Caffeine Suppresses the Activation of Hepatic Stellate Cells cAMP-Independently by Antagonizing Adenosine Receptors

Abstract

During liver injury, hepatic stellate cells (HSCs) are activated by various cytokines and transdifferentiated into myofibroblast-like activated HSCs, which produce collagen, a major source of liver fibrosis. Therefore, the suppression of HSC activation is regarded as a therapeutic target for liver fibrosis. Several epidemiological reports have revealed that caffeine intake decreases the risk of liver disease. In this study, therefore, we investigated the effect of caffeine on the activation of primary HSCs isolated from mice. Caffeine suppressed the activation of HSC in a concentration-dependent manner. BAPTA-AM, an intracellular Ca²⁺ chelator, had no effect on the caffeine-induced suppression of HSC activation. None of the isoform-selective inhibitors of phosphodiesterase1 to 5 affected changes in the morphology of HSC during activation, whereas CGS-15943, an adenosine receptor antagonist, inhibited them. Caffeine had no effect on intracellular cAMP level or on the phosphorylation of extracellular signal-regulated kinase (ERK)1/2. In contrast, caffeine significantly decreased the phosphorylation of Akt1. These results suggest that caffeine inhibits HSC activation by antagonizing adenosine receptors, leading to Akt1 signaling activation.

Liver fibrosis is a major consequence of chronic liver disease, in which the activation of hepatic stellate cells (HSCs) is a critical event. HSCs are located in the space of Disse and attach to sinusoidal endothelial cells, in a similar manner to pericytes on capillaries.¹⁾ Under physiological conditions, HSCs have long protrusions, forming a star-like shape, and are called quiescent HSCs.²⁾ Quiescent HSCs contain lipid droplets, including vitamin A, and play a primary role in storing and supplying vitamin A.³⁾ On the other hand, under conditions of liver injury, quiescent HSCs are activated by cytokines and transdifferentiate into myofibroblast-like activated HSCs.^4,5)

Activated HSCs are characterized by increased proliferation, migration, and contraction, and by a decrease in vitamin A content. Activated HSCs secret cytokines such as hepatocyte growth factor for the regeneration of the liver, and produce collagen, which leads to liver fibrosis.⁶⁾ HSCs are activated by oxidative stress and paracrine signaling from various neighboring cells, such as Kupffer cells, sinusoidal endothelial cells, hepatocytes, and platelets. Furthermore, the activation of HSCs is enhanced by autocrine signaling.^1,4) Thus, HSCs play a critical role in the progression from liver fibrosis to hepatic cirrhosis and hepatoma, which makes them a suitable target for liver fibrosis therapy.⁶⁾

Caffeine is widely ingested in coffee, green tea, black tea, and chocolate. Several epidemiologic studies have suggested that intake of coffee and green tea is related to decreased liver disease.^7,8) However, the mechanism of this caffeine effect has not been investigated in primary HSCs. In the present study, we investigated the underlying mechanism of the inhibition by caffeine of HSC activation induced by fetal bovine serum (FBS) in primary HSCs isolated from mice.

MATERIALS AND METHODS

Isolation and Culture of Hepatic Stellate Cells

HSCs were isolated from male ddY mice (body weight 40–60 g; SLC, Hamamatsu, Japan) by digestion with PRONASE^® (Merck-Millipore, Tokyo, Japan) and collagenase (Yakult, Tokyo, Japan) and fractionated by 13% HistoDenz (Sigma, St. Louis, MO, U.S.A.) according to the method described by Kojima et al.⁹⁾ HSC was identified by autofluorescence derived from vitamin A, which was excited at 340 nm and emitted at 550 nm. The purity of HSC in each dish was confirmed to be greater than 90%. Isolated HSCs were seeded on culture dishes or multi wall plates in Dulbecco’s modified Eagle’s medium (DMEM, Nissui Pharmaceutical, Tokyo, Japan) supplemented with 10% FBS (Gibco, Life Technologies) and incubated for overnight at 37°C in a 95% air-5% CO₂ atmosphere. The next day, the time point before the first medium-change was taken as day 1 and the medium was changed every 2 d with or without reagents thereafter. The time point before 3rd medium-change was taken as day 7. Cells cultured for 2 and 7 d were regarded as quiescent and activated HSCs, respectively.

