Main

Chronic liver injury caused by viral hepatitis, alcohol and drugs results at first in fibrosis, and if the injury persists, leads to liver cirrhosis. The main source of the scar tissue during liver fibrosis and cirrhosis is the hepatic stellate cells (HSC).1, 2 On activation, HSC undergo proliferation and differentiation,1, 2 becoming myofibroblast-like α-smooth muscle actin (αSMA)-positive cells that produce extracellular matrix proteins, particularly type I collagen.1, 2

We have previously shown that hypothyroidism, induced chemically or surgically, prevents liver cirrhosis in a thioacetamide (TAA) rat model of fibrosis3 and recent studies show that hypothyroidism also helps regression of established liver cirrhosis.4

Thyroid hormones, triiodothyronine (T3) and L-thyroxine (T4) have central roles in growth, differentiation and metabolism, acting mainly through transcriptional regulation of target genes.5 The main circulating thyroid hormone, T4, is deiodinated inside cells by various deiodinases and becomes T3. T3 enters the nucleus and binds to nuclear thyroid hormone receptors, bound to target genes containing thyroid response elements (TRE).5 The two major receptor isoforms, thyroid hormone receptor α1 and β1 (TRα1 and TRβ1), are encoded by two separate genes and have tissue-specific distribution, with liver expressing mainly TRβ1, and heart expressing TRα1.5 We have shown that an immortalized hepatic stellate cell line, HSC-T6, expresses TRα1, but not TRβ1, whereas primary hepatocytes express both.6 Recently, T4 was shown to bind and signal through cell-surface integrin αvβ3,7 inducing proliferation in various cells expressing high levels of αvβ3.8 The T4-induced signal transduction pathway includes phosphorylation of signal transducer and activator of transcription 1 (STAT1), STAT3 and mitogen-activated protein kinase (MAPK) and their translocation to the nucleus.8 Moreover, silencing of integrin αv β3 using siRNA, or blocking the signaling of this receptor, induces HSC apoptosis.9

Recent studies have shed light on the function of p75 neurotrophin receptor (p75NTR) in HSC activation and show that in p75NTR−/− mice, HSC are unable to become activated and to express αSMA.10 Interestingly, p75NTR is effective in activating HSC in the absence of its ligand, nerve growth factor (NGF), signaling through activation of guanosine triphosphatase Rho.10 Activation of members of Rho GTPase is reported to be crucial for HSC activation, as Rho inhibition attenuates liver fibrosis in several in vivo models.11, 12

Quiescent HSC do not express p75NTR, but start expressing it on activation,13 whereas hepatocytes do not express this receptor.14 Recent studies show that hyperthyroidism induces expression of p75NTR in normal liver.15

On the basis of our previous studies and on studies describing the role of p75NTR in HSC activation, we have investigated the connection between thyroid hormones and activation of primary HSC in vivo and in vitro, and delineated the mechanisms involved, including transcription of p75NTR and Rho activation, leading to increased expression of αSMA.

MATERIALS AND METHODS

Isolation of Primary Rat HSC

Livers of female rat Wistar (retired breeders) were perfused with Gey's balanced salt solution (GBSS), then GBSS with 100 mg collagenase (Worthington Biochemical Corporation, NJ, USA) and 130 mg pronase (Roche Diagnostics, Mannheim, Germany). The digested liver was minced and incubated with 100 ml of GBSS with 25 mg pronase, 25 mg collagenase, 20 μg/ml DNAse I for 30 min at 37°C with 200 r.p.m. shaking. HSC were isolated on a 17.5% Nycodenz gradient centrifuged at 1400 g for 20 min.9

Primary HSC Culture Conditions

Cells were cultured in DMEM with 10% fetal calf serum (FCS) and 100 μg/ml penicillin/streptomycin. To test effect of T3, we grown the cells for 24 h in serum-free DMEM with 10−7 M T3 (Sigma, St Louis, MI, USA). For other experiments, cells were grown in thyroid hormone-depleted serum medium (TDS), consisting of DMEM supplemented with 10% TDS. Depletion of T3 and T4 was obtained by overnight incubation of 2 g activated charcoal per 100 ml serum at 4°C, followed by centrifugation and filtration.16, 17

Proliferation Assays

Proliferation was by cell counting (using a Coulter counter), by a proliferation kit EZ4U based on formation of tetrazolium salts (Biomedica Medizinprodukte, Vienna, Austria) and by BrdU incorporation (Exalpha Biologicals Inc., Watertown, MA, USA).

