Mammalian target of rapamycin regulates vascular endothelial growth factor–dependent liver cyst growth in polycystin-2–defective mice

Authors

  • Carlo Spirli,

    1. Section of Digestive Diseases, Department of Internal Medicine, Yale University, New Haven, CT
    2. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    Search for more papers by this author
  • Stefano Okolicsanyi,

    1. Section of Digestive Diseases, Department of Internal Medicine, Yale University, New Haven, CT
    2. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    3. Department of Clinical Medicine and Prevention, University of Milano-Bicocca, Milan, Italy
    Search for more papers by this author
  • Romina Fiorotto,

    1. Section of Digestive Diseases, Department of Internal Medicine, Yale University, New Haven, CT
    2. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    Search for more papers by this author
  • Luca Fabris,

    1. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    Search for more papers by this author
  • Massimiliano Cadamuro,

    1. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    Search for more papers by this author
  • Silvia Lecchi,

    1. Section of Digestive Diseases, Department of Internal Medicine, Yale University, New Haven, CT
    2. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    3. Department of Clinical Medicine and Prevention, University of Milano-Bicocca, Milan, Italy
    Search for more papers by this author
  • Xin Tian,

    1. Section of Nephrology, Department of Internal Medicine, Yale University, New Haven, CT
    Search for more papers by this author
  • Stefan Somlo,

    1. Section of Nephrology, Department of Internal Medicine, Yale University, New Haven, CT
    Search for more papers by this author
  • Mario Strazzabosco

    Corresponding author
    1. Section of Digestive Diseases, Department of Internal Medicine, Yale University, New Haven, CT
    2. Center for Liver Research, Ospedali Riuniti di Bergamo, Bergamo, Italy
    3. Department of Clinical Medicine and Prevention, University of Milano-Bicocca, Milan, Italy
    • Section of Digestive Diseases, Department of Internal Medicine, Yale University School of Medicine, 333 Cedar Street, LMP 1080, New Haven, CT 06520
    Search for more papers by this author
    • fax: +1-203-785-7273.


  • Potential conflict of interest: Nothing to report.

Abstract

Polycystic liver disease may complicate autosomal dominant polycystic kidney disease (ADPKD), a disease caused by mutations in polycystins, which are proteins that regulate signaling, morphogenesis, and differentiation in epithelial cells. The cystic biliary epithelium [liver cystic epithelium (LCE)] secretes vascular endothelial growth factor (VEGF), which promotes liver cyst growth via autocrine and paracrine mechanisms. The expression of insulin-like growth factor 1 (IGF1), insulin-like growth factor 1 receptor (IGF1R), and phosphorylated mammalian target of rapamycin (p-mTOR) and the protein kinase A (PKA)–dependent phosphorylation of extracellular signal-regulated kinase 1/2 (ERK1/2) are also up-regulated in LCE. We have hypothesized that mammalian target of rapamycin (mTOR) represents a common pathway for the regulation of hypoxia-inducible factor 1 alpha (HIF1α)–dependent VEGF secretion by IGF1 and ERK1/2. Conditional polycystin-2–knockout (Pkd2KO) mice were used for in vivo studies and to isolate cystic cholangiocytes [liver cystic epithelial cells (LCECs)]. The expression of p-mTOR, VEGF, cleaved caspase 3 (CC3), proliferating cell nuclear antigen (PCNA), IGF1, IGF1R, phosphorylated extracellular signal-regulated kinase, p-P70S6K, HIF1α, and VEGF in LCE, LCECs, and wild-type cholangiocytes was studied with immunohistochemistry, western blotting, or enzyme-linked immunosorbent assays. The cystic area was measured by computer-assisted morphometry of pancytokeratin-stained sections. Cell proliferation in vitro was studied with 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium and bromodeoxyuridine assays. The treatment of Pkd2KO mice with the mTOR inhibitor rapamycin significantly reduced the liver cyst area, liver/body weight ratio, pericystic microvascular density, and PCNA expression while increasing expression of CC3. Rapamycin inhibited IGF1-stimulated HIF1α accumulation and VEGF secretion in LCECs. IGF1-stimulated LCEC proliferation was inhibited by rapamycin and SU5416 (a vascular endothelial growth factor receptor 2 inhibitor). Phosphorylation of the mTOR-dependent kinase P70S6K was significantly reduced by PKA inhibitor 14-22 amide and by the mitogen signal-regulated kinase inhibitor U1026. Conclusion: These data demonstrate that PKA-dependent up-regulation of mTOR has a central role in the proliferative, antiapoptotic, and pro-angiogenic effects of IGF1 and VEGF in polycystin-2–defective mice. This study also highlights a mechanistic link between PKA, ERK, mTOR, and HIF1α-mediated VEGF secretion and provides a proof of concept for the potential use of mTOR inhibitors in ADPKD and conditions with aberrant cholangiocyte proliferation. (HEPATOLOGY 2010.)

Polycystic liver disease may complicate autosomal dominant polycystic kidney disease (ADPKD).1 Liver cysts originating from the biliary epithelium progressively become larger and eventually cause complications related to mass effects, hemorrhages, infection, or rupture. Some patients may require cyst fenestration, liver resection, and even liver transplantation.

