Reduction of advanced liver fibrosis by short-term targeted delivery of an angiotensin receptor blocker to hepatic stellate cells in rats

Authors

  • Montserrat Moreno,

    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Centro de investigación biomédica en red de enfermedades hepáticas y digestivas (CIBERehd), University of Barcelona, Barcelona, Catalonia, Spain
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Teresa Gonzalo,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Robbert J. Kok,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    2. Department of Pharmaceutics, Utrecht Institute for Pharmaceutical Sciences, Utrecht, The Netherlands
    Search for more papers by this author
  • Pau Sancho-Bru,

    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Centro de investigación biomédica en red de enfermedades hepáticas y digestivas (CIBERehd), University of Barcelona, Barcelona, Catalonia, Spain
    Search for more papers by this author
  • Marike van Beuge,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    Search for more papers by this author
  • Josine Swart,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    Search for more papers by this author
  • Jai Prakash,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    Search for more papers by this author
  • Kai Temming,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    2. Kreatech Biotechnology BV, Amsterdam, The Netherlands
    Search for more papers by this author
  • Constantino Fondevila,

    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Centro de investigación biomédica en red de enfermedades hepáticas y digestivas (CIBERehd), University of Barcelona, Barcelona, Catalonia, Spain
    Search for more papers by this author
  • Leonie Beljaars,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    Search for more papers by this author
  • Marie Lacombe,

    1. Kreatech Biotechnology BV, Amsterdam, The Netherlands
    Search for more papers by this author
  • Paul van der Hoeven,

    1. Kreatech Biotechnology BV, Amsterdam, The Netherlands
    Search for more papers by this author
  • Vicente Arroyo,

    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Centro de investigación biomédica en red de enfermedades hepáticas y digestivas (CIBERehd), University of Barcelona, Barcelona, Catalonia, Spain
    Search for more papers by this author
  • Klaas Poelstra,

    1. Department of Pharmacokinetics and Drug Delivery, Groningen University Institute for Drug Exploration, Groningen, The Netherlands
    Search for more papers by this author
  • David A. Brenner,

    1. Department of Medicine, San Diego School of Medicine, University of California, San Diego, CA
    Search for more papers by this author
  • Pere Ginès,

    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Centro de investigación biomédica en red de enfermedades hepáticas y digestivas (CIBERehd), University of Barcelona, Barcelona, Catalonia, Spain
    Search for more papers by this author
  • Ramón Bataller

    Corresponding author
    1. Liver Unit, Hospital Clínic, Institut d'Investigacions Biomèdiques August Pi I Sunyer (IDIBAPS), Centro de investigación biomédica en red de enfermedades hepáticas y digestivas (CIBERehd), University of Barcelona, Barcelona, Catalonia, Spain
    • Liver Unit, Hospital Clínic, IDIBAPS, C/Casanova, 143 Barcelona, Catalonia, Spain
    Search for more papers by this author
    • fax: +34-934515522


  • Potential conflict of interest: Nothing to report.

Abstract

There is no effective therapy for advanced liver fibrosis. Angiotensin type 1 (AT1) receptor blockers attenuate liver fibrogenesis, yet their efficacy in reversing advanced fibrosis is unknown. We investigated whether the specific delivery of an AT1 receptor blocker to activated hepatic stellate cells (HSCs) reduces established liver fibrosis. We used a platinum-based linker to develop a conjugate of the AT1 receptor blocker losartan and the HSC-selective drug carrier mannose-6-phosphate modified human serum albumin (losartan-M6PHSA). An average of seven losartan molecules were successfully coupled to M6PHSA. Rats with advanced liver fibrosis due to prolonged bile duct ligation or carbon tetrachloride administration were treated with daily doses of saline, losartan-M6PHSA, M6PHSA or oral losartan during 3 days. Computer-based morphometric quantification of inflammatory cells (CD43), myofibroblasts (smooth muscle α-actin [α-SMA]) and collagen deposition (Sirius red and hydroxyproline content) were measured. Hepatic expression of procollagen α2(I) and genes involved in fibrogenesis was assessed by quantitative polymerase chain reaction. Losartan-M6PHSA accumulated in the fibrotic livers and colocalized with HSCs, as assessed by immunostaining of anti-HSA and anti–α-SMA. Losartan-M6PHSA, but not oral losartan, reduced collagen deposition, accumulation of myofibroblasts, inflammation and procollagen α2(I) gene expression. Losartan-M6PHSA did not affect metalloproteinase type 2 and 9 activity and did not cause apoptosis of activated HSCs. Conclusion: Short-term treatment with HSC-targeted losartan markedly reduces advanced liver fibrosis. This approach may provide a novel means to treat chronic liver diseases. (HEPATOLOGY 2010.)