Western Blots

Equal amounts of total protein were separated by SDS–PAGE and blotted onto Hybond C extra. Blots were incubated with antibodies for TRα1 and TRβ1 (Affinity Bioreagents, Golden, CO, USA), αv, αSMA, PDGF receptor β, pMAPK (erk1/erk2), erk2, akt 1/2, STAT3 (Santa Cruz Biotechnology, Santa Cruz, CA, USA), p75NTR (Upstate, Lake Placid, NY, USA), pSTAT3, pAkt and p-non-muscle myosin light chain II (pMLC; Cell Signaling Technology, Danvers, MA, USA), then incubated with horseradish-peroxidase-conjugated secondary antibody and detected by chemiluminescence. Expression was normalized to glyceraldehydes-3-phosphate dehydrogenase (GAPDH) or β-actin.

Immunohistochemistry

Hepatic stellate cells were washed with PBS and fixed in methanol: acetone (1:1) for 10 min at −20°C. Cells were blocked for 30 min in PBS with 10% FCS, incubated for 1 h with αSMA antibodies (1:100; Santa Cruz Biotechnology), washed and incubated with secondary anti mouse-Cy3 (Jackson Immunoresearch Laboratories, Bar Harbor, ME, USA).

Oil Red O Staining

Hepatic stellate cells were cultured for 1, 2, 4, 5 and 7 days in TDS with or without 10−7 M T3. Cells were fixed for 30 min in 10% formaldehyde and stained for 1 h with a 0.25% Oil Red O (Sigma) solution in propylene glycol.

Rho Activation Assays

Activated Rho was detected using a pull-down assay or an ELISA-based assay (Cytoskeleton, Denver, CO, USA). For the pull-down assay, 2, or 3 days and 5 days cultures of HSC were kept for 48 h in serum-free DMEM. Cells were pretreated for 30 min with 20 μl of a function-blocking antibody to β3 integrin (BioLegend, San Diego, CA, USA; clone 2C9.G2) or 500 nM RGD (arginine-glycine-aspartate) or 100 nM tetraiodothyroacetic acid (tetrac; Sigma), then T4 (10−7 M) or T3 (10−7 M) were added for 15 or 30 min and cells were scraped in a Mg2+ lysis buffer. Equal amounts of protein were incubated with rhotekin-agarose beads for 45 min at 4°C, beads were washed and Rho-GTP bound to the beads was separated by gel electrophoresis, blotted and membranes were incubated with an Rho monoclonal antibody (Upstate). For the active RhoA ELISA, 5-day HSC were kept for 48 h in serum-free DMEM, then treated for 15 min with 10−7 M T3 or T4 and 100 nM tetrac. Cell lysates were assayed according to the manufacturer's instructions.

Transient Transfections and Reporter Gene Assays

The plasmid containing the −764 bp αSMA promoter-luciferase was from Dr S Phan (University of Michigan, MI, USA). The plasmids containing the −3724/+13 human p75NTR promoter-luciferase and a shorter promoter region, −207/+13, were from Dr P. Barker (McGill University, Montreal, Canada).

Hepatic stellate cells were transfected with 2 μg of the promoter of interest and 0.3 μg β-galactosidase expression vector with Gene Jammer reagent (Stratagene, La Jolla, CA, USA) and treated for 24 h in OptiMEM (Gibco-BRL) with 10−7 M T3 and T4, 2.5 nM thyroid hormone receptor antagonist 1-850 (Calbiochem, Darmstadt, Germany), 10 μM Rho-kinase (ROCK) inhibitor Y-27632 (AG Scientific, San Diego, CA, USA) and 100 nM tetrac. Luciferase activity was determined using the Luciferase Assay Kit (Applied Biosystems, Bedford, MA, USA) and was normalized to β-gal activity.