ADPKD is caused by mutations in the polycystic kidney disease 1 gene or polycystic kidney disease 2 gene.2 These encode for polycystin-1 (PC1) and polycystin-2 (PC2), respectively, two proteins involved in epithelial cell proliferation, morphogenesis, and differentiation.2 In the liver, PC1 and PC2 are expressed by biliary epithelial cells and are located in the primary cilium of the cell. PC2, a nonselective membrane Ca2+ channel, is also located in the endoplasmic reticulum (ER), in which it contributes to Ca2+ regulation.3, 4

Cholangiocytes lining liver cysts in patients with ADPKD show phenotypic and functional peculiarities, such as marked overexpression of vascular endothelial growth factor A (VEGF-A), angiopoietins, insulin-like growth factor 1 (IGF1), and their cognate receptors vascular endothelial growth factor receptor 2 (VEGFR2), TEK tyrosine kinase endothelial (Tie)-2, and insulin-like growth factor 1 receptor (IGF1R).5, 6 Using mice with conditional defects of PC2, we have shown that the cystic epithelium of PC2-defective mice overexpresses VEGF, which in turn stimulates cholangiocyte proliferation, and that the blockade of VEGFR2 signaling inhibits liver cyst growth in vivo.7 We have also shown that protein kinase A (PKA)–mediated up-regulation of ERK1/2 sustains the increased secretion of VEGF as well as its proliferative effects.

IGF1 is concentrated in the fluid drained from liver cysts of ADPKD patients, and IGF1 and its main receptor IGF1R,8 along with phosphorylated protein kinase B (pAKT) and phosphorylated mammalian target of rapamycin (p-mTOR), are overexpressed in the cystic epithelium of human ADPKD.5 Protein kinase B (AKT) is an inhibitor of tuberin and thereby an activator of mammalian target of rapamycin (mTOR), through which it controls hypoxia-inducible factor 1 alpha (HIF1α), the major transcriptional factor for VEGF.9 Through similar mechanisms, IGF1 up-regulates the expression of VEGF in colon cancer cells.10

Expression of mTOR is increased in the cystic epithelium of the kidney cells,11 and this suggests that the mTOR pathway may play a crucial role in the growth of kidney cysts. Rapamycin is an mTOR inhibitor that has been approved for use in humans because of its immunosuppressive properties. Rapamycin slows renal cyst development and renal function deterioration in rodent models of polycystic kidney disease.11–13 A recent retrospective study showed a reduction in liver cyst volume in ADPKD patients who received sirolimus as immunosuppressive therapy after kidney transplantation.14

Recent experimental work has shown that the mTOR inhibitor rapamycin is a potent inhibitor of angiogenesis and proliferation of endothelial, cancer, and stromal cells.15, 16 We hypothesized that inhibition of mTOR by rapamycin could block the progression of polycystic liver disease by interfering with IGF1 and VEGF autocrine signaling. We tested this hypothesis in conditional polycystin-2–knockout (Pkd2KO) mice, which overexpress VEGF, VEGFR2, and IGF1 in a manner similar to human ADPKD.

Abbreviations

ADPKD, autosomal dominant polycystic kidney disease; AKT, protein kinase B; BrdU, bromodeoxyuridine; CC3, cleaved caspase 3; CK, cytokeratin; DMOG, dimethyloxaloyl glycine; ELISA, enzyme-linked immunosorbent assay; ER, endoplasmic reticulum; ERK, extracellular signal-regulated kinase; HIF, hypoxia-inducible factor; IGF, insulin-like growth factor; IGF1R, insulin-like growth factor 1 receptor; LCE, liver cystic epithelium; LCEC, liver cystic epithelial cell; MEK, mitogen signal-regulated kinase; mTOR, mammalian target of rapamycin; MTS, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium; p-mTOR, phosphorylated mammalian target of rapamycin; pAKT, phosphorylated protein kinase B; panCK, pancytokeratin; PC1, polycystin-1; PC2, polycystin-2; PCNA, proliferating cell nuclear antigen; pERK, phosphorylated extracellular signal-regulated kinase; PI3K, phosphoinositol 3 kinase; PKA, protein kinase A; Pkd2KO, polycystin-2–knockout; PKI, protein kinase A inhibitor 14-22 amide; TSC2, tumor suppressor complex 2; VEGF, vascular endothelial growth factor; VEGFR2, vascular endothelial growth factor receptor 2; WT, wild type.

Material and Methods

Reagents.

All reagents were purchased from Sigma Chemical Co. (St. Louis, MO) unless otherwise indicated. Further information can be found in the supporting information.

Animals and Experimental Protocol.

The study was conducted with normal wild-type (WT) mice and ADPKD mouse models generated by an inducible defect in PC2 (Pkd2flox/−:pCxCreERTM), which was targeted through a Cre system and fused to the ligand-binding domain of a mutated estrogen receptor as previously described17 (S.S., unpublished data, 2009). Both males and females were used. Mice that were 28 days old were treated with tamoxifen (0.2 mg/g/day) for 5 days to induce the PC2 gene excision. These mice developed a liver phenotype resembling human ADPKD. To inhibit mTOR signaling, the animals were treated for 8 weeks with rapamycin (LC Laboratories, Woburn, MA; 1.5 mg/kg/day intraperitoneally; n = 10) or with the vehicle dimethyl sulfoxide (n = 10); the treatment was started in their fifth week. All experiments were performed according to protocols approved by the Institutional Animal Care and Use Committee of Yale University.

Immunohistochemical Studies.