Hepatic fibrosis is the consequence of most types of chronic liver diseases.1 There are no effective therapies to treat liver fibrosis in patients in which the causative agent cannot be removed.2 In experimentally-induced liver fibrosis, several agents reduce progression of the disease.3 Inhibitors of the renin-angiotensin system (RAS) are probably the most promising drugs. There is extensive evidence indicating that the RAS regulates liver fibrogenesis.4 RAS components are overexpressed in livers with fibrosis and angiotensin II induces inflammatory and fibrogenic effects in vivo and in activated hepatic stellate cells through AT1 receptors (HSC).5, 6 The blockade of AT1 receptors reduces the accumulation of activated HSCs and attenuates liver fibrosis in rats7 and AT1 receptor–deficient mice exhibit attenuated response to hepatic inflammation and fibrosis.8 However, the efficacy of AT1 receptor blockers to reverse established fibrosis is unknown. We propose an innovative approach to deliver drugs to activated HSCs, increasing the concentration in the liver at the sites of active fibrogenesis. Moreover, drug delivery can be useful to avoid systemic undesirable effects such as renal dysfunction.

The drug delivery system applied in this study uses mannose 6-phosphate modified human serum albumin (M6PHSA), a carrier that delivers drugs to activated HSCs.9 M6PHSA binds to the mannose-6-phosphate/insulin growth factor type II receptor (M6P/IGII-R), a surface exposed receptor that is de novo expressed in activated HSCs during liver fibrogenesis.10 Prior studies demonstrated rapid and efficient accumulation of drug-M6PHSA conjugates in the fibrotic liver.11, 12 To conjugate losartan to M6PHSA, we employed a novel type of platinum linker called ULS (Universal Linkage System), which can bind losartan via a coordinative bond at one of the aromatic nitrogen atoms in the tetrazole group.13–15 Application of this coordinative linker technology has several important advantages, for instance straightforward coupling of drugs, adequate stability of conjugates, and slow-release of the active pharmacon within target cells.11

In the present study, we administered losartan-M6PHSA for a short period of time to rats with advanced fibrosis. We demonstrate that losartan-M6PHSA accumulates exclusively in the fibrotic liver at the sites of activated HSCs. Importantly, treatment with losartan-M6PHSA, but not free losartan given orally, reduced both hepatic inflammation and fibrosis.

Abbreviations

AT1, angiotensin type 1 receptor; CCl4, carbon tetrachloride; HSC, hepatic stellate cell; IGF II, insulin-like growth factor II; M6PHSA, mannose-6-phosphate modified human serum albumin; RAS, renin-angiotensin system; ULS, universal linkage system.

Materials and Methods

Synthesis of Losartan-M6PHSA.

Losartan and human serum albumin (HSA) were obtained from Synfine (Ontario, Canada) and Sanquin (Amsterdam, The Netherlands), respectively. Losartan was first coupled to Universal Linkage System (ULS; Kreatech BV, The Netherlands). ULS was prepared as described elsewhere.11 ULS (32 μmol) in dimethylformamide (DMF) was added to a solution of losartan (32 μmol, 10 mg/mL of the potassium salt of losartan in DMF). Mass spectrometry analysis confirmed the presence of the 1:1 losartan-ULS species after completion of the reaction, whereas 195Pt-NMR confirmed the coordination of Pt(II) to a nitrogen donor site. Ion-spray mass spectrometry (ESI+) mass-to-charge ratio (m/z): 711-717 [losartan-ULS-Cl]+, 748-754 [losartan-ULS-DMF]+ 195Pt NMR of losartan-ULS (CD3OD): −2491 and −2658 ppm. M6PHSA was prepared and characterized as described previously.16 A total of 10 mg M6PHSA (14.3 nmol) was dissolved in 1 mL 20 mM tricine/NaNO3 buffer (pH 8.5) and reacted with losartan-ULS (143 nmol) in 10-fold molar excess overnight at 37°C. The losartan-M6PHSA product was purified by dialysis against PBS at 4°C, filter-sterilized and stored at −20°C. Protein content of the conjugates was assessed by the BCA assay (Pierce, Rockford, IL). ULS content per losartan-M6PHSA was evaluated by inductively coupled plasma–atomic emission spectroscopy (ICP-AES) at 214.424 nm and at 265.945 nm with a VISTA AX CCD Simultaneous ICP-AES (Varian, Palo Alto, CA). Standards (cisplatin) and unknown samples were spiked with yttrium as an internal standard (360.074 nm). Losartan conjugated to M6PHSA was determined after competitive dissociation of drug-ULS bonds by potassium thiocyanate, as described previously.11, 15 High performance liquid chromatography (HPLC) analyses were performed on a thermostated C18 column (Sunfire; Waters Inc., Milford, MA) with an isocratic mobile phase consisting of acetonitrile–water–trifluoroacetic acid (30:70:0.1, vol/vol/vol; pH 2.0). Losartan-M6PHSA and M6PHSA were subjected to anion-exchange and size exclusion chromatography as described.9

Animal Experimental Procedures.