RT–PCR for Cytokines

Total RNA extracted using an EZ-RNA kit (Biological Industries, Bet Haemek, Israel) was DNAse-treated and reverse-transcribed. The cDNAs were used for PCR, using the following primers: tumor necrosis factor-α (TNFα): 5′-ggagtggatcaggctgtctc-3′, 5′-ccagatggggatagctggta-3′; transforming growth factor-β (TGFβ1): 5′-tacagggctttcgcttcagt-3′, 5′-tggttgtagagggcaaggac-3′; stromal-derived factor 1α (SDF-1α): 5′-gagtcagacgcctgaggaac-3′, 5′-atccttgggggtaggaacac-3′; monocyte chemoattractant protein 1(MCP-1): 5′-tagcatccacgtgctgtctc-3′, 5′-tgaggtggttgtggaaaaga-3′; p75NTR: 5′-tacagtggcggatatggtga-3′,5′-ctctacctcctcacgcttgg-3′; GAPDH: 5′-agacagccgcatcttcttgt-3′, 5′-cttgccgtgggtagagtcat-3′. The PCR reaction contained 2 μl of cDNA reaction, 0.4 μM of each primer, 200 μM dNTP, 10 mM Tris-HCl (pH 8.3), 50 mM KCl, 1.1 mM MgCl2, 0.01% gelatin, 2.5 U RED TaqDNA polymerase (Sigma). After denaturation for 10 min at 94°C, the PCR was started using 32 cycles at 94°C for 50 s, 60°C for 50 s and 72°C for 50 s, with a final 5 min elongation step at 72°C.

In Vivo Induction of Early Fibrosis and Hyperthyroidism

Male Wistar rats (300 g) were injected i.p. with 150 mg/Kg body weight TAA twice a week for 3 weeks. The groups of animals were: (1) control group receiving TAA (n=5); (2) TAA and T3 10 μg per rat injected s.c. five times a week (n=6); (3) TAA and T4 (Eltroxin) in the drinking water (10 mg/l) (n=6).

Statistical Analysis

Results are presented as fold inductions compared to control values considered 100% and are averages +s.d. of at least three separate experiments. Statistical significance was assessed using an unpaired Student's t-test with P-value <0.05 considered significant.

RESULTS

Expression of TRα1 Decreases and αv Increases with HSC Transdifferentiation into Myofibroblasts

Primary HSC undergo spontaneous, time-dependent activation in culture, with 2- to 4-day cultures considered quiescent, 7-day cultures becoming activated and 14-day cultured considered fully activated. We were interested to assess expression of T3 nuclear receptor TRα1, as well as expression of the newly described T4 receptor integrinαvβ3, as a function of HSC activation. TRα1 expression in primary cultures of HSC has been assessed using TRα1-specific antibodies (not cross-reactive with TRα2 or TRβ1). Expression of TRα1 is high in 4-day (quiescent) and 7-day (activated) HSC, and dramatically decreases to only 30% in 14-day myofibroblasts (Figure 1a and b). TRβ1 has not been detected in HSC by TRβ1-specific antibodies (not cross-reactive with TRα1 or TRα2; data not shown). The pattern of expression of αv and PDGF receptor-β is the opposite of TRα1 during in vitro activation of HSC, with low expression of αv and PDGF receptor-β in quiescent cells and high expression in fully activated 14-day HSC (Figure 1a and b).

Figure 1
figure 1

Expression of TRα1 is reduced in myofibroblasts. (a) Western blot showing TRα1, αV and PDGF receptor β in 4, 7 and 14-day HSC. (b) Means of fold-induction +s.d. of five westerns blot experiments with TRα1 normalized to GAPDH. *P=0.000029 vs 4 days. (c) TRα1 determined by western blot in HSC isolated from 3 weeks TAA-treated rats and compared with TRα1 in 4-day HSC cultures from normal rats (N). Results are means of fold-induction from three experiments +s.d. *P<0.03.

Full size image

We have tested whether in vivo activation of HSC also causes TRα1 reduction. Activation of HSC in vivo was obtained by 3 weeks injection of TAA, then HSC were isolated and tested for expression of TRα1 after 4, 7 and 14 days in culture. TRα1 is strongly reduced in HSC from TAA-treated rats compared with expression of TRα1 in 4-day quiescent HSC from normal rats (Figure 1c).

Thyroid Hormones do not Affect HSC Proliferation

T4 has been shown to induce proliferation of various cell types acting through integrin αvβ3.8 The effect of T3 or T4 on proliferation of 7 and 14-day HSC has been determined by cell counting, by a proliferation kit based on formation of tetrazolium salts and by BrdU incorporation. T3 and T4 have no effect on proliferation tested by any of these methods (Figure 2a–c).

Figure 2
figure 2

T3 and T4 do not affect HSC proliferation. (a) Proliferation of HSC determined by cell counting. Cells were serum-starved overnight in 0.5% FCS and grown with 10−7 M T3 or 10−7 M T4 for 48 h. Results are means of fold-induction +s.d. of five separate experiments in triplicates. (b) Proliferation determined by EZ4U kit. HSC were replated, serum-starved overnight in 0.5% FCS and cultured for 48 h with 10−11, 10−9 and 10−7 M T3 or 10−7 M T4. Results are means of fold-induction +s.d. of three experiments in triplicates. (c) Proliferation determined by BrdU incorporation. HSC were replated, serum-starved overnight in 0.5% FCS and cultured for 24 h with 10−7 M T3 or T4. BrdU was added for 4 h and its incorporation determined using an ELISA assay. Results are means of fold-induction +s.d. of three experiments in triplicates.