Paraffin-fixed liver sections (5 μm thick) were deparaffinized and stained with hematoxylin and eosin. A pancytokeratin (panCK) antibody (56-and 64-kDa keratins, 1:300; Dako, Carpinteria, CA) was used to identify the biliary cysts. To detect the antigen of interest, serial liver tissue sections were immunostained as described.6 The source of antibodies is detailed in the online supporting information. For all immunoreactions, negative controls were also included and showed no staining.

Morphometric Analysis of the Cystic Area and PanCK-Positive Structures.

Serial sections of the two main liver lobes, treated as previously described, were mounted onto 0.1% poly(L-lysine)-coated glass slides and immunostained with a panCK antibody to allow the correct discrimination of the biliary cyst structures from the vessels. The relative area covered by the biliary cysts (five random, nonoverlapping fields per main liver lobe) was recorded with a digital camera at a ×10 magnification (for a total of 10 fields per mouse). In each field, the cystic areas were measured by two investigators blinded to the treatment modality with ImageJ software (National Institutes of Health, Bethesda, MD).7

On the same liver sections, we also calculated the percentage of the whole liver lobe area occupied by panCK-positive cells with a motorized stage system able to scan the whole liver lobes. Images taken at a ×4 magnification were analyzed with Metamorph software (Molecular Devices, Downington, PA; see the supporting information).

Quantification of Proliferation, Apoptosis, and Phosphorylated Extracellular Signal-Regulated Kinase (pERK)-Positive Structures in Liver Tissue Sections.

Liver sections were immunostained with an anti–proliferating cell nuclear antigen (anti-PCNA) antibody (1:200; Santa Cruz Biotechnology, Santa Cruz, CA) to measure the percentage of cystic cholangiocytes entering the cell cycle. Immunodetection of the cleaved form of caspase 3 [cleaved caspase 3 (CC3); 1:50; R&D Systems, Minneapolis, MN] was used to detect cells undergoing apoptosis, and immunodetection of pERK (1:100; Cell Signaling Technology, Danvers, MA) was used to assess the activation of the ERK pathway. The amount of pERK and CC3-positive structures was estimated by computer-assisted morphometric analysis as described previously.

Isolation and Characterization of Cholangiocytes.

Mouse cholangiocytes and cystic epithelial cells were isolated and cultured from WT and Pkd2KO mice essentially as previously described.6, 7 Microdissected intrahepatic bile ducts were used to obtain WT cholangiocytes, whereas in the case of conditional knockout mice, cells were isolated from microdissected cysts as described.6, 7 WT and Pkd2KO mouse cholangiocytes were maintained in 25-cm2 tissue culture flasks in a medium enriched with 10% fetal bovine serum at 37°C in a humidified, 5% CO2 atmosphere (for a detailed characterization of the cultured cells, see the online supporting information). The biliary phenotype and maintenance of the normal polarity were confirmed via staining for cytokeratin-19 (CK-19) and acetylated α-tubulin, by the measurement of transepithelial resistance, and by electron microscopy as described.7

Determination of HIF1α in Cultured Cells.

Cells were incubated in the presence of the prolyl 4-hydroxylase inhibitor, the 2-oxoglutarate analogue dimethyloxaloyl glycine (DMOG; 3 mM, 18 hours), or IGF1 (10 ng/mL) with or without the phosphoinositol 3 kinase (PI3K) inhibitor LY294002 (10 μM with a 10-minute pretreatment) or rapamycin (10nM with a 10-minute pretreatment), and they were compared with control cells. The nuclear fraction of each sample was isolated with a nuclear extraction kit (NE-PER, Pierce Biotechnology, Rockford, IL; for the purity of the nuclear extract, see Supporting Fig. 1). The concentration of protein was determined by the Bradford method (Pierce Biotechnology, Rockford, IL). The amount of HIF1α was measured with an HIF1α kit (R&D Systems, Minneapolis, MN) by the Duoset enzyme-linked immunosorbent assay (ELISA) according to the manufacturer's protocol. The amount of HIF1α was then normalized to the amount of nuclear protein.

Measurement of VEGF Secretion.

An ELISA assay (Biosource International) was used to quantify VEGF in a culture medium collected from cholangiocytes isolated from polycystic and control mice, as we previously described.7 A VEGF standard curve was generated for each individual experiment. Readings were normalized for the total protein in the well.

Determination of Cell Proliferation.

Cells were plated into 96-multiwell plates (5000 cells/well) and serum-starved.7 After 24 hours, cells were supplemented with IGF1 (10 ng/mL) alone and with rapamycin (10nM) or a competitive VEGFR2 inhibitor, SU5416 (5 μM), as shown in the Results section. Cell proliferation was measured with (1) CellTiter 96 AQueous One Solution (Promega Italia, Milan, Italy), which exploits 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) compound colorimetric bioreduction by the cells, and (2) the Biotrak ELISA system (GE Healthcare, Piscataway NJ), which measures the incorporation of the pyrimidine analogue 5-bromo-2′-deoxyuridine during DNA synthesis in proliferating cells.

Western Blots.

Methodological details of western blots can be found in the online supporting information.

Microvascular Density.

To study the changes in pericystic microvascular density induced by treatment with rapamycin, liver sections were stained with rat anti-CD3418 and counterstained with panCK.7 The biliary and vascular areas were calculated as reported in the supporting information.

Statistical Analysis.

