammatory responses.8 During alcohol-induced liver injury, it was reported that ICAM-1 expression and the resultant leukocyte infiltration are involved in the deterioration of alcohol-induced liver injury.9 Therefore, the suppression of inflammatory responses may be achieved by selective knockdown of ICAM-1 in HECs.

RNA interference is an important endogenous mechanism for gene regulation by cleaving specific messenger RNA (mRNA) possessing the complementary sequence using small interfering RNA (siRNA).10, 11 Although siRNA is a promising candidate for molecular therapy, an effective method for siRNA transfer into the cytoplasm of targeted cells in vivo is still being developed. The effective methods for in vivo siRNA delivery involve nonviral carriers, including liposomes, emulsions, micelles, and polymers.12-18 However, because the nonviral carriers are taken up into the cells via endocytosis, degradation within endosomes and escape from endosomes are major obstacles for the improvement of siRNA therapeutics. Moreover, efficient and selective siRNA delivery into HECs is essential to achieve the potent anti-inflammatory effects produced by ICAM-1 siRNA.

Recently, the benefits have become appreciated of delivery of nucleic acids into cells using microbubbles and ultrasound (US) (also known as “sonoporation methods”), due to the high transfer efficiency into the cytoplasm.19-22 Our group has developed US-responsive and mannose-modified liposomes/plasmid DNA complexes for in vivo gene transfer and successfully obtained efficient gene expression in mannose receptor-expressing cells, such as HECs and splenic dendritic cells.23-25 Moreover, we demonstrated that a large amount of plasmid DNA could be directly transferred into the cytoplasm through a mechanism involving transient pores created on the cell membrane by the destruction of microbubbles after US exposure.26 Therefore, the efficient transfer of ICAM-1 siRNA into HECs might be achieved by applying this method to siRNA delivery.

In the present study, we developed an ICAM-1 siRNA transfer system based on US-responsive and mannose-modified liposome/siRNA complexes (Man-PEG2000 bubble lipoplexes [Man-PEG2000 BLs]) for anti-inflammatory therapy. ICAM-1 siRNA delivered by Man-PEG2000 BLs and US exposure was selectively and efficiently transferred into HECs in vitro and in vivo. Furthermore, sufficient ICAM-1 suppression and potent anti-inflammatory effects were achieved by ICAM-1 siRNA transfer against various types of inflammation induced by lipopolysaccharide (LPS), dimethylnitrosamine (DMN), carbon tetrachloride (CCl4), and IR. To our knowledge, this is the first report of a gene transfer method using Man-PEG2000 BLs and US exposure for the selective and efficient transfer of siRNA to HECs. This novel siRNA transfer method could be valuable for medical treatments that target HECs.

Abbreviations

ALT, alanine aminotransferase; AST, aspartate aminotransferase; BL, bubble lipoplex; CCl4, carbon tetrachloride; DAPI, 4′,6-diamidino-2-phenylindole; DMN, dimethylnitrosamine; FITC, fluorescein isothiocyanate; H&E, hematoxylin and eosin; HEC, hepatic endothelial cell; ICAM, intracellular adhesion molecule; IFN-γ, interferon-γ; IL, interleukin; IR, ischemia/reperfusion; iv, intravenous; LPS, lipopolysaccharide; MCP-1, monocyte chemoattractant protein 1; MDA-5, melanoma differentiation-associated gene 5; mRNA, messenger RNA; RIG-1, retinoic acid-inducible gene 1; siRNA, small interfering ribonucleic acid; TLR, Toll-like receptor; TNF-α, tumor necrosis factor α; US, ultrasound.