Full size image

Thyroid Hormones Induce Activation of HSC and Increase Expression of αSMA

Because we have not observed effects of thyroid hormones on HSC proliferation, we have determined effects of these hormones on HSC activation. Spontaneous activation of HSC requires serum and is time dependent, so we needed to use serum for the duration of activation of HSC in culture, but use serum without thyroid hormones. Therefore, we have determined activation of HSC in the absence or presence of thyroid hormones using TDS.16, 17 Both 4-day (not shown) and 7-day cultures of HSC grown in TDS show more αSMA-positive cells in TDS+T3 and TDS+T4 (Figure 3a). With activation, HSC lose their retinoic acid granules, so this is another marker of HSC activation. Most HSC after 4 days in culture still show retinoic acid granules determined by Oil Red staining in both TDS and TDS+T3 (Figure 3b). However, although 5-day cultures in TDS have 70% Oil Red-positive cells, HSC grown in TDS+T3 have only 50% Oil Red-positive cells (Figure 3b), suggesting faster activation of HSC in the presence of T3. After 7 days in culture, cells contain almost no Oil Red-positive granules (data not shown).

Figure 3
figure 3

T3 causes more rapid activation of HSC and induces expression of αSMA in HSC. (a) Immunohistochemistry for αSMA expression in HSC cultured for 7 days in TDS, with or without 10−7 M T3 or T4, showing more αSMA-positive cells in the presence of T3 and T4. nc in the first picture is staining with second antibody only; (b) Oil Red staining showing that after 5 days in culture, there are less Oil Red vesicles in HSC grown in TDS+T3. (c) Western blot showing αSMA in HSC cultured for 14 days in TDS with or without 10−7 M T3 or T4. (d) Results are means of fold-induction +s.d. of western blot experiments with αSMA normalized to GAPDH. *P=0.00007 (six experiments); **P=0.032 (four experiments). (e) Effect of 10−7 M T3 on transcription of the −764 αSMA promoter in 7-day HSC. Cells were treated overnight with 10−7 M T3 in serum-free medium, with or without T3 inhibitor 1-850 (2.5 nM) and ROCK inhibitor Y-27632 (10 μM). Results are means of fold-induction +s.d. of four experiments in duplicates, with luciferase activity normalized to β-galactosidase activity. *P=0.002 vs control; **P=0.03 vs T3; ***P=0.0003 vs T3.

Full size image

To quantitate expression of αSMA, we performed western blot analyses of HSC cultured for 14 days in TDS, with or without 10−7 M T3 and T4. Presence of T3 or T4 causes a twofold, statistically significant increase in the levels of αSMA protein, compared with TDS alone (Figure 3c and d). The same effects are observed in 7-day cultures (not shown).

We have next investigated whether T3 and T4 induction of αSMa is transcriptionally mediated, as the αSMA promoter contains a TRE at −173/−149 bp and T3 induces αSMA transcription in cardiomyocytes.17, 18 T3 addition to 7-day cultures of HSC induces a modest (20%), but statistically significant increase in transcription (Figure 3e), which is abolished in the presence of 2.5 nM of the thyroid hormone receptor antagonist 1-850 (Figure 3e). Moreover, αSMA transcription is known to be induced by Rho/ROCK signaling and the presence of 10 μM of ROCK inhibitor Y-27632 reduces transcription of αSMA by itself, as expected, but also abolishes T3 induction of αSMA transcription, indicating that T3 upregulation of αSMA transcription requires Rho/ROCK signaling (Figure 3e).

Thyroid Hormones Induce Expression of P75NTR

We have examined other possible mechanisms responsible for HSC activation by thyroid hormones. A receptor crucial for differentiation of HSC, p75NTR,10 is upregulated in the livers of hyperthyroid rats, following injection of T3 for 10 days.15 We have therefore assessed the effect of T3 on the expression of p75NTR in HSC. Expression of p75NTR is low in 7-day HSC cultures grown in TDS (Figure 4a), and is induced two- and threefold, respectively, in the presence of 10−7 M T3 and T4 in 14-day cultures, as shown by western blot (Figure 4a and b) and by RT–PCR (Figure 4c and d).