Results are shown as means and standard deviations. Statistical comparisons were made with a one-way analysis of variance or the Wilcoxon-Mann-Whitney two-sample rank-sum test, as appropriate. In the latter, the P value was obtained from the exact permutation null distribution. The statistical analysis was performed with SAS software (SAS, Cary, NC). P values < 0.05 were considered significant.

Results

p-mTOR, IGF1, and IGF1R Are Overexpressed in Pkd2KO Cystic Cholangiocytes.

Pkd2KO mice developed a liver phenotype similar to human ADPKD.7 VEGF, p-mTOR (the phosphorylated, active form of mTOR), IGF1, and its receptor IGF1R were expressed in the cystic epithelium (n = 3) by immunohistochemistry (Fig. 1). These findings are consistent with previous reports showing overexpression of mTOR, VEGF, VEGFR2, IGF1, and IGF1R in liver cysts of patients with ADPKD5, 6 and establish that the Pkd2KO mouse is an adequate model for studying the role of mTOR, VEGF, and IGF1 in liver cyst growth.

Figure 1.

Expression of p-mTOR, IGF1, IGF1R, and VEGF in cystic cholangiocytes of Pkd2KO mice. Paraffin-embedded liver sections (5 μM) of WT and Pkd2flox/−:pCxCreERTM (Pkd2KO) mice were labeled with specific antibodies against the phosphorylated form of mTOR, IGF1, IGF1R, and VEGF. Immunoreactivity for p-mTOR, IGF1, IGF1R, and VEGF (arrowhead) was up-regulated on the biliary epithelium in the Pkd2KO mice with respect to the control mice (arrow). The magnification was ×40.

Rapamycin Inhibits Liver Cyst Growth in Conditional PC2-Knockout Mice.

To understand the pathophysiological relevance of increased p-mTOR expression, we treated Pkd2KO mice with the mTOR inhibitor rapamycin. Preliminary experiments using rapamycin at the dose of 5 mg/kg/day11 encountered significant toxicity (two of three mice died before completing the 8-week treatment). The daily dose of 1.5 mg/kg intraperitoneally for 8 weeks11, 13 was well tolerated without mortality or liver toxicity (Supporting Table 1). After 8 weeks of treatment, mice were sacrificed. PanCK staining revealed a significant reduction in the cystic area in rapamycin-treated mice in comparison with untreated Pkd2KO mice (cystic area: 27,027 ± 10,810 μM2 in Pkd2KO control mice versus 10,810 ± 4324 μM2 in treated mice, n = 10, P < 0.001; Fig. 2). Similarly, the percentage of the total area of the lobe covered by CK-positive structures was lower in rapamycin-treated mice (4.5% ± 1.3% in Pkd2KO control mice versus 2.2% ± 1.1% in rapamycin-treated mice, n = 10, P < 0.001; not shown). As a result of the reduction in liver cysts, treatment with rapamycin decreased the liver weight/body weight ratio of Pkd2KO mice (0.056 ± 0.008 in rapamycin-treated mice versus 0.0777 ± 0.016 in untreated Pkd2KO mice, n = 10, P < 0.01; liver weight/body weight ratio = 0.039± 0.002 in WT mice used as controls, n = 6; Fig. 2 and Supporting Fig. 2). The cystic area was greater in female mice; both females and males responded to rapamycin treatment (Supporting Fig. 3). Furthermore, confirming that rapamycin treatment was active, we found a significant reduction in P70S6K activation (Supporting Fig. 4). Furthermore, in agreement with an inhibitory effect on angiogenic factors, the pericystic microvascular density (Fig. 3A) and VEGF (Supporting Fig. 4) were significantly reduced.

Figure 2.

Rapamycin reduced the cystic area and liver weight/body weight percentage in Pkd2KO mice. (A) The micrographs are representative of vehicle-treated mice (left) and rapamycin (1.5 mg/kg/day)-treated mice (right). (B) As shown in the bar graph, a significant reduction in the cystic area was observed in Pkd2KO-treated animals (°P < 0.001, n = 10). (C) The decrease in cyst growth was also reflected in the significant reduction of the liver weight/body weight ratio. In fact, the liver weight/body weight ratio was higher in mice treated with the vehicle (n = 10) versus WT mice (**P < 0.001, n = 6), and it was significantly reduced in rapamycin-treated mice (*P < 0.01, n = 10).

Figure 3.

Rapamycin reduced microvascular density (CD34) and cell proliferation (PCNA) while increasing apoptosis (CC3 expression) in Pkd2KO mice. Each pair of micrographs is representative of vehicle-treated mice (top) and rapamycin-treated mice (bottom). For the quantitative analysis, five random, nonoverlapping fields per slide at a magnification of ×40 were recorded with a digital camera connected to Spot Advanced imaging software (version 3.5) by an observer who was blinded to the treatment modality. (A) Rapamycin treatment reduced microvascular density in Pkd2KO mice. Paraffin-embedded liver sections (5 μM) were stained with anti-CD34 and wide-spectrum anti-cow CK (panCK). To calculate the vascular and biliary areas, two different thresholds were set for CD34-positive (red fluorescence) and panCK-positive (green fluorescence) structures, and then they were expressed as the percentages of pixels above the threshold per field. As shown in the bar graphs, treatment with rapamycin significantly reduced the microvascular density (#P < 0.05 versus the vehicle;). (B) Cystic cholangiocytes showed strong proliferative activity (PCNA staining). As shown in the bar graph, a significant reduction in PCNA expression, as assessed by morphometric analysis, was observed in Pkd2KO-treated animals (*P < 0.001, n = 10). On average, 1000 nuclei were counted per mouse, and the percentage of PCNA-positive nuclei was then calculated. Only strongly positive immunostained nuclei were considered PCNA-positive. (C) To account for the decrease in liver cysts in rapamycin-treated mice, we quantified the amount of apoptosis (CC3 staining) in cystic cells. The bar graph shows a significant increase in CC3 expression in Pkd2KO-treated animals (*P < 0.01, n = 10). The CC3-positive area was expressed as a percentage of the CK-positive area.