Materials and Methods

In vitro siRNA Delivery

After incubation of HECs for 72 hours, the culture medium was replaced with Opti-MEM I (Invitrogen, Carlsbad, CA) containing lipoplexes/BLs (1 μg siRNA). At 5 minutes after siRNA transfer, HECs were exposed to US (frequency, 2.062 MHz; duty, 50%; burst rate, 10 Hz; intensity, 4.0 W/cm2) for 20 seconds. In the siRNA delivery using naked siRNA and conventional nanobubbles, at 5 minutes after addition of naked siRNA (1 μg) and conventional nanobubbles (60 μg total lipids), cells were immediately exposed to US. US was generated using a Sonopore-4000 sonicator (Nepa Gene, Chiba, Japan). At 1 hour after US exposure, the medium was replaced with RPMI-1640 and incubated for an additional 23 hours. Lipofectamine 2000 (Invitrogen) was used according to the recommended procedures with an exposure time of 1 hour, which is the same exposure time in other experiments using lipoplexes.

In Vivo siRNA Delivery

Six-week-old C57BL/6 female mice were intravenously injected with BLs containing 10 μg siRNA via the tail vein. At 5 minutes after the injection of the bubble lipoplexes, US (frequency, 1.045 MHz; duty, 50%; burst rate, 10 Hz; intensity 1.0 W/cm2; time, 2 minutes) was applied transdermally to the abdominal area using a Sonopore-4000 sonicator. In the siRNA delivery using naked siRNA and conventional nanobubbles, at 4 minutes after intravenous injection of conventional nanobubbles (500 μg total lipid), naked siRNA (10 μg) was intravenously injected and US was exposed at 1 minute after naked siRNA injection.

Statistical Analyses

Results are presented as the mean ± SD of more than three experiments. Analysis of variance was used to test the statistical significance of differences among groups. Two-group comparisons were performed using the Student t test and multiple comparisons between control and other groups were performed using the Dunnett's test.

Results

Suppression Effects of ICAM-1 siRNA

The suppression of LPS-induced ICAM-1 expression by ICAM-1 siRNAs (Supporting Fig. 1A) was investigated in primary mouse HECs. As shown in Supporting Fig. 1B, the suppression of ICAM-1 was the highest in ICAM-1 siRNA with sequence 1, and not observed in scrambled siRNA. Therefore, ICAM-1 siRNA containing sequence 1 and scrambled siRNA were used in the following examinations.

Figure 1.

Suppression effects of icam-1 mRNA expression and cytotoxicity followed by ICAM-1 siRNA delivery in LPS-stimulated primary mouse HECs. (A) In vitro confocal images of cellular associated ICAM-1 siRNA (1 μg siRNA) transferred by various methods 1 hour after treatment in primary mouse HECs. US was directly exposed to HECs at 5 minutes after addition of BLs. The lipoplexes were constructed with AlexaFluor-594–labeled ICAM-1 siRNA (red), and the endosomes were labeled with AlexaFluor-488 transferrin conjugates (green). Nuclei were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (blue). Scale bars, 10 μm. (B,C) The level of icam-1 mRNA expression (B) and in vitro confocal images of ICAM-1 expression (C) obtained by ICAM-1 siRNA transfer (1 μg siRNA) using various types of methods 24 hours after LPS stimulation in primary mouse HECs. US was directly exposed to HECs at 5 minutes after addition of BLs, and cells were exposed to LPS (100 ng/mL) at 24 hours after the addition of siRNA or lipoplexes/BLs. ICAM-1 was labeled with anti-mouse ICAM-1 antibody and fluorescein isothiocyanate (FITC)-conjugated secondary antibody (green), and nuclei were counterstained by DAPI (blue). Scale bars, 10 μm. (D,E) Comparison of the suppression of icam-1 mRNA expression (D) and cell viability (E) obtained by siRNA transfer using Man-PEG2000 BLs (1 μg siRNA) and US exposure with that by Lipofectamine 2000. *P < 0.05, **P < 0.01 versus no treatment. Each value represents the mean + SD (n = 5). N.T., no treatment.

Physicochemical Properties of Man-PEG2000 BLs

Following enclosure of US imaging gas into Man-PEG2000 BLs, lipoplexes became cloudy (data not shown) and the average particle size increased (Supporting Fig. 2A). Following gel electrophoresis experiments, the formation of siRNA complexes in BLs was confirmed (Supporting Fig. 2B). Moreover, ζ-potentials of BLs were lower than that of liposomes (Supporting Fig. 2A), suggesting that siRNA was attached to the surface of cationic bubble liposomes. These physicochemical properties are consistent with our previous reports using plasmid DNA.23-26

Figure 2.