Figure 4
figure 4

T3 and T4 induce expression of p75NTR. (a) Western blot analysis showing p75NTR in HSC grown for 7 days in TDS, with or without 10−7 M T3 or T4. (b) Results are means of fold-induction from five western blot experiments +s.d. of p75NTR normalized to GAPDH in 14-day cultures. *P=0.03. (c) RT–PCR showing expression of p75NTR in HSC grown for 14 days in TDS or TDS with 10−7 M T3. (d) RT–PCR of p75NTR normalized to GAPDH, results are means of fold-induction +s.d. of four experiments. *P=0.000005. (e) Effect of 10−7 M T3 or T4 in 7-day cultures on transcription of p75NTR using two fragments of the p75NTR promoter, the −3724/+13 and the −207/+13. Results are means of fold-induction +s.d. of five experiments. *P<0.05 in one-tailed unpaired Student's t-test. (f) Effect of thyroid hormones on transcription of p75NTR using the −3724/+13 promoter of human p75NTR in 7-day HSC. Cells were treated overnight with 10−7 M T3 or T4 in serum-free medium, with or without T3 inhibitor 1-850 (2.5 nM) and T4 inhibitor tetrac (100 nM). Results are means of fold-induction +s.d. of five experiments in duplicates, with luciferase activity normalized to β-galactosidase activity. *P=0.028 vs control; **P=0.049 vs control, #P=0.05 vs T3; ##P=0.036 vs T4.

Full size image

We have next determined whether T3 and T4 affected transcription of p75NTR, by transfecting vectors containing the luciferase reporter gene under the control of the proximal p75NTR promoter in 7-day HSC. We observed that 10−7 M T3 and T4 enhance 1.4- and 1.7-fold, respectively, transcription of the full-length p75NTR promoter, −3724/+13 bp and both hormones cause a twofold increase in the transcription of the shorter fragment, −207/+13 bp (Figure 4e). Moreover, both T3 antagonist 1-850 and T4 inhibitor tetrac abolish the transcriptional effect of T3 and T4 on the full-length p75NTR promoter (Figure 4f).

T3 and T4 Signal through Rho Activation in HSC

We have next studied signaling mechanisms induced by thyroid hormones, which can explain their effects on activation of HSC. Several signal transduction pathways have been shown for thyroid hormones: T4, through integrin αvβ3, induces phosphorylation of MAPK and STAT3,8 whereas T3, in addition to signaling through nuclear receptors TRα1 and TRβ1, triggers association of TRα1 with the p85α subunit of phosphatidyl-inositol-3-kinase (PI3K), leading to Akt phosphorylation.19 Treatment of 7-day cultures of HSC with 10−7 M T3 or T4 for 30 min has no effect on phosphorylation of MAPK, Akt or STAT3 (Figure 5a).

Figure 5
figure 5

T3 and T4 induce Rho activation, but do not affect phosphorylation of MAPK, STAT3 or Akt in HSC. (a) T3 and T4 do not affect phosphorylation of MAPK, STAT3 and Akt. HSC were grown for 5 days in 10% FCS, serum-starved for 48 h, then 10−7 M T3 and T4 were added for 30 min. Results are means of fold-induction +s.d. of three experiments, with expression of pMAPK, pSTAT3 and pAkt normalized to erk2, STAT3 and Akt. (b) Representative western blot analysis showing activated Rho (Rho-GTP) after T3 and T4 treatment in 4- and 7-day HSC. HSC were grown for 2 or 5 days in 10% FCS, serum-starved for 48 h and treated with 10−7 M T3 for 30 min, or 10−7 M T4 and 100 nM tetrac for 15 min. Rho-GTP was assayed using a pull-down assay. Total Rho expression is shown. Rho-GTP expression was normalized to total Rho, and is shown below the blots, as fold-induction vs control. Experiments were repeated qtwice with similar results. (c) Representative western blot showing activated Rho (Rho-GTP) after T3 and T4 treatment in 5-day HSC. HSC were grown for 3 days in 10% FCS, serum-starved for 48 h and pretreated for 30 min with 20 μl (concentration 1 μg/μl) function-blocking antibody to integrin β3 and 500 nM RGD. Then 10−7 M T3 and 10−7 M T4 were added for 15 min and Rho-GTP was assayed using a pull-down assay. Rho-GTP expression was normalized to total Rho, and is shown below the blots, as fold-induction vs control. (d) T3 and T4 induce RhoA activation. HSC were grown for 5 days in 10% FCS, serum-starved for 48 h and treated with various doses of T3 and T4 for 15 min. Activated RhoA was determined using ELISA. Histogram shows representative experiment performed in duplicates, from three experiments with similar results. (e) T3 and T4 induce phosphorylation of non-muscle myosin light chain II. Representative western blot analysis showing pMLC in 5-day cultures of HSC. After 3 days in culture, cells were serum-starved for 48 h, then treated with 10−7 M T3 and T4 for 1/2 or 1 h, with or without 100 nM tetrac. Antibodies to β-actin show equal protein loading.