Rapamycin Treatment Affects the Cell Proliferation/Apoptosis Rate in Pkd2KO Mice.

The proliferative activity of cystic cholangiocytes increased with respect to controls.7 In this study, we found that, after 8 weeks of treatment, PCNA immunostaining was lower in rapamycin-treated mice (21.4% ± 6.1% of cyst nuclei were positive for PCNA) versus vehicle-treated Pkd2KO mice (44% ± 20.6% were PCNA-positive; n = 10, P < 0.01), and this suggests that rapamycin reduces the proliferation of liver cyst cells (Fig. 3B).

Given the role of mTOR in cell survival, we hypothesized that rapamycin could also increase apoptosis in cystic cholangiocytes. With computer-assisted morphometry, apoptosis in vivo was analyzed from the immunohistochemical expression of the cleaved, activated form of caspase 3 (CC3) in the same liver specimens used for PCNA staining. As shown in Fig. 3C, the CC3-positive area was 18.7% ± 10.3% of the CK-positive area in rapamycin-treated mice versus 9.8% ± 5.4% in untreated mice (n = 10, P < 0.05). These data suggest that mTOR inhibition reduces cyst growth through the combined reduction of proliferation and the increase in the apoptosis rate in the cystic epithelium.

Rapamycin Inhibits IGF1-Induced HIF1α Accumulation and VEGF Secretion in Cultured Cholangiocytes.

We have previously shown that VEGF and VEGFR2 regulate cyst growth and cholangiocyte proliferation in Pkd2KO mice.7 Therefore, we studied the effect of rapamycin on VEGF secretion and on the nuclear expression of its main transcription factor, HIF1α, in cystic cholangiocytes cultured from Pkd2KO mice (n = 8 isolations) and in cholangiocytes cultured from WT mice (n = 5 isolations). Preliminary experiments showed that rapamycin significantly inhibited HIF1α and VEGF after the administration of 3mM DMOG, an inhibitor that blocks prolyl 4-hydroxylase–dependent HIF1α degradation, and this indicates that mTOR controls HIF1α-dependent VEGF secretion in cystic cholangiocytes (not shown). We next studied the effects of IGF1 (10 ng/mL for 18 hours) in the presence or absence of rapamycin. As shown in Fig. 4A and Supporting Table 2, nuclear expression of HIF1α was significantly higher in Pkd2KO cells than in WT cells both at the baseline (1875 ± 605 versus 625 ± 286 pg/mg of protein in WT cells, n = 15, P < 0.001) and after IGF1 administration (3598 ± 860 versus 1167 ± 959 pg/mg in WT cells after IGF1 stimulation, n = 7, P < 0.001). Rapamycin significantly inhibited IGF1-induced HIF1α accumulation in Pkd2KO cholangiocytes (1516 ± 288 ng/mg, n = 4, P < 0.001). Similarly, VEGF released into the culture medium was significantly higher in Pkd2KO cholangiocytes than in WT cholangiocytes both at the baseline (1440 ± 52 ng/mg of protein in cells versus 596 ± 167 ng/mg in WT cells, n = 15, P < 0.001) and after IGF1 administration (2381 ± 997 ng/mg in Pkd2KO cells versus 665 ± 205 ng/mg in WT cells after IGF1 stimulation, n = 7, P < 0.001). In Pkd2KO cholangiocytes, rapamycin significantly decreased VEGF secretion after stimulation with IGF1 (1368 ± 462 ng/mg, n = 4, P < 0.05). These results indicate that IGF1 stimulates HIF1α and VEGF via the mTOR pathway and that inhibition of mTOR by rapamycin inhibits HIF1α-dependent VEGF secretion in cystic cholangiocytes.

Figure 4.

Rapamycin and LY294002 inhibited IGF1-induced HIF1α accumulation and VEGF secretion. The effect of rapamycin on VEGF secretion and on the nuclear expression of its main transcription factor, HIF1α, in primary cultures of cystic cholangiocytes cultured from Pkd2KO mice was assessed with ELISA. (A) In cultured Pkd2KO cholangiocytes, HIF1α accumulation and VEGF secretion induced by IGF1 were significantly higher with respect to WT cholangiocytes. (B) This effect was completely blunted in cells treated with rapamycin (5 μM; n = 4) and in cells treated with the PI3K inhibitor LY204002 (10 μM; n = 3). #P < 0.005 versus the WT control, ∧P < 0.05 versus the WT control, · P < 0.05 versus WT+IGF1, *P < 0.001 versus the Pkd2KO control, and **P < 0.001 versus Pkd2KO+IGF1.

PI3K Inhibition Blocks IGF1-Induced HIF1α Accumulation and VEGF Secretion.