Assessment of HSC Activation

HSCs were cultured on a 96-well plate for 7 d. Cells were fixed with 2% paraformaldehyde (Wako Pure Chemical Industries, Ltd., Osaka, Japan) for 45 min, permeabilized with 0.1% Triton X-100 for 1 h, and blocked with 3% bovine serum albumin fraction V (BSA, Roche, Mannheim, Germany) for 30 min in phosphate-buffered saline (PBS; 137 mM NaCl, 8.10 mM Na₂HPO₄·12H₂O, 2.68 mM KCl, 1.47 mM KH₂PO₄). The cells were then incubated with mouse monoclonal anti-α-smooth muscle actin (α-SMA) antibody (1 : 1000, Sigma) in PBS containing 1% BSA overnight at 4°C, followed by incubation with Alexa Fluor^® 546 conjugated goat anti-mouse immunoglobulin G (IgG) antibody (1 : 750, Invitrogen, Carlsbad, CA, U.S.A.) or Alexa Fluor 488 conjugated phalloidin (1 : 750, Invitrogen) in PBS containing 1% BSA for 1 h at room temperature. Nuclei were stained with Hoechst 33342 (1 : 1500, Dojindo, Kumamoto, Japan). The α-SMA positive area was measured as an index of the activation of HSC using ImageJ software.¹⁰⁾

RT-PCR

Total RNA was extracted from quiescent and activated HSCs using an RNeasy mini kit according to manufacturer’s instructions (Qiagen, Tokyo, Japan). RT-PCR was performed using a Qiagen One step RT-PCR kit (Qiagen). The primers for RT-PCR are listed in Table 1.

Table 1. Sequences of Primers Used for RT-PCR

Gene	Forward sequences (5′→3′)	Reverse sequences (5′→3′)
PDE1A	CTAAAGATGAACTGGAGGGATCTTCGGAAC	TGGAGAAAATGGAAGCCCTAATTCAGC
PDE1B	CCACCCAACTAAGCAGTGGT	GGTTGGCTCCACAATGAAGT
PDE1C	ATGGTTGGGCTGAGCTATCCACC	CCAGTTTGCCACTCCTGTCTTATAAAGGAGG
PDE2A	GCATGTGTCATGACCTGGAC	AGCATGCGCTGATAGTCCTT
PDE3A	TGCTATACAACGACCGTTCT	ATGACTAGGAAGCGGAAGTG
PDE3B	AACCATGAAACGGAAACCAA	GTGATCGTAATCGTGCATGG
PDE4A	GCGGGACCTACTGAAGAAATTCC	CAGGGTGGTCCACATCGTGG
PDE4B/C	CGCTGGAGGAACTAGACTGG	CAGACACCTGGTTCCCTGAT
PDE4D	CCTCTGACTGTTATCATGCACCACC	GATCCACATCATGTATTGCACTGGC
PDE5A	GGCCTCTCCATTGACGAATA	ATTCAGCTGCCACGAGTTCT
PDE7A	CGGAACTGGGAATTAAGCAA	GACAGCCGTGTGTCTGAAAA
PDE7B	CAAAGTTGGAACGTGGGACT	GGGTTGTGACCGTGGTAATC
RyR1	TACTTCGACACAACCCCACA	ACAGTCTCCAGCAGGGAAGA
RyR2	GCGAGCTGGCTACTATGACC	CGTTGCTAATGCTCACGAAA
RyR3	AACGTCCTGCTCTTGGAAA	ATGTCCTCCACCTTGTCTGG
GAPDH	AACGACCCCTTCATTGAC	TCCACGACATACTCAGCAC
β-Actin	CTCTCAGCTGTGGTGGTGAA	AGCCATGTACGTAGCCATCC

PDE: phosphodiesterase, RyR: ryanodin receptor, GAPDH: glyceraldehydes-3-phospate dehydrogenase.

cAMP Level

On the 2nd day after cell seeding, HSCs were treated with forskolin or caffeine. After 15 min, the reaction was terminated with 0.1 N HCl, and the sample was collected by cell scraper. Precipitation was removed by centrifugation (1000×g, 10 min). The concentration of cAMP in the supernatant was measured using a Cyclic AMP EIA kit (Cayman Chemical, Ann Arbor, MI, U.S.A.). cAMP level was normalized by DNA content for each sample. The DNA content was measured by spectrophotometer.