Full size image

One of the known signals regulating activation of HSC is Rho activation,10 and p75NTR has been shown to induce Rho activation in the absence of its ligand NGF.10, 20 We assume that T3 and T4 may enhance Rho activation either through increased expression of p75NTR, or may directly induce Rho activation. Rho activation has first been tested by pull-down assay with Rhotekin agarose beads. Both T3 and T4 strongly induce Rho activation in both 4- and 7-day HSC, and Rho activation by T4 is abolished by T4 antagonist, tetrac (Figure 5b). Pretreatment of cells with a function-blocking antibody to integrin β3 or with RGD inhibits T4, but not T3, activation of Rho (Figure 5c). RhoA activation is also shown using ELISA. Both T3 and T4 activate RhoA in a dose-dependent manner after 15 min treatment, with a T4 dose of 10−8 M sufficient for inducing RhoA activation (Figure 5d).

The effects of T3 and T4 have next been assayed on the downstream signaling of Rho and ROCK activation, which consists of phosphorylation of pMLC. We have determined effects of T3 and T4 in 5-day cultures of HSC and have observed that T4 induces phosphorylation of MLC after 30 min, whereas T3 maximal effect is after 30 min. Tetrac abolishes the effect of both hormones (Figure 5e).

T3 Increases Expression of TNFα

Activated HSC are known to express many inflammatory and profibrotic cytokines. We have determined mRNA for various cytokines in HSC cultured in TDS, with or without 10−7 M T3, for 4, 7 and 14 days. There is a dramatic increase in SDF-1α expression during HSC activation, with strong SDF-1α expression in 14-day cultures (Figure 6a). T3 increases expression of TNFα and SDF-1α in 4- and 7-day cultures, with no effect on MCP-1 and TGFβ1 (Figure 6a). Moreover, short-term treatment with T3 also induces expression of TNFα, but not of MCP-1 and TGFβ1, as shown after overnight treatment in serum-free medium in 14-day HSC (Figure 6b and c).

Figure 6
figure 6

Effect of T3 on cytokine expression after long-term culture in TDS or after 24 h treatment with T3. (a) Cytokine expression detected by RT–PCR in HSC cultured for 4, 7 and 14 days in TDS with or without 10−7 M T3. (b) RT–PCR showing that T3 treatment for 24 h increases expression of TNFα in 14-day cultures. (c) Expression of TNFα, MCP-1 and TGFβ1 by RT–PCR in HSC cultured for 14 days in 10% FCS, then changed to DMEM with or without 10−7 M T3 for 24 h. Results are means of fold-induction +s.d. of three experiments after normalization to GAPDH. *P=0.00004.

Full size image

Hyperthyroidism Accelerates Development of Liver Fibrosis and Enhances Activation of HSC In Vivo

We have examined whether thyroid hormones also accelerate activation of HSC, not only in vitro, but also in vivo, using a rat model of TAA-induced fibrosis. Rats have received TAA, or TAA+T3 and TAA+T4, for a period of 3 weeks. Injection of T3 to normal rats for 10 days has been shown to induce liver damage, as determined by increased values for ALT and AST.17 We have obtained similar results in TAA+T3-treated rats, which have significantly higher ALT than TAA alone, but less ALT and AST than TAA+T4 (Table 1).