In vivo distribution of ICAM-1 siRNA delivered by Man-PEG2000 BLs and US exposure. (A) Tissue distribution and pharmacokinetics of radiolabeled bare- and Man-PEG2000 BLs complexed with 10 μg ICAM-1 siRNA after intravenous (iv) administration into mice. Tissue distribution of lipoplexes was measured at 6 hours after iv administration of lipoplexes. Inset shows blood concentration of lipoplexes at predetermined times after iv administration. *P < 0.05, **P < 0.01 versus the corresponding group of bare-PEG2000 lipoplexes. Each value represents the mean ± SD (n = 5). (B) Hepatic cellular localization of AlexaFluor-594 labeled ICAM-1 siRNA delivered by bare- and Man-PEG2000 BLs (10 μg siRNA) and US exposure at 6 hours after iv administration of lipoplexes into mice. Liver was separated to hepatocytes, Kupffer cells, and endothelial cells by collagenase perfusion, one-step density gradient centrifugation, and magnetic cell sorting as described in the Supporting Materials and Methods. **P < 0.01 versus the corresponding group of hepatocytes. Each value represents the mean + SD (n = 5). (C) Fluorescent images of hepatic localization of AlexaFluor-594–labeled ICAM-1 siRNA (red) delivered by bare- and Man-PEG2000 BLs (10 μg siRNA) and US exposure. HECs were labeled with anti-mouse CD146 antibody and FITC-conjugated secondary antibody (green), and nuclei were counterstained with DAPI (blue). Livers were harvested at 6 hours after iv administration of lipoplexes into mice, and magnified images corresponding to the areas enclosed in boxes are shown in the inset (i). Scale bars, 100 μm.

Intracellular Transport Characteristics of ICAM-1 siRNA

The siRNA transfer efficiency was investigated in primary mouse HECs expressing mannose receptors (Supporting Fig. 4). The amount of siRNA delivered by BLs and US exposure was significantly higher than that by lipoplexes only (Supporting Fig. 3A). Moreover, the amount of siRNA delivered by Man-PEG2000 BLs and US exposure was higher than unmodified BLs. However, the amount of siRNA was significantly suppressed in the presence of mannan but not suppressed in the presence of chlorpromazine, an endocytosis inhibitor (Supporting Fig. 3B,C). Confocal microscopy analysis of cells after siRNA transfer by bubble lipoplexes with US exposure revealed that siRNA was not colocalized in endosomes (Fig. 1A). These observations suggest that siRNA is directly transferred into the cytoplasm of targeted cells and is not mediated by endocytosis in this siRNA transfer method.

Figure 3.

Suppression effects of ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure on icam-1 mRNA and protein expression in HECs of an LPS/D-galactosamine-induced inflammatory mouse model. (A) Evaluation schedule of ICAM-1 expression in LPS/D-galactosamine-stimulated mice. (B-D) The expression level of icam-1 mRNA in cells (B) and protein on the cell membrane (C, D) obtained by siRNA delivery (10 μg siRNA) using various methods in HECs. At 24 hours after siRNA delivery, LPS/D-galactosamine (1 μg/100 mg/kg) was intraperitoneally administered into mice to induce the acute inflammatory responses. HECs were isolated by collagenase perfusion, one-step density gradient centrifugation, and magnetic cell sorting as described in the Supporting Materials and Methods. The icam-1 mRNA and protein expression in HECs was determined via quantitative reverse-transcription polymerase chain reaction (B), western blotting/enzyme-linked immunosorbent assay (C), and confocal images (D). The expression levels of mRNA and protein were detected at 3 and 6 hours after LPS/D-galactosamine stimulation, respectively. *P < 0.05, **P < 0.01 versus no treatment. Each value represents the mean + SD (n = 5). ICAM-1 was labeled with anti-mouse ICAM-1 antibody and FITC-conjugated secondary antibody (green), and nuclei were counterstained with DAPI (blue). Scale bars, 100 μm. N.T., no treatment.