Western Blot

HSCs treated with caffeine (3 mM) or CGS-15943 (1 µM) from day 2 for 24 h were collected and disrupted with a sonicator (UH-50, SMT Co., Ltd., Tokyo, Japan) in a homogenization buffer (20 mM Tris-HCl (pH 7.4), 2 mM ethylenediaminetetraacetic acid (EDTA), 50 µg/mL phenylmethylsulfonyl fluoride, 10 µg/mL aprotinin, 10 µg/mL leupeptin, and 250 mM sucrose) on ice. The cell lysates were loaded with one-fifth volume of a sampling buffer (125 mM Tris-HCl (pH 6.8), 4% sodium dodecyl sulfate (SDS), 20% glycerol, 10% 2-mercaptoethanol, and 0.001% bromophenol blue) and boiled at 95°C for 5 min. After SDS-polyacrylamide gel electrophoresis, the protein was transferred to a polyvinylidene difluoride membrane, and blocked with Blocking One-P^® (Nacalai Tesque, Kyoto, Japan) in TBS-T buffer (10 mM Tris, 137 mM NaCl, 0.1% Tween 20) for 1 h at room temperature. The membrane was then incubated with anti-phospho-extracellular signal-regulated kinase (ERK)1/2 antibody (1 : 2500, Cell Signaling Technology, Danvers, MA, U.S.A.), anti-ERK1/2 antibody (1 : 2500, Proteintech, Manchester, U.K.), anti-phospho-Akt1 antibody (1 : 10000, Abcam, Cambridge, U.K.), anti-Akt1 antibody (1 : 10000, Abcam) or anti-vinculin antibody (1 : 50000, Abcam) in Can Get Signal^® Solution 1 (Toyobo, Osaka, Japan) overnight at 4°C. After washing away the free antibodies, the membrane was incubated with anti-rabbit IgG horseradish peroxidase (HRP)-linked antibody (1 : 5000, Sigma) in Can Get Signal Solution 2 (Toyobo) for 1 h at room temperature. Signals were detected using ImmunoStar LD reagents (Wako).

Reagents

O,O′-Bis(2-aminophenyl)ethyleneglycol-N,N,N′,N′-tetraacetic acid, tetraacetoxymethyl ester (BAPTA-AM) was purchased from Dojindo (Kumamoto, Japan); erythro-9-(2-hydroxy-3-nonyl) adenine, HCl (EHNA) and rolipram were from Enzo Life Sciences (Farmingdale, NY, U.S.A.); caffeine (anhydrous), cilostamide, 8-methoxymethyl-3-isobutyl-1-methylxanthine (8-Me-IBMX), and zaprinast were from Sigma; and other chemicals were purchased from Wako.

Statistics

Data are expressed as mean±standard error of the mean (S.E.M.) Groups were compared using the Tukey-Kramer multiple comparison test or the unpaired Student’s t-test. Statistical results of less than 5% (p<0.05) were considered significant.

RESULTS

The area of α-SMA-positive staining, which is a marker of activated HSC, was markedly increased, and the autofluorescence of vitamin A, which is a marker of quiescent HSC, disappeared due to the culturing of the primary HSCs in DMEM containing FBS for 7 d (Figs. 1A, B). Caffeine decreased the α-SMA-positive area in a concentration-dependent manner (Figs. 1C, D). In the presence of 10 mM caffeine, the autofluorescence of vitamin A was still observed at day 7 (Fig. 1B). These results suggest that caffeine suppressed the activation of HSCs induced by FBS.

Fig. 1. Effect of Caffeine on Activation of HSCs Cultured for 7 d

(A) Immunostaining images with anti-α-smooth muscle actin (α-SMA; red) of HSCs cultured for 2 and 7 d in DMEM containing FBS, which represent quiescent HSCs (qHSC) and activated HSCs (aHSC), respectively. Nuclei were stained with Hoechst 33342 (blue). (B) Phase contrast and vitamin A autofluorescence images of HSCs cultured for 2 d and those cultured for 7 d in the absence or presence of 10 mM caffeine. (C) Immunostained α-SMA in HSCs cultured in the absence (control) and presence of 3 and 10 mM caffeine. The α-SMA-positive area was measured and normalized to per cell (D). Data are expressed as mean±S.E.M. (n=4 to 5). ** p<0.01 vs. control.