Liver fibrosis was induced in 250 g male Wistar rats (Harlan, Zeist, The Netherlands) by either bile duct ligation or chronic treatment with CCl4. For the bile duct ligation,17 rats were anesthesized with isoflurane (2% isoflurane in 2:1 O2/N2O, 1 L/minute; Abbot Laboratories Ltd., Queensborough, UK). The common bile duct was doubly ligated with 4-0 silk and transected between the two ligations. Sham operation was performed similarly with exception of ligating and transecting the bile duct. Animals were sacrificed 15 days after surgery. Arterial blood pressure was measured immediately before tissue harvesting. Animals were anesthesized with pentobarbital (30 mg/kg intraperitoneally) and the right carotid artery was cannulated (PE-90; Transonics Systems Inc., Ithaca, NY). The mean arterial blood pressure was recorded using a pressure analyzer (LPA-200; Digi-Med, Louisville, KY) for 10 minutes. In the model of CCl4-induced liver fibrosis, rats were fed ad libitum with standard chow and drinking water containing phenobarbital (0.3 g/L). Fibrosis was induced by inhalation of CCl4 for 8 weeks as described previously.18 In both experimental models, rats received a daily injection by the penis vein of saline, losartan-M6PHSA (3.3 mg/kg/day, corresponding to 125 μg losartan/kg), M6PHSA alone (3.3 mg/kg/day), or an oral administration of losartan by gavage (5 mg/kg/day) at 72, 48, and 24 hours before sacrifice. For pharmacokinetic studies, a subset of rats received an additional dose of the treatments 10 minutes before sacrifice. To determine the efectiveness of long-term treatment with losartan-M6PHSA on advanced fibrosis, rats were treated by CCl4 inhalation for 10 weeks. During the last 3 weeks, rats received saline, losartan-M6PHSA (3.3 mg/kg/day), or M6PHSA alone (3.3 mg/kg/day) by the penis vein twice a week. At least 10 rats were included per group in both models. Animal procedures were approved by the Committee for Care and Use of Laboratory Animals of the Hospital Clínic, Barcelona, and are in accordance with National Institutes of Health guidelines.

Analysis of Losartan-M6PHSA Biodistribution.

The presence of losartan-M6PHSA or M6PHSA in tissue cryosections was demonstrated by immunostaining using an anti-HSA antibody (Cappel ICN Biomedicals, Zoetermeer, The Netherlands).19 The colocalization of losartan-M6PHSA with HSC was assessed by double immunostaining of anti-HSA (Cappel ICN Biomedicals, Zoetermeer, The Netherlands) and anti–α-SMA (Abcam, Cambridge, UK). To avoid cross-reactivity of anti-HSA antibody with rat albumin, normal rat serum was added to the antibody. Sections from rats that did not receive HSA were completely negative after the anti-HSA staining. The amount of losartan in liver tissue homogenates was analyzed by HPLC as described above. Two different procedures were employed to isolate losartan from tissue homogenates. The first method consisted of direct extraction from the livers, whereas the second method comprised an additional incubation of tissues overnight with potassium thiocyanate in order to chemically release conjugate-bound losartan, as described above.

Quantification of Collagen Accumulation and Infiltration by Myofibroblastic Cells and Apoptosis.

The degree of hepatic fibrosis was estimated as the percentage of area stained with picrosirius Red (Sirius Red F3B; Gurr-BDH Lab Supplies, Poole, UK).6 The amount of fibrogenic myofibroblasts was estimated by measuring the percentage of area stained with α-SMA antibody (DAKO, Carpinteria, CA). For morphometric assessment of percentage of area with positive staining, an optic microscope (Nikon Eclipse E600) connected to a high-resolution camera (CC12 Soft-Imaging System, Münster, Germany) was used. Images were analyzed in an automated image-analysis system (AnalySIS, Soft-Imaging System, Münster, Germany). Results are given as percentage of positive area. Cell apoptosis was quantified by using In Situ Death Detection Kit, POD (Roche Applied Science, Barcelona, Spain) based on terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL). TUNEL-positive cells per high-power field (200×) were counted. All measurements were performed blindly.

Statistical Analysis.

Results are expressed as the mean ± standard error of the mean. Significance was established using Student t test, two-way analysis of variance with Bonferroni's post hoc test and Mann-Whitney assay. Differences were considered significant if P < 0.05.

Other methods are shown in Supporting Materials and Methods.

Results

Synthesis of Losartan-M6PHSA and Internalization by HSCs.