Figure 4.

Suppression effects of ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure on leukocyte infiltration and proinflammatory cytokine production in an LPS/D-galactosamine-induced inflammatory mouse model. (A) Evaluation schedule of leukocyte infiltration and proinflammatory cytokine production in LPS/D-galactosamine–stimulated mice. (B,C) Levels of IL-8 and MCP-1 expression in the liver (B) and the levels of TNF-α, IFN-γ, and IL-6 secretion in the serum (C) after siRNA delivery (10 μg siRNA) using various delivery methods 12 hours after LPS/D-galactosamine stimulation. **P < 0.01 versus no treatment. Each value represents the mean + SD (n = 5). N.T., no treatment. (D) Photomicrographs of infiltrated leukocytes after siRNA delivery using Man-PEG2000 BLs (10 μg siRNA) and US exposure in LPS/D-galactosamine–stimulated mouse liver. Leukocytes were labeled with anti-mouse Gr-1 (Ly-6G) antibody and rhodamine isothiocyanate–conjugated secondary antibody (red), and nuclei were counterstained with DAPI (blue). Scale bars, 100 μm. **P < 0.01 versus no treatment. Each value represents the mean + SD (n = 5). N.T., no treatment.

Suppression Effects of LPS-Induced ICAM-1 Expression In Vitro

As shown in Fig. 1B,C, ICAM-1 expression induced by LPS stimulation was suppressed by approximately 80% in siRNA transfer using Man-PEG2000 BLs and US exposure. The suppression effect of ICAM-1 expression was not observed for scrambled siRNA. Moreover, this suppression effect was comparable to that by Lipofectamine 2000 (Fig. 1D) but with decreased cytotoxicity (Fig. 1E).

In Vivo distribution of ICAM-1 siRNA

We investigated the pharmacokinetic profiles and the tissue distribution of BLs after intravenous administration into mice. Compared with nonmodified BLs, the retention time of Man-PEG2000 BLs in the blood was reduced, and localization in both the liver and spleen were increased (Fig. 2A). Moreover, a large amount of ICAM-1 siRNA was distributed in HECs that abundantly express mannose receptors when delivered using Man-PEG2000 BLs and US exposure (Fig. 2B,C).

Suppression Effects of Drug-Induced Hepatic ICAM-1 Expression In Vivo

The suppression of ICAM-1 expression by siRNA delivery was investigated in an LPS/D-galactosamine–induced acute hepatitis mouse model (Fig. 3A). As shown in Fig. 3B-D, ICAM-1 mRNA and protein levels in HECs induced by LPS/D-galactosamine stimulation were suppressed by approximately 80% using Man-PEG2000 BLs and US exposure. Moreover, ICAM-1 expression induced by CCl4 and DMN stimulation was also significantly suppressed by the same ICAM-1 siRNA delivery system (Supporting Figs. 6B and 7B). The effects of siRNA dose on ICAM-1 suppression and the duration of ICAM-1 suppression were examined in an LPS/D-galactosamine-induced inflammatory mouse model. Following siRNA delivery using Man-PEG2000 BLs and US exposure, suppression was obtained at 10 μg of ICAM-1 siRNA (Supporting Fig. 5A), and was sustained for at least 3 days (Supporting Fig. 5B).

Figure 5.

Suppression effects of ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure on liver toxicity in an LPS/D-galactosamine–induced inflammatory mouse model. (A) The level of serum ALT/AST activities after siRNA delivery (10 μg siRNA) using various methods at predetermined times after LPS/D-galactosamine stimulation. Each value represents the mean ± SD (n = 5). (B) Fluorescent images of apoptosis after siRNA delivery using Man-PEG2000 BLs (10 μg siRNA) and US exposure in LPS/D-galactosamine–stimulated mice. Apoptosis (green) was detected via terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling, and nuclei were counterstained with DAPI (blue). Scale bars, 100 μm. (C) Liver histology with H&E staining 24 hours after siRNA delivery using Man-PEG2000 BLs (10 μg siRNA) and US exposure in LPS/D-galactosamine–induced inflammatory mouse model. Black arrows: destruction of tube formation in hepatic central vein. Scale bars, 100 μm.