Caffeine is well known to function as an agonist of ryanodine receptor Ca²⁺ release channels, as an inhibitor of phosphodiesterase (PDE), and as a non-selective adenosine receptor antagonist. To elucidate the mechanism of the inhibitory effect of caffeine on HSC activation, we first investigated the possible involvement of ryanodine receptors in the caffeine effect. We confirmed that both quiescent and activated HSCs expressed the mRNAs of RyR1 and RyR2. The contribution of Ca²⁺ to the inhibitory effect of caffeine on the activation of HSCs was examined using an intracellular Ca²⁺ chelator BAPTA-AM. BAPTA-AM (5 µM) with 0.05% cremophor EL was loaded into quiescent HSCs 1.5 h before the addition of 3 mM caffeine. After 7-d culturing, there was no difference in cell morphology or α-SMA-positive area between cells treated with or without BAPTA-AM (Figs. 2B, C), suggesting that Ca²⁺ release through ryanodine receptors contributed little to the suppressive effect of caffeine on HSC activation.

Fig. 2. Possible Involvement of Ca²⁺ Release from Endoplasmic Reticulum in the Suppressive Effect of Caffeine on HSC Activation

(A) RT-PCR analysis of the expression of ryanodine receptor (RyR) 1, 2, 3 in quiescent and activated HSCs. (B) Immunostaining of α-SMA in HSCs cultured for 7 d in the absence or the presence of caffeine (3 mM). Before the application of caffeine, the cells were treated with or without BAPTA-AM (5 µM). The α-SMA-positive area of HSCs was measured and normalized to per cell (C). Data are expressed as mean±S.E.M. (n=4). Not significant (N. S.) vs. without BAPTA-AM.

We next examined the contribution of PDE inhibition to the suppressive effect of caffeine on HSC activation. Using RT-PCR, we confirmed the mRNA expression of PDE1A, 1B, 1C, 2A, 3A, 3B, 4A, 4B/C, 4D, 5A, 7A, and 7B in both quiescent and activated HSCs (Fig. 3A). The effect of PDE inhibition on the activation of HSCs was investigated with subtype-specific PDE inhibitors, i.e., the PDE1 inhibitor 8-Me-IBMX (30 µM), the PDE2 inhibitor EHNA (10 µM), the PDE3 inhibitor cilostamide (1 µM), the PDE4 inhibitor roripram (10 µM), and the PDE5 inhibitor zaprinast (3 µM). None of the PDE inhibitors had a significant effect on morphology or on the α-SMA-positive area of the HSCs (Figs. 3B, C). We further measured intracellular cAMP level using an enzyme-linked immunosorbent assay (ELISA) kit. In quiescent HSCs, forskolin (10 µM), an adenylate cyclase activator, drastically increased intracellular cAMP levels, whereas caffeine (3 mM) had no significant effect on these levels under the current experimental conditions (Fig. 3D).

Fig. 3. Possible Involvement of PDE Inhibition in the Suppressive Effect of Caffeine on HSC Activation

(A) RT-PCR analysis of the expression of PDE subtypes in quiescent and activated HSCs. mRNA expression of PDE1A, 1B, 1C, 2A, 3A, 3B, 4A, 4B/C, 4D, 5A, 7A, 7B was detected in both quiescent and activated HSCs. (B) Immunostaining of α-SMA in HSCs cultured for 7 d in the presence of 8-Me-IBMX (30 µM), EHNA (10 µM), cilostamide (CSM; 1 µM), rolipram (RLP; 10 µM), and zaprinast (ZPN; 3 µM). The α-SMA-positive area was measured and normalized to the control (C). (D) Intracellular cAMP levels in HSCs cultured for 2 d in the absence and presence of caffeine (3 mM) or forskolin (10 µM). Data are expressed as mean±S.E.M. (n=5). Not significant (N. S.) vs. control.

Finally, we investigated whether the adenosine receptor antagonistic effect of caffeine is also involved in caffeine’s suppressive effect on HSC activation. CGS-15943 (1 µM), a subtype-nonselective adenosine receptor antagonist, significantly decreased the area of α-SMA positive staining (Figs. 4A, B). It is well known that adenosine A₁ and A₃ receptors inhibit cAMP production via Gi, whereas adenosine A_2A and A_2B receptors stimulate it via Gs.¹¹⁾ However, we did not find any changes in intracellular cAMP levels in the HSCs due to caffeine. Therefore, we explored the possible involvement of other signaling pathways. Caffeine (3 mM) and CGS-15943 (1 µM) significantly decreased the phosphorylation of Akt1 (Figs. 4C, E). Concomitantly, caffeine tended to inhibit and CGS-15943 significantly inhibited expression level of Akt1 (Figs. 4C, E). In contrast, both inhibitors showed significant effect on neither the phosphorylation level nor the total level of ERK1/2 (Figs. 4D, F).