Losartan was conjugated to manose-6-phosphate coupled to human serum albumin (M6PHSA) (Fig. 1A). After its reaction to the linker at a stoichiometric ratio (Fig. 1B), the losartan-ULS adduct was conjugated to M6PHSA. An average of seven losartan-ULS molecules were coupled to M6PHSA, as assessed by HPLC and confirmed by inductive coupled plasma-atomic emission spectroscopy (ICP-AES) (data not shown). Conjugation of losartan to M6PHSA did not change the charge or size features of M6PHSA, as assessed by anion-exchange chromatography and size exclusion chromatography, respectively (Fig. 1C,D). Because ULS is a derivative of cisplatin, an antitumor agent that may cause cell toxicity, we studied the effects of losartan-M6PHSA on cultured HSCs. Losartan-M6PHSA did not cause cell toxicity, while cisplatin induced cell death, suggesting that occupation of the coordinative sites of platinum with drug and carrier prevents its disruptive reactivity with cellular components (Fig. 1E). To test whether losartan-M6PHSA is biologically active in cultured HSCs, cells were stimulated with angiotensin II in the presence or absence of either free losartan or losartan-M6PHSA. We found that both treatments equally blunted angiotensin II–induced intracellular calcium increase (Fig. 1F). Also, we detected intracellular staining for HSA after incubating HSCs with losartan-M6PHSA for 10 minutes. This staining was strongly blunted by excess of M6P sugars and an antibody against the M6P/IGF II receptor. We found 25.2 ± 2.4, 0.2 ± 0.1, and 5.3 ± 0.6 positive cells in cultures incubated with isotype-matched antibody, excess of M6P, and anti-IGFRII antibody, respectively (P < 0.001 of isotype-matched antibody respect to the other two conditions) (Fig. 2A). These results indicate that losartan-M6PHSA directly interacts with IGF II receptors present in HSCs, and is internalized to inhibit angiotensin II–induced biological actions.

Figure 1.

Synthesis and characterization of losartan-M6PHSA. (A) Coordinative linkage between losartan, ULS, and M6PHSA. M6PHSA consists of an albumin core modified with approximately 28 mannose-6-phosphate (M6P) groups. (B) The formation of losartan-ULS was confirmed by ion-spray mass spectrometry, which demonstrated the typical isotopic pattern of platinum compounds. MonoQ anion exchange chromatography (C) confirmed that the charge of losartan-M6PHSA was not affected by losartan coupling. Bold line: losartan-M6PHSA; thin line: M6PHSA. (D) Size exclusion chromatography showed that losartan-M6PHSA and M6PHSA consisted for >90% of monomeric material and a minor fraction of dimeric albumin. Bold line: losartan-M6PHSA; thin line: M6PHSA. (E) Losartan-M6PHSA induced no cytotoxicity, as assessed by the Alamar Blue viability assay when incubated with rat HSCs for 24 hours. Indicated compounds were equivalent to 100 μM of platinum (cisplatin, losartan-M6PHSA) or equivalent to the amount of drug (100 μM) or carrier (1 mg/mL) in the losartan-M6PHSA preparation. *P < 0.001 compared to control HSCs and losartan-M6PHSA using Student t test. (F) Changes in [Ca++]i in cultured HSCs stimulated with saline or angiotensin II (10−8 M) in the presence of different compounds. Results are the mean of at least 15 cells per condition.

Figure 2.

Analysis of biodistribution of losartan-M6PHSA in HSCs and in rats with advanced fibrosis induced by prolonged bile duct ligation. (A) Representative pictures of cultured hepatic stellate cells (HSCs) visualized with confocal microscopy. Cells were incubated with losartan-M6PHSA (1 mg/mL) for 4 hours in the presence of an isotype-matched antibody (left picture), an anti–IGF-II antibody (middle picture), and with an excess of M6P, a ligand for the M6P/IGF-II receptor (right picture). Losartan-M6PHSA was clearly seen inside HSCs treated with isotype antibody (arrows), whereas both an anti–IGF-II antibody and M6P-HSA markedly prevented losartan-M6PHSA uptake. (B) Losartan-M6PHSA was not detected in the lung, spleen, heart, or kidney, but was detected in the liver within the nonparenchymal cells of rats treated with losartan-M6PHSA (magnification 40×). Staining was absent in rats treated with saline (not shown). (C) Losartan-M6PHSA colocalized with stellate cells in rat liver (arrow), as assessed with double immunostaining with anti-HSA and anti–α-SMA. Epifluorescence microscopy verified the colocalization of anti–α-SMA (red fluorescence color) and anti-HSA (green fluorescence color) in the regions colored in yellow (arrow) (magnification 400×). Five rats were studied per group. (D) Quantification of losartan in liver homogenates by HPLC. Absolute levels of losartan in the liver were highest for orally administered losartan (left panel), but represented a five-fold lower relative accumulation (right panel) in view of the different doses administered. *P < 0.05 (losartan).