To better understand the relationship between the PI3K/pAKT/mTOR pathway and VEGF production in cystic cholangiocytes, we studied the effects of the PI3K inhibitor LY294002 (10 μM) on HIF1α nuclear expression and on VEGF secretion in the presence of IGF1 (10 ng/mL). Figure 4B and Supporting Table 3 show that both HIF1α production and VEGF production were significantly reduced in Pkd2KO cholangiocytes treated with LY294002 (HIF1α: 4483 ± 586 pg/mg of protein after IGF1 stimulation versus 1589 ± 95 pg/mg after IGF1 and LY24002 treatment, n = 3, P < 0.001; VEGF: 4629 ± 304 ng/mg of protein after IGF1 stimulation versus 1838 ± 313 ng/mg after IGF1 and LY24002 treatment, n = 3, P < 0.01). This is consistent with the hypothesis that IGF1 stimulates HIF1α-dependent VEGF production through the PI3K/AKT/mTOR pathway.

IGF1-Induced Cell Proliferation in Cystic Cholangiocytes Is Mediated by mTOR and Inhibited by the Blockade of VEGF Autocrine Signaling.

We previously showed that administration of VEGF increased proliferation of cystic cholangiocytes through VEGFR2 stimulation.7 IGF1 (10 ng/mL) increased cell proliferation, as assessed by the MTS assay and by bromodeoxyuridine (BrdU) incorporation, in both normal and cystic cholangiocytes (Fig. 5). Administration of rapamycin (10 nM) to IGF1-treated Pkd2KO and WT cholangiocytes significantly inhibited IGF1-induced cell proliferation, as measured by BrdU incorporation and the MTS assay (Fig. 5). These data indicate that the PI3K/AKT/mTOR pathway mediates IGF1R signaling in cholangiocytes and, to a larger extent, in cystic cholangiocytes. Interestingly, the VEGFR2 inhibitor SU5416 (5 μM) significantly decreased IGF1-induced proliferation of cystic cholangiocytes by 50%, and this suggests that IGF1 proliferative effects in cholangiocytes may be in part mediated through the increased secretion of VEGF.

Figure 5.

Rapamycin and SU5614 inhibited IGF1-induced cell proliferation in cultured Pkd2KO cystic cholangiocytes. With two different cell proliferation assays, similar results were obtained: (A) the MTS assay results and (B) BrdU incorporation. In both WT and Pkd2KO cholangiocytes, IGF1 significantly enhanced cell proliferation in comparison with untreated cells (*P < 0.05 with respect to WT control cholangiocytes and · P < 0.01 with respect to control Pkd2KO cholangiocytes). The increase in cell proliferation stimulated by Pkd2KO cholangiocytes was significantly higher in Pkd2KO cholangiocytes (&P < 0.05 with respect to WT cholangiocytes exposed to IGF1). IGF1-induced cell proliferation was significantly inhibited by treatment with rapamycin (5 μM) or with SU5416, a VEGFR2 inhibitor (°P < 0.05 in WT cholangiocytes treated with IGF1 plus rapamycin or SU5416 with respect to IGF1-treated WT cholangiocytes and #P < 0.01 in Pkd2KO cholangiocytes treated with IGF1 plus rapamycin or SU5416 with respect to IGF1-treated Pkd2KO cholangiocytes, n = 3).

Crosstalk Between ERK1/2 and mTOR in Cystic Cholangiocytes.

Our data show that IGF1, through PI3K/pAKT/mTOR, increases HIF1α-dependent VEGF secretion and cell proliferation. A recent article suggested that polycystins control mTOR activity by inhibiting ERK.19 We have previously shown that PKA-mediated phosphorylation of pERK1/2 is increased in cystic cholangiocytes, and this correlates with increased secretion of VEGF and response to VEGFR2 stimulation.7 To better understand the relationships between mTOR activation and PKA-mediated phosphorylation of ERK in cystic cholangiocytes, we measured the phosphorylation of P70S6K, a kinase activated by mTOR, after inhibition of PKA with protein kinase A inhibitor 14-22 amide (PKI; 1 μM, n = 3) and after inhibition of the ERK pathway with the mitogen signal-regulated kinase (MEK) inhibitor U1026 (10 μM). As shown in Fig. 6, phosphorylation of P70S6K was increased in Pkd2KO cholangiocytes and was inhibited by PKI and by U1026, and this suggests that the PKA/ERK pathway activates mTOR.19

Figure 6.

Phosphorylation of mTOR-activated P70S6K was PKA-mediated and ERK-mediated. (A) In Pkd2KO cystic cholangiocytes, the phospho-P70S6K/P70S6K ratio was higher with respect to WT cells; it was completely inhibited by rapamycin, as expected, and was significantly inhibited by the PKA inhibitor PKI (*P < 0.05 with respect to untreated WT cholangiocytes, ∧P < 0.001 with respect to untreated Pkd2KO cholangiocytes, and #P < 0.01 with respect to untreated Pkd2KO cholangiocytes). (B) In Pkd2KO cholangiocytes, the phospho-P70S6K/P70S6K ratio was significantly inhibited by the MEK inhibitor U1026 (10 μM; *P < 0.05 with respect to untreated WT cholangiocytes and °P < 0.05 with respect to untreated Pkd2KO cholangiocytes).