Table 1 Effect of TAA and hyperthyroidism on liver functions, thyroid hormones levels and fibrosis
Full size table

Thyroid-stimulating hormone values are lower in the TAA+T3-treated animals and their total T3 is higher than TAA-treated rats, whereas animals receiving T4 have both elevated T3 and free T4 (Table 1). TAA+T3- and TAA+T4-treated rats have a fibrosis score of 1.6 and 2.8, respectively, compared with 0.8 in rats treated with TAA alone (Table 1). Higher fibrosis score in TAA+T4- vs TAA+T3-treated rats is not due to increased hepatic damage, as ALT is lower in livers of TAA+T4 than in TAA+T3 (Table 1). Interestingly, T4 treatment resulted in significantly higher total T3 levels than treatment with T3 (Table 1). Collagen staining is observed in livers of TAA+T3- and TAA+T4-injected rats, but not in TAA-treated rats (Figure 7a). There is weak expression of αSMA and p75NTR in normal rat liver, yet comparable in its levels to expression of both proteins in livers of rats treated for 3 weeks with TAA (Figure 7b). The source of αSMA in the normal liver can be endothelial cells, and also 3-week treatment with TAA is only the beginning of activation of HSC, explaining weak levels of αSMA. Expression of αSMA is increased in a statistically significant manner in the livers of both TAA+T3- and TAA+T4-treated rats, compared with rats receiving TAA (Figure 7c and d). However, only TAA+T4 treatment, but not TAA+T3, results in statistically higher levels of p75NTR (Figure 7e and f). We assume that T3 may enhance p75NTR expression at an earlier time, and as we observed that T3 induces a twofold increase and T4 a threefold increase of p75NTR in vitro (Figure 4b), T4 may also be more potent than T3 in enhancing expression of p75NTR (Figure 8).

Figure 7
figure 7

Hyperthyroidism induces fibrosis and enhances expression of αSMA in livers of TAA-treated rats. (a) Collagen expression as detected by Masson Trichrome staining of livers after 3 weeks treatment with TAA, TAA+ T3 and TAA+ T4. (b) Western blot analysis showing αSMA, p75NTR and β-actin expression in normal liver (N) and livers of rats injected for 3 weeks with TAA; (c) western blot showing the effect of TAA with or without T3 or T4 administration on αSMA expression. (d) Results are means +s.d. of αSMA normalized to β-actin (n=3). *P<0.05 by unpaired one-tailed Student's t-test. (e) Western blot analysis showing the effect of TAA with or without T3 or T4 administration on p75NTR expression (f) Results are means +s.d. of p75NTR normalized to β-actin. *P=0.02.

Full size image
Figure 8
figure 8

Diagram of mechanisms of activation of HSC mediated by T3 and T4. (a) T4 binds to cell-surface integrin αvβ3 and induces Rho activation, leading to increased expression of αSMA. T4 binding to integrin αvβ3 is blocked by tetrac, RGD and an antibody against β3. (b) T3 and T4 enter the cell, and induce transcription of p75NTR, through either TRα1 binding to TREs in the p75NTR promoter, or inducing transcription of other transcription factor (eg, egr-1) that enhance synthesis of the p75NTR gene. Upregulation of p75NTR on the cell surface results in activation of Rho, leading to increased expression of αSMA. (c) T3 binds to TRα1, which activates αSMA transcription via TREs in the αSMA promoter.

Full size image

DISCUSSION

We have found that thyroid hormones, T3 and T4, are important for activation of primary HSC both in vitro and in vivo. We show that T3 and T4 enhance activation of HSC by three different mechanisms: direct upregulation of αSMA, increased expression of p75NTR and rapid, direct activation of Rho. Activation of Rho is a novel signaling pathway, previously unreported for these hormones.

The profibrogenic effect of hyperthyroidism has been reported by several studies from our laboratory, where hyperthyroidism increases and hypothyroidism alleviates liver fibrosis in a TAA rat model,3, 4 yet the mechanisms involved are unknown.

We show for the first time that TRα1 expression is regulated by the state of activation of HSC, as strongly activated HSC after 14 days in culture have reduced expression of TRα1, yet the same cells show increased levels of T4 receptor αv, as was previously reported.9 Interestingly, expression of other nuclear hormone receptors involved in lipid metabolism, such as peroxisome proliferator-activated receptors-γ, is also reduced in activated HSC, and overexpression of these receptors in activated HSC induces a quiescent phenotype.21

The main effect of T3 and T4 is upregulation of αSMA expression and accelerated activation of HSC in vitro, and the effect of T3 on αSMA is transcriptional and abolished by T3 antagonist 1-850. Similar findings were reported for cardiomyocytes, where T3 enhances αSMA transcription through TRα1 binding to a TRE in the αSMA promoter and T3 addition to TDS induces αSMA in a dose-dependent manner.17, 18