Pharmacokinetics of Losartan-M6PHSA in Bile Duct –Ligated Rats.

M6PHSA binds to M6P/IGFII-R, which is expressed in activated HSCs in the fibrotic liver.16 In the bile duct ligation model, we administered losartan-M6PHSA (3.3 mg/kg, corresponding to 125 μg losartan/kg) daily from day 12-14 and animals were sacrificed at day 15. For pharmacokinetic purposes, a subgroup of the animals received an additional dose of the conjugate at 10 minutes before sacrifice. Control groups were treated with equivalent doses of M6PHSA (3.3 mg/kg), saline, or free losartan given orally at a dose (5 mg/kg) that has been shown to attenuate liver fibrosis when given for a prolonged period of time.20, 21 We first assessed whether losartan-M6PHSA preferentially accumulates in the fibrotic rat liver. The liver and other organs (lungs, heart, spleen, and kidneys) were stained with anti-HSA to detect the presence of the albumin-based conjugate. Losartan-M6PHSA was only detected in the liver (Fig. 2B). Injection of the carrier alone (M6PHSA) followed a similar distribution pattern (not shown). Importantly, losartan-M6PHSA colocalized with activated HSCs, as assessed by double immunostaining with anti-HSA and anti–α-SMA antibodies (Fig. 2C). To further demonstrate the selective homing of losartan-M6PHSA in the liver, tissue levels of losartan were quantified by HPLC. Animals receiving losartan-M6PHSA showed losartan levels which corresponded to 81% of the last injected dose being at least 20% of the cumulative dose (Fig. 2D). In contrast, oral losartan yielded liver tissue levels corresponding to only 4% of the cumulative dose (15% of the last dose administered). These results illustrate the preferential hepatic accumulation of losartan-M6PHSA. However, because free losartan was administered at a 40-fold higher dose as compared to targeted losartan, the control treatment yielded nine-fold higher absolute concentrations.

Treatment with Losartan-M6PHSA, but not Oral Losartan, Reduces Advanced Liver Fibrosis.

Rats were submitted to prolonged ligation of the common bile duct, which induces profound changes in the hepatic architecture including bridging fibrosis.17 As expected, bile duct ligation for 15 days resulted in a marked increase in serum bilirubin and aminotransferase levels, which were unaffected by any of the treatments. Bile duct–ligated rats treated with saline or M6PHSA alone showed severe septal fibrosis (Fig. 3A). Hepatic collagen, as assessed by morphometric analysis of Sirius red staining and hydroxyproline content, was markedly increased in these rats as compared to sham-operated rats (Fig. 3A,B). In contrast, bile duct–ligated rats treated with losartan-M6PHSA showed less collagen deposition with less frequent formation of bridging fibrosis. Importantly, short-term oral treatment with losartan alone did not reduce histological fibrosis or the amount of collagen content. To confirm these results, hepatic procollagen α2(I) gene expression was quantified. Procollagen α2(I) was up-regulated 10-fold in bile duct–ligated rats treated with saline compared with sham-operated animals. Losartan-M6PHSA, but not oral losartan or M6PHSA alone, reduced procollagen α2(I) by 60% (Fig. 3C). These results indicate that short-term treatment with losartan-M6PHSA, but not oral losartan, attenuates advanced liver fibrosis. To provide additional evidence of the antifibrotic effects of HSC-targeted losartan, liver fibrosis was also induced by CCl4 for 8 weeks.18 Rats treated with CCl4 for 8 weeks showed a marked distortion of the hepatic architecture with bridging fibrosis. At the end of the treatment period, rats received three consecutive daily doses of saline, oral losartan, losartan-M6PHSA, or M6PHSA alone. Similar to bile duct–ligated rats, we administered a final dose 10 minutes before sacrifice, to enable the detection of losartan-M6PHSA in the tissues. Losartan-M6PHSA accumulated in the fibrotic liver to a similar extent (13% ± 6% of the cumulative dose, n = 10, data not shown) as observed in bile duct–ligated rats. Hepatic collagen content, as assessed by morphometric analysis of Sirius red staining, hydroxyproline content, and procollagen α2(I) gene expression, was reduced in rats treated with losartan-M6PHSA (Fig. 3D,E,F). Finally, none of the treatments in both experimental models induced changes in renal function, as indicated by normal serum creatinine levels, nor histological changes in the heart or the kidney (data not shown). Both losartan-M6PHSA and oral losartan induced a slight decrease in arterial pressure (data not shown). All together, these results demonstrate that short-term treatment with losartan targeted to HSCs is highly effective in attenuating liver fibrosis in rats. To investigate whether long-term treatment with losartan-M6PHSA was also effective, a new experimental procedure was carried out. Advanced liver fibrosis was established by CCl4 inhalation for 10 weeks. During the last 3 weeks, rats were given saline, losartan-M6PHSA, or M6PHSA alone twice a week. We found that losartan-M6PHSA was able to reduce collagen synthesis, as assessed by reduced expression of procollagen α1(I) and procollagen α2(I). However, the amount of activated HSCs (as assessed by α-SMA expression) and the degree of collagen accumulation (as assessed by Sirius red staining) were not significantly reduced (Supporting Fig. 1). Further studies identifying the ideal route and drug dosage from long-term studies are clearly required.