Conversely, to determine if mTOR affects ERK1/2 activity, we studied pERK1/2 expression after administration of IGF1 with or without rapamycin or the VEGFR2 inhibitor SU5416. As shown in Fig. 7, IGF1-induced ERK1/2 phosphorylation was significantly inhibited by treatment with rapamycin (5 μM) and also by the VEGFR2 inhibitor SU5416 (the pERK/ERK ratio in control Pkd2KO cells was 1.21 ± 0.4 versus 2.1 ± 0.4 after IGF1 administration, P < 0.05). The pERK/ERK ratio was reduced to 1.34 ± 0.5 after IGF1 and rapamycin (P < 0.05, n = 5) and to 1.34 ± 0.5 after IGF1 and SU5416 (P < 0.05, n = 5). As shown in Supporting Fig. 5, SU5416 had no inhibitory effects on IGFR-1; therefore, these findings suggest that mTOR does not directly activate pERK1/2, but rather the increased secretion of VEGF caused by IGF1 via the mTOR pathway activates the MEK/ERK1/2 pathway downstream of VEGFR2.

Figure 7.

Rapamycin and SU5614 inhibited IGF1-induced ERK phosphorylation in cultured Pkd2KO cystic cholangiocytes. IGF1 significantly enhanced ERK phosphorylation. IGF1-induced ERK phosphorylation was significantly inhibited by treatment with rapamycin (5 μM) or with SU5416, a VEGFR2 inhibitor; the bar graph illustrates the pERK/ERK ratio as assessed by densitometry (°P < 0.05 with respect to WT cells, ∧P < 0.05 with respect to Pkd2KO controls, and *P < 0.05 with respect to Pkd2KO controls, n = 4).

Discussion

The progressive growth of liver cysts can cause significant morbidity in patients with ADPKD.1 Understanding the mechanisms by which liver cysts become larger may lead to novel treatment paradigms. Liver cysts are not connected to the biliary tree, and their growth is dependent on the autocrine effect of cytokines and growth factors produced by the cystic epithelium. Among these factors, VEGF and IGF1, along with their cognate receptors, are expressed by cystic cholangiocytes and are capable of autocrine stimulation of the liver cyst epithelium.5, 6

In this study, using mice with conditional inactivation of PC2, we have demonstrated that mTOR plays a central role in cyst growth. Furthermore, we have shown that IGF1 and VEGF signaling are linked through the PI3K/AKT/mTOR pathway, that there is significant crosstalk between mTOR and ERK1/2, and that the mTOR inhibitor rapamycin reduces the growth of liver cysts in vivo through the repression of VEGF secretion, with reduced cell proliferation and increased apoptosis.

m-TOR is a signaling molecule that integrates a broad spectrum of signals, including growth factors.20 Hormones and growth factors activate several downstream transduction pathways, which include the PI3K/AKT and Ras/MEK/ERK pathways. Both pathways converge to activate mTOR by inhibiting the activity of its negative regulator tuberin [more specifically tumor suppressor complex 2 (TSC2)].21 It has been shown that AKT and ERK may directly phosphorylate different serine residues on TSC2 and thereby inhibit its activity.22, 23 A number of functions modulated by mTOR are potentially relevant for liver cyst growth. Among them, mTOR stimulates HIF1α, a main transcription factor for VEGF.24

Rapamycin, an inhibitor of mTOR commonly used as an antirejection agent, has shown promising oncological applications because of its ability to promote chemotherapy-induced apoptosis and inhibit angiogenesis.16 Previous studies in animal models of polycystic kidney diseases nonorthologous to polycystin defects, such as the Han:Sprd rats25 and orpk and bpk mice,11 reported that treatment with rapamycin reduced kidney cysts and improved kidney function. Retrospective studies showed a reduction in kidney and liver cysts in patients with advanced-stage ADPKD who received a renal transplant and were treated with a rapamycin-containing antirejection regimen.14

We found that administration of rapamycin significantly decreased the cystic area of the liver and the liver/body weight ratio in Pkd2KO mice. At a daily dose of 1.5 mg/kg, rapamycin was well tolerated with no significant changes in liver function tests in comparison with untreated controls. Treatment with rapamycin decreased the PCNA index of liver cysts while increasing the expression of CC3, and this suggests that rapamycin alters the balance between proliferation and apoptosis by reducing the number of proliferating cells and enhancing cyst apoptosis in vivo.

Because of the role of VEGF in polycystic liver disease progression and the reported anti-angiogenic effects of rapamycin on cancer, we studied the effects of rapamycin on VEGF production in cystic cholangiocytes cultured from PC2-defective mice. We found that rapamycin suppressed the increased HIF1α nuclear expression and VEGF production typical of PC2-defective cells. This indicates that VEGF production in cystic cholangiocytes is controlled by mTOR and that the inhibitory effects of rapamycin on liver cysts could be explained in part by the inhibition of VEGF expression.

IGF1 is a cholangiocyte growth factor able to stimulate the PI3K/AKT pathway. IGF1 is overexpressed by the cystic epithelium and reaches a high concentration in the fluid of hepatic cysts in ADPKD patients.5 IGF1R is overexpressed in human cholangiopathies, including cholangiocarcinoma and human liver ADPKD.5, 26 Here we show that administration of IGF1 significantly increased HIF1α and VEGF in cystic cholangiocytes with respect to WT cholangiocytes. Stimulation of IGF1R is known to activate different common transduction pathways that modulate proliferation/survival.27 In agreement with earlier studies showing that IGF1 activates mainly the PI3K/AKT/mTOR pathway in cholangiocytes,5, 27, 28 we have found that IGF1-stimulated, HIF1α-dependent VEGF expression is inhibited by rapamycin. The significant inhibitory effect of the PI3K inhibitor LY294002 on IGF1-induced, HIF1α-dependent VEGF secretion is consistent with a major role of PI3K/AKT in mediating IGF1 signaling in cholangiocytes.