Fig. 4. Possible Involvement of Adenosine Receptor Antagonist Effect in the Suppressive Effect of Caffeine on HSC Activation

(A) Immunostaining of α-SMA in HSCs cultured for 7 d in the presence of CGS-15943 (1 µM). The α-SMA-positive area was measured and normalized to per cell (B). Western blot analysis for phosphorylation or total level of ERK1/2 (C) and Akt1 (D) in HSCs cultured in the absence or presence of 3 mM caffeine for 24 h. Western blot analysis for phosphorylation or total level of ERK1/2 (E) and Akt1 (F) in HSCs cultured in the absence or presence of 1 µM CGS-15943 for 24 h. Each band was normalized to vinculin level. Data are expressed as mean±S.E.M. (n=4). Not significant (N. S.), * p<0.05, ** p<0.01 vs. control.

DISCUSSION

The activation of HSCs in primary culture can be induced by FBS, which includes various growth factors.⁴⁾ In the present study, we have clearly demonstrated that the activation of mouse primary HSCs by FBS was suppressed by caffeine: Caffeine decreased the area of α-SMA positive staining, which is a marker of activated HSCs, and resulted in the retention of vitamin A autofluorescence, which is a marker of quiescent HSCs. These results support epidemiological studies which have shown the suppression of liver fibrosis due to the intake of caffeine or caffeine-containing beverages.^7,8) We have further investigated the mechanism underlying this caffeine effect. Among the well-known functions of caffeine, we suggest that its antagonistic effects on adenosine receptors are involved in the suppressive effect on HSC activation, which is likely to be mediated via the inhibition of Akt1.

Caffeine is known to function as an adenosine receptor antagonist. In the present study, we found that the adenosine receptor antagonist CGS-15943, like caffeine, decreased the area of α-SMA-positive staining in mouse HSCs cultured in the presence of FBS for 7 d, which indicates the suppression of HSC activation. On the other hand, the other functions of caffeine, i.e., agonistic effects on ryanodine receptors or inhibition of PDE, are unlikely to be involved, as discussed below. We confirmed using LC/MS that adenosine was present in the culture medium under our experimental conditions (data not shown). This suggests that caffeine inhibits the activation of HSCs by antagonizing adenosine receptors. Several studies have suggested the contribution of adenosine to the development of hepatic fibrosis: Ethanol, CCl₄, and thioacetamide induce extracellular generation of adenosine by ecto-5′-nucleotidase (CD73) in hepatic slices, which contributes to the development of alcohol-induced fatty liver and hepatic fibrosis.¹²⁾ Ethanol has also been shown to stimulate adenosine release from the hepatoma cell line HepG2.¹³⁾

Some G protein-coupled receptors (GPCRs) have been shown to activate the Akt pathway.¹⁴⁾ There are several reports suggesting that the Akt pathway mediates responses to adenosine receptor stimulation: The phosphorylation of Akt and ERK1/2 via the activation of adenosine A_2B receptors regulates the function of glioblastoma stem cells,¹⁵⁾ and the activation of adenosine A₃ receptors induces the activation of Akt in melanoma cells.¹⁶⁾ The focal adhesion kinase-phosphatidylinositol 3 (PI3)-kinase-Akt signaling pathway has been suggested to be critical for HSC proliferation and type I collagen expression in activated rat HSCs.¹⁷⁾ In the present study, caffeine and CGS-15943 significantly inhibited the phosphorylation of Akt1 in mouse HSCs. The phosphorylation of ERK1/2 seems also to lead to the activation of HSCs,^18,19) and several GPCRs, including adenosine receptors, can activate ERK1/2.^15,20) However, the present study shows that caffeine and CGS-15943 had no effect on the phosphorylation of ERK1/2 in mouse HSCs. This suggests that adenosine receptors induce the activation of Akt1, thereby leading to the activation of HSCs, and that caffeine suppresses this mechanism by antagonizing adenosine receptors. Interestingly, similarly to the present results in mouse HSCs, caffeine and CGS-15943 have been shown to suppress proliferation of human hepatic cancer cells by inhibiting the PI3 kinase-Akt pathway, independently of ERK1/2.²¹⁾ As the present study has determined neither the adenosine receptor subtype involved in the caffeine effect nor the detailed signaling components of the Akt pathway, a more detailed analysis is needed to confirm this notion.