Figure 6.

Suppression effects of ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure on icam-1 mRNA and protein expression in HECs of an IR-induced hepatic inflammatory mouse model. (A) Evaluation schedule of ICAM-1 expression in hepatic IR-stimulated mice. (B,C) Expression level of icam-1 mRNA in cells (B) and protein on the cell membrane (C) obtained by siRNA delivery (10 μg siRNA) using various delivery methods in HECs. HECs were isolated via collagenase perfusion, one-step density gradient centrifugation, and magnetic cell sorting as described in the Supporting Materials and Methods. The icam-1 mRNA and protein expression in HECs was determined via quantitative reverse-transcription polymerase chain reaction (B) and western blotting/enzyme-linked immunosorbent assay (C). Expression levels of mRNA and protein were detected at 3 and 6 hours after IR stimulation, respectively. *P < 0.05, **P < 0.01 versus sham operation. Each value represents the mean + SD (n = 5).

Figure 7.

Suppression effects of ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure on leukocyte infiltration and proinflammatory cytokine production in IR-induced hepatic inflammatory mouse model. (A) Evaluation schedule of leukocyte infiltration and proinflammatory cytokine production in hepatic IR-stimulated mice. (B,C) Levels of IL-8 and MCP-1 expression in the liver (B) and TNF-α, IFN-γ, IL-6 secretion in the serum (C) after siRNA delivery (10 μg siRNA) using various delivery methods 6 hours after IR stimulation. **P < 0.01 versus sham operation. Each value represents the mean + SD (n = 5).

Anti-inflammatory Effects Against Drug-Induced Hepatitis

First, the suppression of leukocyte infiltration by ICAM-1 siRNA delivery was evaluated in an LPS/D-galactosamine–induced inflammatory mouse model (Fig. 4A). As shown in Fig. 4B,D, the expression of interleukin (IL)-8 and monocyte chemotactic protein 1 (MCP-1) was suppressed, and a significantly decreased number of infiltrated leukocytes were detected after siRNA delivery using Man-PEG2000 BLs and US exposure. Moreover, the production of proinflammatory cytokines (tumor necrosis factor α [TNF-α], interferon-γ [IFN-γ], and IL-6) were also suppressed by this siRNA delivery (Fig. 4C).

The anti-inflammatory effects obtained by ICAM-1 siRNA delivery were investigated next. As shown in Fig. 5A, alanine aminotransferase (ALT)/aspartate aminotransferase (AST) activities in the serum were markedly suppressed by siRNA delivery using Man-PEG2000 BLs and US exposure (Fig. 5A). As shown in Fig. 5B, hepatic apoptosis induced by LPS/D-galactosamine stimulation was significantly suppressed by this ICAM-1 siRNA delivery. Moreover, we performed hematoxylin and eosin (H&E) staining of liver sections to evaluate the effects on hepatic structural features. Although the circular and tube formations of the hepatic central vein were observed in normal liver section (Fig. 5C, left), they were crushed in the LPS-stimulated liver section (Fig. 5C, middle). On the other hand, destruction of the hepatic central vein induced by LPS stimulation was significantly suppressed by ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure (Fig. 5C, right), suggesting that the liver injury induced by LPS-stimulation is suppressed by this siRNA delivery. Similar effects by this ICAM-1 siRNA delivery were also observed for CCl4- and DMN-induced inflammatory mouse models (Supporting Figs. 6C,D and 7C,D).