Recently, it was reported that HSC from p75NTR−/− mice failed to activate and did not express αSMA.10 We show that HSC grown in TDS have low levels of p75NTR and addition of T3 and T4 results in a strong increase in the expression of this receptor, similar to the expression of αSMA under the same conditions. Moreover, T3 and T4 treatment induces transcription of the p75NTR promoter, T3 acting through TRα1 and T4 through cell-surface integrin αvβ3. The −207/+13 p75NTR promoter is sufficient for the T3- and T4-induced transcriptional effects. The in vitro effects of T3 and T4 are also observed in livers of rats injected for 3 weeks with TAA, where treatment with T3 and T4 increased expression of αSMA in a statistically significant manner. The lack of effect of TAA and T3 on induction of p75NTR in vivo can be explained both by the fact that in vitro experiments show that T4 is more potent than T3 in increasing p75NTR (threefold vs twofold) and by the fact that in vivo administration of T4 resulted in significantly higher serum concentrations of total T3 and free T4, compared to administration of T3.

P75NTR, depending on partnering receptors and on the balance between neurotrophins or proneurotrophins, leads to cell survival or apoptosis.22 Expression of p75NTR is increased after injury and inflammation, helping recovery from injury, as in remyelination by Schwann cells.23 TNFα, lipopolysaccharide and transcription factor egr-1, are some of the known regulators of p75NTR transcription.23 Thyroid hormones may induce transcription of p75NTR directly by binding to TREs in the p75NTR promoter or alternatively, T3 may induce transcription of egr-1, as reported for Schwann cells.24

In HSC, p75 NTR has a dual role: in the absence of its ligand NGF, it induces cell activation,23 whereas in the presence of NGF, it induces HSC apoptosis.21 In the absence of NGF, p75NTR binds Rho inhibitor, Rho-GDI, thus allowing Rho activation.25 We show that T3 and T4 rapidly and directly induce activation of Rho, with no effect on expression of total Rho. This is a newly described signaling pathway for the two hormones. Rho activation regulates cell adhesion, motility, invasion and metastasis through reorganization of cytoskeleton components.26 Rho activation is induced by growth factors and extracellular matrix components.26 Integrin β3 signaling mediates Rho activation,11 explaining T4 activation of RhoA by binding to integrin αvβ3. T3 activates Rac1, another GTPase, following association of TRβ1 with the p85α subunit of the PI3K.25 However, in HSC, PI3K is not activated by T3, as shown by unchanged levels of pAkt.

Rho signaling is central to activation of HSC, and inhibition of its downstream signal ROCK by Y-27632, inhibits liver fibrosis and HSC proliferation and activation.11, 12 Moreover, activation of Rho induces a myofibroblast, αSMA-expressing differentiation program in several cell types.27, 28 Activation of Rho is followed by activation of ROCK, which phosphorylates non-muscle myosin II, inducing translocation of transcription factors serum response factor and myocardin-related transcriptional activator from the cytoplasm to the nucleus and transcription of αSMA.28 In accordance with their effect on Rho activation, we show that thyroid hormones also increase phosphorylation of non-muscle myosin II.

In addition to its effect on HSC activation, T3 has also a pro-inflammatory effect on HSC phenotype, enhancing mRNA expression of TNFα and SDF1α, which may also contribute to increasing fibrosis in vivo. Addition of T3 to TDS does not further increase expression of SFD1α in fully activated HSC (14 days in culture), as seen in 4 and 7-day cultures, suggesting that T3 is necessary for regulation of SDF1α expression only during earlier stages of HSC activation. Expression of SDF1α is strongly induced by in vitro activation of HSC.29 Moreover, SDF1α increases activation of HSC by upregulating αSMA and collagen I and expression levels of both SDF1α and its receptor CXRC4 have been shown to be increased in cirrhotic livers of hepatitis C patients.30

MCP-1 expression in HSC can be induced by leptin, prostaglandins, TNFα and lipopolysaccharide,31, 32 and its expression is regulated by different factors, depending on the activation stage of HSC.33 We have not detected any effect of T3 on modulation of MCP-1 mRNA levels in quiescent, activated or fully activated HSC.

Interestingly, we have observed that treatment of T3 and T4 together with TAA caused liver damage, as shown by high levels of ALT. However, development of liver fibrosis cannot be attributed to apoptosis of damaged hepatocytes, as T3 causes significantly more liver damage than T4, yet T4 is more potent in inducing fibrosis, indicating that fibrosis results from activation of HSC.

In conclusion, thyroid hormones contribute to an activated HSC phenotype by increasing levels of p75NTR, activating Rho and inducing αSMA and thereby participate in accelerated development of liver fibrosis. Exploring these novel mechanisms may increase the range of therapeutic options for treatment of liver fibrosis.