The present results using BAPTA-AM exclude the possible involvement of intracellular Ca²⁺ in the caffeine effect on HSC activation. There are several reports that Ca²⁺ influences proliferation and apoptosis in activated HSCs.^22,23) Caffeine at high concentrations causes a pronounced increase in the sensitivity of ryanodine receptors to Ca²⁺, thereby inducing Ca²⁺ release even at basal cytosolic Ca²⁺ concentrations. In addition, we have confirmed the expression of RyR1 and RyR2 in both quiescent and activated HSCs. However, BAPTA-AM, a chelator of intracellular Ca²⁺, did not affect the suppressive effect of 3 mM caffeine on HSC activation. It is plausible that Ca²⁺ is not involved in the effect of caffeine, at least at the 3 mM level, on HSC activation.

The present study also provides evidence against the involvement of cAMP elevation through PDE inhibition in the caffeine effect. Several studies have suggested that cAMP mediates the inhibition of HSC activation. Kawada et al. reported that dibutyryl cAMP inhibited the myofibroblastic transformation of rat HSCs in vitro.²⁴⁾ Moreover, ATP and adenosine have been shown to suppress the activation of HSCs via cAMP elevation.^25,26) In agreement with these findings, Shim et al. reported that caffeine decreased fibrosis and α-SMA expression in human HSC line LX-2 cells cultured in DMEM supplemented with FBS, and that these decreases were accompanied by an elevation in cAMP level.²⁷⁾ However, other studies have reported contradictory results: The inhibition by caffeine of acetaldehyde-induced activation of HSC-T6 cell lines²⁸⁾ and rat HSCs²⁹⁾ was accompanied by a decrease in cAMP level by antagonizing Gs-coupled adenosine A_2A receptors. Therefore, the involvement of cAMP in the caffeine effect on HSC activation is still controversial. In the present study, no selective PDE inhibitors suppressed the activation of HSCs, even though the concentration of each PDE inhibitor used was 10 times higher than IC₅₀.^30–32) Moreover, caffeine did not increase intracellular cAMP level in HSCs. We also confirmed that the adenylate cyclase inhibitor SQ22,536 (10 µM) did not affect the decrease in the area of α-SMA-positive staining induced by 3 mM caffeine (data not shown). Taken together, these results suggest that cAMP elevation through PDE inhibition is unlikely to be involved in the suppressive effect of caffeine on the HSC activation, at least under the experimental conditions of this study. The production of cAMP in isolated mouse HSCs in this study may have been at too small a volume to cause any effects even after the inhibition of PDE.

In a clinical study, plasma caffeine levels have been 0.01–0.2 mM by intake of a cup of coffee (caffeine 100 mg/150 mL coffee).³³⁾ It is important to point out that caffeine contacts to HSCs before metabolizing by hepatocytes, and the plasma caffeine level in the portal vein should be higher than that in whole body. It is, therefore, possible to say that caffeine acts on HSCs at mM range. Moreover, the plasma caffeine concentration in liver disease patients has been shown to be 7 times higher than that in healthy subjects.³³⁾ This also supports the clinical use of caffeine.

In summary, in the present study, isolated mouse HSCs were transformed into activated HSCs by culturing in a medium supplemented with FBS. We found a suppressive effect of caffeine on the activation of HSCs by FBS and investigated the signaling mechanism. These results suggest that caffeine suppresses the activation of HSCs by inhibiting the adenosine receptor-mediated signaling pathway involving Akt1. They also offer new insight into the role of the adenosine receptors that stimulate the Akt1 pathway in the pathogenesis of liver fibrosis. These findings may help to develop a rationale for the use of adenosine receptor antagonists to prevent the progression of liver fibrosis.