Figure 3.

Effect of different treatments on the degree of liver fibrosis in two different experimental models of liver fibrosis. (A) Bile duct ligation model. Severe bridging was observed in rats receiving saline, M6PHSA, or oral losartan. In contrast, rats treated with losartan-M6PHSA showed fewer areas with collagen accumulation (magnification 40×). Morphometric quantification of the area with Sirius red staining in rat livers showed significant inhibitory effects by losartan-M6PHSA, but not by other treatments. Pictures represent a reconstruction of 16 different areas of the liver biopsy; 40× magnification. (B) Analysis of collagen deposition in bile duct–ligated rats by measuring hydroxyproline content. Hepatic hydroxyproline markedly increased in rats with bile duct ligation compared to controls. Losartan-M6PHSA, but not oral losartan, reduced hydroxyproline content. (C) Quantification of the messenger RNA expression of procollagen α2(I). Expression was reduced by losartan-M6PHSA treatment. Losartan- M6PHSA but not oral losartan also reduced liver fibrosis in rats treated with carbon tetrachloride (CCl4) as assessed by (D) Sirius red staining, (E) hydroxyproline content, and (F) procollagen α2(I) gene expression. Pictures represent a reconstruction of 16 different areas of the liver biopsy; 40× magnification. *P < 0.05 versus sham; #P < 0.05 versus other fibrotic groups. Results are the mean of at least five different samples per condition.

Mechanisms of the Antifibrotic Effect of Losartan-M6PHSA.

To explore the mechanisms involved in the potent antifibrotic effect of losartan-M6PHSA, we first assessed the accumulation of fibrogenic myofibroblasts by morphometric quantification of α-SMA–positive cells. Bile duct ligation resulted in the accumulation of abundant α-SMA–positive cells around proliferating bile ducts as well as in the hepatic sinusoids (Fig. 4A,B). Treatment with losartan-M6PHSA, but not oral losartan or M6PHSA alone, was associated with a significant reduction in the accumulation of myofibroblasts, as determined by morphometric analysis of the positively stained area (Fig. 4C). This effect was not associated with increased HSC apoptosis (data not shown). In the CCl4 model of liver fibrosis, α-SMA hepatic immunostaining was also reduced by losartan-M6PHSA treatment.(Fig. 4D,E) Next, we assessed hepatic expression of metalloproteinases (MMP) 3 and 9 and tissue inhibitor of metalloproteinase-1 (TIMP-1). Bile duct ligation resulted in a marked increase in these four genes, which was not reduced by losartan-M6PHSA or oral losartan (Fig. 5A,B,D). However, TIMP-1 protein expression was reduced, as assessed by immunohistochemistry (Supporting Fig. 2). We also assessed the activity of metalloproteinases MMP2 and MMP9 by gelatin zymography. We found that losartan-M6PHSA did not modify MMP2 and MMP9 activity in bile duct-ligated rats (Fig. 5C). Also, we explored the hepatic expression of transforming growth factor β1 (TGF-β1), a cytokine that mediates the fibrogenic actions of angiotensin II.22 Bile duct ligated rats showed increased TGF-β1 gene expression, which was not reduced in rats treated with losartan-M6PHSA (Fig. 5E). Further studies should analyze protein expression of TGF-β1 to confirm these results. Furthermore, we explored whether losartan-M6PHSA reduces hepatic inflammation. First, we analyzed in HSCs the expression of proinflammatory genes (ICAM-1 and interleukin-8 [IL-8]). Both genes were up-regulated by angiotensin II treatment. Treatment by losartan and losartan-M6HSA reduced this effect (Fig. 6A,B). Next, in vivo liver inflammation was assessed by quantifying the infiltration of inflammatory cells (CD43-positive) in the hepatic parenchyma by immunohistochemistry. Compared to sham-operated rats, bile duct–ligated rats showed a marked increase in the infiltration of CD43-positive inflammatory cells (Fig. 7A). This effect was blunted by treatment with losartan-M6HSA and, to a lesser extent, by oral losartan. In contrast, monocyte chemotactic protein 1 expression was not modified by any of the treatments (Fig. 7C). The number of CD43-positive cells was also decreased in CCl4-treated rats (Fig. 7B).