Anti-inflammatory Effects Against IR-Induced Liver Injury

The effects of ICAM-1 suppression by delivery of siRNA was evaluated for IR-induced liver injury (Fig. 6A). As shown in Fig. 6B,C, ICAM-1 expression induced by IR stimulation was suppressed by siRNA delivery using Man-PEG2000 BLs and US exposure. Moreover, IL-8/MCP-1 expression and proinflammatory cytokine production were also suppressed (Fig. 7B,C). Following the examination of liver toxicity, ALT/AST activities in the serum and hepatic apoptosis were significantly suppressed (Fig. 8A,B). Moreover, after H&E staining of liver sections, the circular and tube formations of hepatic central vein in the normal liver (Fig. 8C, left) section is destructed by IR stimulation (Fig. 8C, middle), on the other hand, IR-derived destruction of hepatic central vein was suppressed by this ICAM-1 siRNA delivery (Fig. 5C, right).

Figure 8.

Suppression effects of ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure on liver toxicity in an IR-induced hepatic inflammatory mouse model. (A) The level of serum ALT/AST activities after siRNA delivery (10 μg siRNA) using various delivery methods 24 hours after hepatic IR stimulation. **P < 0.01versus the corresponding sham operation group. Each value represents the mean + SD (n = 5). (B) Fluorescent images of apoptosis followed by siRNA delivery using Man-PEG2000 BLs (10 μg siRNA) and US exposure in IR-induced hepatic inflammatory mouse model. Apoptosis (green) was detected via terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate nick-end labeling, and nuclei were counterstained with DAPI (blue). Scale bars, 100 μm. (C) Liver histology at 24 hours after siRNA delivery using Man-PEG2000 BLs (10 μg siRNA) and US exposure in IR-induced hepatic inflammatory mouse model. Arrows indicate the destruction of tube formation in the hepatic central vein. Scale bars, 100 μm.

Discussion

In the sonoporation method, transient pores are created on the cell membrane followed by the destruction of microbubbles, and a large amount of nucleic acids can be directly transferred into the cytoplasm.21, 26, 27 Because siRNA is functionalized in the cytoplasm, gene transfer using Man-PEG2000 BLs and US exposure23-26 would be also suitable for siRNA delivery. In the present study, we applied this gene transfer method for the selective and efficient delivery of siRNA to HECs in vivo and investigated the anti-inflammatory effects in various types of inflammatory responses.

The innate inflammatory responses based on the interaction with siRNA and Toll-like receptor (TLR)-3, -7, and -8 should be excluded for evaluating the gene suppression effects of siRNA, but should be considered for clinical applications of siRNA.28, 29 The proinflammatory cytokines (such as TNF-α, IFN-γ, and IL-6) can be induced by siRNA interaction with endosomal TLR-3, -7, and -8 in siRNA transfer using conventional nonviral carriers.28, 29 Transfer of siRNA using Man-PEG2000 BLs and US exposure results in the direct deposition into the cytoplasm and is not mediated by endocytosis (Fig. 1A and Supporting Fig. 3C).26 Therefore, the inflammatory responses followed by the interaction with TLRs are expected to be low, but siRNA is also recognized by cytoplasmic retinoic acid-inducible gene 1 (RIG-1)/melanoma differentiation-associated gene 5 (MDA-5) involved in inflammatory responses.28, 30 Because the modification of 3′-overhang sequences is suppressed by the activation of interferon-responsive factors 3/7, transcriptional factors that exist downstream of the RIG-1/MDA-5 pathway,31, 32 we used siRNAs with 3′-dTdT overhang sequences (Supporting Fig. 1A).

As shown in Figs. 1B-D and 3, ICAM-1 expression in LPS-stimulated HECs was significantly suppressed by ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure, both in vitro and in vivo. Similarly, tissue infiltration of leukocytes and proinflammatory cytokine production were both suppressed after ICAM-1 suppression by siRNA delivery using this method (Fig. 4). Furthermore, potent anti-inflammatory effects were obtained by this ICAM-1 siRNA delivery in an LPS-stimulated inflammatory mouse model (Fig. 5). The delivery of siRNA to HECs, which express mannose receptors (Supporting Fig. 4),33 was selective and efficient using Man-PEG2000 BLs with US exposure (Fig. 2B,C). Moreover, because a large amount of siRNA was directly transferred into the cytoplasm (Fig. 1A and Supporting Fig. 3C),26 endosomal escape and degradation within endosomes could be evaded. These data may indicate that nucleic acid transfer using Man-PEG2000 BLs and US exposure can be applied for siRNA delivery.