Acknowledgment

We thank Philip Hawke of the University of Shizuoka Scientific English Program for correcting the English in the manuscript.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Friedman SL. Hepatic stellate cells: protean, multifunctional, and enigmatic cells of the liver. Physiol. Rev., 88, 125–172 (2008).
• 2) Atzori L, Poli G, Perra A. Hepatic stellate cell: a star cell in the liver. Int. J. Biochem. Cell Biol., 41, 1639–1642 (2009).
• 3) Senoo H. Structure and function of hepatic stellate cells. Med. Electron Microsc., 37, 3–15 (2004).
• 4) Friedman SL. Cytokines and fibrogenesis. Semin. Liver Dis., 19, 129–140 (1999).
• 5) Nwosu ZC, Alborzinia H, Wölfl S, Dooley S, Liu Y. Evolving insights on metabolism, autophagy, and epigenetics in liver myofibroblasts. Front. Physiol., 7, 191 (2016).
• 6) Li J-T, Liao Z-X, Ping J, Xu D, Wang H. Molecular mechanism of hepatic stellate cell activation and antifibrotic therapeutic strategies. J. Gastroenterol., 43, 419–428 (2008).
• 7) Jin X, Zheng R, Li Y. Green tea consumption and liver disease: a systematic review. Liver Int., 28, 990–996 (2008).
• 8) Modi AA, Feld JJ, Park Y, Kleiner DE, Everhart JE, Liang TJ, Hoofnagle JH. Increased caffeine consumption is associated with reduced hepatic fibrosis. Hepatology, 51, 201–209 (2010).
• 9) Kojima N, Hori M, Murata T, Morizane Y, Ozaki H. Different profiles of Ca²⁺ responses to endothelin-1 and PDGF in liver myofibroblasts during the process of cell differentiation. Br. J. Pharmacol., 151, 816–827 (2007).
• 10) Abramoff MD, Magelhães PJ, Ram SJ. Image processing with imageJ. ResearchGate, 11, 36–42 (2003).
• 11) Klinger M, Freissmuth M, Nanoff C. Adenosine receptors: G protein-mediated signalling and the role of accessory proteins. Cell. Signal., 14, 99–108 (2002).
• 12) Peng Z, Fernandez P, Wilder T, Yee H, Chiriboga L, Chan ESL, Cronstein BN. Ecto-5′-nucleotidase (CD73) -mediated extracellular adenosine production plays a critical role in hepatic fibrosis. FASEB J., 22, 2263–2272 (2008).
• 13) Chan ESL, Montesinos MC, Fernandez P, Desai A, Delano DL, Yee H, Reiss AB, Pillinger MH, Chen J-F, Schwarzschild MA, Friedman SL, Cronstein BN. Adenosine A_2A receptors play a role in the pathogenesis of hepatic cirrhosis. Br. J. Pharmacol., 148, 1144–1155 (2006).
• 14) New DC, Wu K, Kwok AWS, Wong YH. G protein-coupled receptor-induced Akt activity in cellular proliferation and apoptosis. FEBS J., 274, 6025–6036 (2007).
• 15) Liu TZ, Wang X, Bai Y, Liao H, Qiu S, Yang Y, Yan X-H, Chen J, Guo H, Zhang S. The HIF-2alpha dependent induction of PAP and adenosine synthesis regulates glioblastoma stem cell function through the A2B adenosine receptor. Int. J. Biochem. Cell Biol., 49, 8–16 (2014).
• 16) Madi L, Rosenberg-Haggen B, Nyska A, Korenstein R. Enhancing pigmentation via activation of A3 adenosine receptors in B16 melanoma cells and in human skin explants. Exp. Dermatol., 22, 74–77 (2013).
• 17) Reif S, Lang A, Lindquist JN, Yata Y, Gabele E, Scanga A, Brenner DA, Rippe RA. The role of focal adhesion kinase-phosphatidylinositol 3-kinase-akt signaling in hepatic stellate cell proliferation and type I collagen expression. J. Biol. Chem., 278, 8083–8090 (2003).
• 18) Sun W-Y, Song Y, Hu S-S, Wang Q-T, Wu H-X, Chen J-Y, Wei W. Depletion of β-arrestin2 in hepatic stellate cells reduces cell proliferation via ERK pathway. J. Cell. Biochem., 114, 1153–1162 (2013).