Figure 4.

Effect of different treatments on α-SMA–positive cells in fibrotic livers. (A) Effect of different treatments on the accumulation of myofibroblasts and activated HSCs. Liver sections of bile duct–ligated animals stained with anti–smooth muscle α-actin expression (α-SMA) antibody. Bile duct–ligated animals showed a marked accumulation of α-SMA–positive cells. Rats treated with losartan-M6PHSA showed fewer α-SMA–positive cells. Pictures represent a reconstruction of 16 different areas of the liver biopsy, 40× magnification. (B) High power magnification (400×) photomicrograph of a liver from a bile duct–ligated rat treated with saline. α-SMA staining was detected in cells located in the sinusoids corresponding to activated HSCs as well as in myofibroblasts around proliferating bile ducts. (C) Morphometric quantification of the area with α-SMA staining in rat liver specimens (*P < 0.05 versus sham; #P < 0.05 versus saline, M6PHSA and oral losartan). (D) In the CCl4 model, treatment with losartan-M6PHSA reduced α-SMA staining in the liver as compared to diseased animals treated with saline. (E) Quantification of the area with α-SMA staining in liver specimens (#P < 0.05).

Figure 5.

Gene expression and metalloproteinase activity in the liver of rats submitted to different treatments. (A,B) Matrix metalloproteinases 3 and 9 (MMP3 and MMP9) expression in sham-operated rats and bile duct–ligated rats receiving various treatments. (C) Zymogram analysis showing activity for MMP2 and MMP9 in bile duct–ligated rats submitted to different treatments. (D) TIMP1 expression in sham-operated rats and bile duct–ligated rats. (E) TGF-β1 in sham-operated rats and rats with bile duct ligation receiving various treatments. Results are the mean of four independent experiments. *P < 0.05 versus sham-operated rats.

Figure 6.

Effects of different treatments on expression of proinflammatory genes in cultured primary hepatic stellate cells (HSCs). Gene expression of (A) intercellular adhesion molecule-1 (ICAM-1) and (B) interleukin-8 were measured by quantitative polymerase chain reaction in cultured HSCs. Angiotensin II (10−8 M) stimulated the expression of both genes. This effect was attenuated by both free losartan and losartan-M6PHSA. Results are the mean of four independent experiments.*P < 0.05 versus saline; #P < 0.05 versus angiotensin alone.

Figure 7.

Treatment with losartan-M6PHSA and, to a lesser extent, oral losartan reduced the number of infiltrating leukocytes in the liver parenchyma. Liver sections of (A) bile duct–ligated rats or (B) CCl4-treated rats receiving the specified treatments were processed for immunohistochemistry and stained with anti-CD43 (magnification 40×). Rats receiving saline or M6PHSA showed intense infiltration of CD43-positive leukocytes. Treatment with losartan-M6PHSA reduced the inflammatory infiltrate. Quantification of the number of positive cells in 20 randomly chosen high-power fields (*P < 0.05 versus sham; #P < 0.05 versus saline and M6PHSA). (C) Representative pictures of immunostaining for monocyte chemotactic protein 1 in bile-duct ligated rats. No differences were detected between groups.

Discussion

This study demonstrates that advanced liver fibrosis can be attenuated by short-term administration of an antifibrotic drug selectively targeted to activated HSCs. We provide evidence that the delivery of the AT1 receptor blocker losartan to activated HSCs reduces hepatic inflammation and collagen deposition. This novel approach appears to be more effective than conventional treatment with oral losartan.

The new drug conjugate losartan-M6PHSA was successfully synthesized by applying a novel linker system that binds losartan via a transition-metal coordination bond. Traditionally, linking drugs to carrier moieties represents a complex issue involving tedious drug-derivatization reaction steps.23 A key property of our platinum linker, ULS, is that it can be applied for conjugation of many valuable drug molecules containing aromatic nitrogens, forming a bond of intermediate binding strength. The ligand-exchange behavior of platinum compounds is quite slow, giving them a high kinetic stability.24 The slow rate of drug release from the linker11, 15 will cause sustained drug release within target cells and will effectuate only very low concentrations of reactive platinum in target cells, which are orders of magnitude lower than applied in cisplatin cancer therapy. One therefore would predict rapid detoxification of ULS by binding to cytosolic platinophilic ligands. The HSC viability studies with losartan-M6PHSA are in agreement with the safety data of other drug-M6PHSA conjugates prepared with the ULS linker.11, 25

An important finding of the current study is that oral losartan given for short periods of time, did not reduce established fibrosis. This is not surprising, since in the vast majority of studies in which losartan reduces the extent of liver fibrosis, losartan is given concomitantly with the agent causing liver injury, and for prolonged periods of time (i.e., several weeks).26, 27 This finding suggests that antifibrotic drugs may be not as active as expected when administered to rats with established fibrosis, which is in line with the poor clinical usefulness of many preclinical drug-candidates. Here, we demonstrate that the selective delivery of antifibrotic drugs to the main fibrogenic cell type in the liver (i.e., activated HSCs) markedly increased the antifibrotic effect.