In agreement with the aforementioned data, IGF1 stimulated cell proliferation in PC2-defective cells (Fig. 5), and this effect was significantly inhibited by rapamycin. However, IGF1 stimulates secretion of VEGF, also a strong mitogen for cystic cholangiocytes. As shown in Fig. 5, IGF1-induced cell proliferation in cystic cholangiocytes was significantly inhibited by the VEGFR2 inhibitor SU5416. SU5416 did not affect the phosphorylation of the IGF1 receptor, in contrast to the specific IGF1R inhibitor AG102429 used as a positive control, and this indicates that the effects of SU5416 on VEGFR2 are specific (Supporting Fig. 5). These data strongly argue for the presence of an autocrine loop in which IGF1 stimulates PI3/AKT/mTOR and mTOR stimulates HIF1a-dependent secretion of VEGF, which in turn, interacting with its receptor VEGFR2, activates an MEK/ERK1/2-dependent proliferative effect. In agreement with this interpretation, both rapamycin and SU5416 inhibited the increase in pERK expression induced by IGF1 in cystic cholangiocytes. A further indication of this autocrine loop involving IGF, mTOR, and VEGF secretion is the strong reduction in pericystic microvascular density in mice treated with rapamycin (Fig. 3A).

An open question is the mechanistic relationships between the aforementioned mechanisms and the polycystin defect. Shillingford et al.11 and Distefano et al.19 proposed that PC1 acts as an inhibitor of TSC2; this mechanism would be lost in ADPKD, and increased activity of mTOR would result. These data represent an important clue; however, they have been generated by overexpression of PC1. In our study, we instead used a strategy involving the loss of function of PC2. In addition to its functional relationships with PC1, PC2 participates in the cellular and ER homeostasis of calcium.3, 4 Lower cellular and ER calcium stimulates PKA and ERK phosphorylation.30-33 We have recently shown that in PC2-defective cholangiocytes, the increase in pERK1/2 is PKA-dependent.7 We here provide evidence that baseline p-mTOR activation in PC2-defective cholangiocytes (as judged by its downstream kinase P70S6) is PKA-dependent and ERK-dependent and is thus linked to the altered calcium homeostasis.

We can only speculate about why the cystic epithelium produces this vast array of growth factors. The mechanisms are unclear, but this is akin to what happens to WT biliary epithelium during liver repair. Thus, cystic cholangiocytes resemble activated cholangiocytes in terms of their ability to generate autocrine and paracrine growth factors.34 This property may be a result of a relative lack of terminal differentiation, as we suggested earlier, or a response needed to cope with cellular stress. In this study, we did not explore the relationship between these mechanisms and estrogens; however they must be taken into account in the overall scenario, as also shown by the more severe phenotype of female mice. A number of clinical observations35 indicate that estrogens play a role in polycystic liver diseases. Estrogen receptor-β is up-regulated in liver cysts of ADPKD patients, and 17-β-estradiol stimulates the proliferation of cystic cholangiocytes obtained from patients with ADPKD; this has also shown that ADPKD epithelium is sensitive to the proliferative effects of estrogens and IGF1.5 Estrogens also promote the synthesis and release of growth factors, including IGF1, from the cyst epithelium.5

In conclusion, our study demonstrates that mTOR plays a central role in liver cyst growth in mice with defective PC2 (Fig. 8). The mTOR pathway regulates HIF1α-dependent VEGF secretion and appears central to the proliferative, antiapoptotic, and pro-angiogenic effects of IGF1, one of the major factors generated by the cystic epithelium. The mTOR inhibitor rapamycin inhibits VEGF secretion and signaling and significantly reduces liver cyst growth by reducing proliferation and increasing apoptosis of the cystic epithelium. This study also reveals a mechanistic link between mTOR and ERK and HIF1α-mediated VEGF secretion and provides a proof of concept for the potential use of mTOR inhibitors in polycystic liver disease and in conditions with aberrant cholangiocyte proliferation.

Figure 8.

Working model to explain cyst growth in PC2-defective cholangiocytes. The schematic summarizes the main findings of this study in the context of the most recent literature.7, 19 Cells with defective PC2 are characterized by increased PKA production and ERK1/2 phosphorylation.7 Erk 1/2 stimulates HIF1α-dependent VEGF secretion directly and by inhibiting tuberin, a negative regulator of mTOR.23 mTOR has a central role in IGF1-stimulated proliferation of cystic cholangiocytes. IGF1, a growth factor secreted by the cystic epithelium and by cholangiocyte under stress, binds to its receptor IGF1R and activates the PI3K/AKT/mTOR pathway; mTOR stimulates proliferation through cyclins and through an HIF1α/VEGF-dependent autocrine loop. Rapamycin is an mTOR inhibitor, LY294002 is a PI3K inhibitor, AG1024 is an IGF1R inhibitor, and SU5416 is a VEGFR2 inhibitor.

Ancillary

Advertisement