• 19) Rietz A, Volkov Y, Davies A, Hennessy M, Spiers JP. Okadaic acid induces matrix metalloproteinase-9 expression in fibroblasts: crosstalk between protein phosphatase inhibition and β-adrenoceptor signalling. Br. J. Pharmacol., 165, 274–288 (2012).
• 20) Tohgo A, Choy EW, Gesty-Palmer D, Pierce KL, Laporte S, Oakley RH, Caron MG, Lefkowitz RJ, Luttrell LM. The stability of the G protein-coupled receptor-beta-arrestin interaction determines the mechanism and functional consequence of ERK activation. J. Biol. Chem., 278, 6258–6267 (2003).
• 21) Edling CE, Selvaggi F, Ghonaim R, Maffucci T, Falasca M. Caffeine and the analog CGS 15943 inhibit cancer cell growth by targeting the phosphoinositide 3-kinase/Akt pathway. Cancer Biol. Ther., 15, 524–532 (2014).
• 22) Huang Y, Li X, Wang Y, Wang H, Huang C, Li J. Endoplasmic reticulum stress-induced hepatic stellate cell apoptosis through calcium-mediated JNK/P38 MAPK and Calpain/Caspase-12 pathways. Mol. Cell. Biochem., 394, 1–12 (2014).
• 23) Ping A, Yihao T, Jingxing D, Minkai C, Hesheng L. Ca²⁺/calmodulin-dependent protein kinase II mediates platelet-derived growth factor-induced human hepatic stellate cell proliferation. Dig. Dis. Sci., 57, 935–942 (2012).
• 24) Kawada N, Kuroki T, Kobayashi K, Inoue M, Kaneda K. Inhibition of myofibroblastic transformation of cultured rat hepatic stellate cells by methylxanthines and dibutyryl cAMP. Dig. Dis. Sci., 41, 1022–1029 (1996).
• 25) Ikeda N, Murata S, Maruyama T, Tamura T, Nozaki R, Kawasaki T, Fukunaga K, Oda T, Sasaki R, Homma M, Ohkohchi N. Platelet-derived adenosine 5′-triphosphate suppresses activation of human hepatic stellate cell: in vitro study. Hepatol. Res., 42, 91–102 (2012).
• 26) Sohail MA, Hashmi AZ, Hakim W, Watanabe A, Zipprich A, Groszmann RJ, Dranoff JA, Torok NJ, Mehal WZ. Adenosine induces loss of actin stress fibers and inhibits contraction in hepatic stellate cells via Rho inhibition. Hepatology, 49, 185–194 (2009).
• 27) Shim SG, Jun DW, Kim EK, Saeed WK, Lee KN, Lee HL, Lee OY, Choi HS, Yoon BC. Caffeine attenuates liver fibrosis via defective adhesion of hepatic stellate cells in cirrhotic model. J. Gastroenterol. Hepatol., 28, 1877–1884 (2013).
• 28) Wang H, Guan W, Yang W, Wang Q, Zhao H, Yang F, Lv X, Li J. Caffeine inhibits the activation of hepatic stellate cells induced by acetaldehyde via adenosine A2A receptor mediated by the cAMP/PKA/SRC/ERK1/2/P38 MAPK signal pathway. PLoS ONE, 9, e92482 (2014).
• 29) Wang Q, Dai X, Yang W, Wang H, Zhao H, Yang F, Yang Y, Li J, Lv X. Caffeine protects against alcohol-induced liver fibrosis by dampening the cAMP/PKA/CREB pathway in rat hepatic stellate cells. Int. Immunopharmacol., 25, 340–352 (2015).
• 30) Podzuweit T, Nennstiel P, Müller A. Isozyme selective inhibition of cGMP-stimulated cyclic nucleotide phosphodiesterases by erythro-9-(2-hydroxy-3-nonyl) adenine. Cell. Signal., 7, 733–738 (1995).
• 31) Reeves ML, Leigh BK, England PJ. The identification of a new cyclic nucleotide phosphodiesterase activity in human and guinea-pig cardiac ventricle. Implications for the mechanism of action of selective phosphodiesterase inhibitors. Biochem. J., 241, 535–541 (1987).
• 32) Tang KM, Jang EK, Haslam RJ. Photoaffinity labelling of cyclic GMP-inhibited phosphodiesterase (PDE III) in human and rat platelets and rat tissues: effects of phosphodiesterase inhibitors. Eur. J. Pharmacol., 268, 105–114 (1994).
• 33) Hasegawa M, Yamada S, Hirayama C. Fasting plasma caffeine level in cirrhotic patients: relation to plasma levels of catecholamines and renin activity. Hepatology, 10, 973–977 (1989).