Although LPS is widely used to evaluate the induction of acute inflammatory responses, they are induced by not only various medicines but also surgical operations.34 Aiming for the clinical application of anti-inflammatory therapy using our siRNA delivery method, the anti-inflammatory effects against various inflammatory models in mice were investigated. After evaluation of the anti-inflammatory effects against CCl4-, DMN-, and IR-stimulated inflammation, ICAM-1 expression in HECs and the inflammatory responses was significantly suppressed by ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure in these inflammatory mouse models (Figs. 6-8 and Supporting Figs. 6 and 7). Although the mechanisms of inflammatory responses as a result of LPS, CCl4, DMN, and IR stimulation are different,5, 6, 35, 36 ICAM-1 expression in HECs is reported in various types of inflammation, including drug-induced hepatic inflammation and IR-induced liver injury.7 These data suggest that anti-inflammatory effects obtained by ICAM-1 siRNA delivery using Man-PEG2000 BLs and US exposure may be beneficial for acute hepatitis and liver injury.

In the present study, efficient ICAM-1 suppression was obtained at a dose of 1 μg siRNA/mouse (0.05 mg/kg) for siRNA delivery using Man-PEG2000 BLs and US exposure in vivo (Supporting Fig. 5A). This dose of siRNA is lower than those reported for other studies evaluating the therapeutic effects using siRNA, although the therapeutic mechanism and delivery methods of each siRNA are likely to be different.37-39 These findings suggest that the increased distribution of siRNA into HECs by mannose modification (Fig. 2) and the enhancement of intracytoplasmic siRNA transfer by sonoporation (Fig. 1A and Supporting Fig. 3) could contribute to the potent anti-inflammatory effects observed at a low dose of siRNA in our siRNA delivery method.

ICAM-1 suppression effects were only sustained for 72 hours by siRNA delivery using Man-PEG2000 BLs and US exposure (Supporting Fig. 5B). However, because the disease target of this study was acute inflammation, the potent therapeutic effects might be obtained in short duration and single administration of siRNA. Recently, it has been reported that ICAM-1 is involved in various diseases not only for acute/chronic hepatic failure, but also Crohn's disease, ulcerative colitis, and ileus.40-42 In addition, antisense oligonucleotides against ICAM-1 (ISIS-2302; Alicaforsen) are currently under development for the treatment of Crohn's disease and ulcerative colitis.43, 44 However, most of these inflammatory diseases are based on chronic inflammation. In the present study, it is strongly suggested that transfer of ICAM-1 siRNA using Man-PEG2000 BLs and US exposure enables a large amount of siRNA to be delivered the cytoplasm of targeted cells (Fig. 1A and Supporting Fig. 3). Therefore, to prolong the duration of gene suppression using this siRNA delivery system, future studies using cholesterol-modified siRNA45 or locked nucleic acid,46 which are forms of stable siRNA resistant to enzymatic degradation, might be necessary for application to a variety of chronic inflammatory diseases.

In conclusion, ICAM-1 siRNA was transferred into HECs selectively and efficiently, and sufficient ICAM-1 suppression effects were obtained by ICAM-1 siRNA transfer using Man-PEG2000 BLs and US exposure, both in vitro and in vivo. Moreover, potent anti-inflammatory effects were achieved against various types of inflammation by this ICAM-1 siRNA transfer. These findings contribute to overcoming the poor efficiency of siRNA transfer into the cytoplasm of the targeted cells using nonviral carriers, and this novel siRNA delivery method using Man-PEG2000 BLs and US exposure may offer a valuable system for medical treatment where the cellular targets are HECs.

Ancillary