Different mechanisms may explain the strong antifibrotic effect achieved with our drug-targeting construct. First, targeting losartan to activated HSCs via the modified albumin, M6PHSA, increases the fraction of the dose that accumulates within the fibrogenic cells. Since HSCs only represent a small fraction of the total liver, the drug levels found in liver homogenates may underestimate the actual accumulation of losartan-M6PHSA within HSCs. However, orally administered losartan resulted in higher hepatic concentrations due to the much higher dose, which however produced weaker antifibrogenic effects. Thus, the strong effects of losartan-M6PHSA cannot be attributed to an increase in drug concentrations within the liver, but to the selectivity of losartan to activated HSCs. Secondly, the activity of losartan-M6PHSA may be enhanced by the specific interaction that M6PHSA provides. The M6P/IGFII receptor participates in the activation of latent TGF-β1, which may be affected by M6PHSA.10 However, the finding that treatment with M6PHSA alone did not affect fibrosis or inflammation in bile duct-ligated rats does not support this hypothesis. Thirdly, we show that targeted losartan rapidly reduces the accumulation of activated HSCs in the fibrotic liver. This is consistent with previous reports showing that angiotensin II is a powerful mitogen for HSCs.28 And finally, targeted losartan strongly attenuated infiltration of inflammatory cells, a major pathogenic event in liver fibrogenesis.29 This latter effect is consistent with previous reports showing that Ang II exerts pro-inflammatory actions both in cultured cells and in vivo.6, 30 Although the cell type mediating the anti-inflamatory effect is unknown, activated HSCs are potential candidates. In fact, losartan-M6PHSA attenuated the inflammatory effects induced by angiotensin II on cultured HSCs. The beneficial effect of losartan-M6PHSA is not related with increased expression or activity of the collagenolytic enzymes MMP2, MMP3, and MMP9.

Our results may have implications for the treatment of chronic liver diseases. First, we provide evidence that short-term treatment with a highly active oral compound—losartan—is capable to attenuate the inflammatory response but it is not strong enough to reduce liver fibrosis. Therefore, the current assumption that RAS blockers are highly effective in attenuating experimental liver fibrosis should be tempered. Secondly, our results support the current research to develop innovative systems to deliver drugs to activated HSCs. This approach would be particularly useful in conditions with rapidly aggressive hepatic fibrosis (e.g., acute alcoholic hepatitis) in which the use of AT1 receptors blockers may induce undesirable side effects such as renal failure. Thirdly, our results suggest the possibility to use drugs known to block other pathogenic functions of activated HSCs, such as cell contractility and angiogenic effects. These pathogenic actions of activated HSCs could participate in the pathogenesis of portal hypertension and the progression of hepatocellular carcinoma, respectively.28, 31

Although the current study demonstrates that a short treatment of an antifibrotic drug to HSCs is able to reduce liver fibrosis, further studies should be performed to assess whether this strategy is also feasible for long periods of time. This aim includes initial pharmacodynamic studies to investigate the optimal route and dosage to ensure a stable and continuous release of the compounds to the fibrotic liver. We attempted to address this issue by giving losartan-M6PHSA for 3 weeks in rats with advanced fibrosis. This regime was able to reduce collagen synthesis but not the degree of fibrosis. This partial result can be explained by the lack of previous studies identifying the best regime for chronic administration of targeted drugs to HSCs. It is plausible that more frequent injections or the use of alternative routes (e.g., subcutaneous osmotic pumps) would have yielded positive results. We are currently performing complex pharmacological studies to address this issue.

Acknowledgements

We thank Anna Planagumà for kind help in animal handling and Elena Juez and Cristina Millán for excellent technical support. We also thank the Department of Pharmaceutical Analysis (University of Groningen) for the losartan-ULS mass spectrometry analysis, Jan Visser (Department of Pharmacokinetics and Drug Delivery) for assistance in HPLC analysis and the Unitat de Microscopia confocal (UB) for the analysis with the epifluorescence microscopy. Klaas Sjollema and Michel Meijer are also acknowledged for their kind assistance with the confocal pictures at the UMCG Microscopy and Imaging Center. Frank Opdam, Jack Veuskens, and Roel Schaapveld (Kreatech Biotechnology) are acknowledged for critical reading of the manuscript.

Ancillary