ATP synthase, canonically mitochondrially located, is reported to be ectopically expressed on the plasma membrane outer face of several cell types. We analysed, for the first time, the expression and catalytic activities of the ecto- and mitochondrial ATP synthase during liver regeneration. Liver regeneration was induced in rats by two-thirds partial hepatectomy. The protein level and the ATP synthase and/or hydrolase activities of the hepatocyte ecto- and mitochondrial ATP synthase were analysed on freshly isolated hepatocytes and mitochondria from control, sham-operated and partial hepatectomized rats. During the priming phase of liver regeneration, 3 h after partial hepatectomy, liver mitochondria showed a marked lowering of the ATP synthase protein level that was reflected in the impairment of both ATP synthesis and hydrolysis. The ecto-ATP synthase level, in 3 h partial hepatectomized hepatocytes, was decreased similarly to the level of the mitochondrial ATP synthase, associated with a lowering of the ecto-ATP hydrolase activity coupled to proton influx. Noteworthily, the ecto-ATP synthase activity coupled to proton efflux was completely inhibited in 3 h partial hepatectomized hepatocytes, even in the presence of a marked intracellular acidification that would sustain it as in control and sham-operated hepatocytes. At the end of the liver regeneration, 7 days after partial hepatectomy, the level and the catalytic activities of the ecto- and mitochondrial ATP synthase reached the control and sham-operated values. The specific modulation of hepatocyte ecto-ATP synthase catalytic activities during liver regeneration priming phase may modulate the extracellular ADP/ATP levels and/or proton influx/efflux trafficking, making hepatocyte ecto-ATP synthase a candidate for a novel player in the liver regeneration process.
FoF1-ATP synthase (ATP synthase, EC 22.214.171.124) is a protonmotive multimeric enzyme localized in the coupling membrane of mitochondria, chloroplasts and bacteria, which synthesizes or hydrolyses ATP depending on the energy state of the membrane. For ATP synthesis, the ATP synthase uses the mitochondrial transmembrane proton gradient (ΔμH+) generated by the respiratory chain, whereas ATP hydrolysis produces a proton gradient [1–3]. ATP synthase subunits are localized on the plasma membrane of different mammalian cell lines [4–6], overcoming initial doubts concerning their ectopic presence.
Currently, the role of the plasma membrane (ecto) ATP synthase together with its upstream and downstream effectors is not uniquely defined, as it has been shown to be highly dependent on the cell type and the culture conditions. Although the mitochondrial ATP synthase is primarily committed to aerobic ATP synthesis , the ecto enzyme has been involved in various biological functions including angiogenesis, lipoprotein metabolism, immune recognition of tumours [7,8] and the pathogenesis of different diseases [9–12].
The liver is the largest gland in the body, and hepatocytes, which carry out most of the synthetic and metabolic liver functions, make up its cellular bulk, accounting for ∼80% of all the tissue cells . Liver has an extremely high capacity for regeneration, the importance of which in humans is rising together with the incidence of liver pathologies (e.g. cancer, ischaemic injury and trauma). Since the pioneering work of Higgins and Anderson , LR (liver regeneration), induced by the surgical resection in rats of a large tissue portion [i.e. PH (partial hepatectomy)], is still a state-of-the-art experimental model in which hepatocyte proliferation recovers the original mass of the organ. Hepatocyte regeneration after PH is an orchestrated process not yet fully understood, spanning 1 min to 7 days, involving cellular, morphological, biochemical and metabolic changes that proceed along a linked sequence of three distinct phases: priming, proliferation and termination [15,16]. Specific tunings of the catalytic activities and expression level of the mitochondrial ATP synthase have been reported during the different phases of LR [17,18].
Our group has characterized previously the ecto-ATP synthase on the plasma membrane of quiescent rat hepatocytes  (referred to in the present paper as control hepatocytes), reporting a similar subunits composition as the mitochondrial complex. We have demonstrated that, as for the mitochon-drial enzyme, the proton efflux and proton influx across the hepatocyte plasma membrane are coupled respectively to ecto-ATP synthesis and hydrolysis. Hepatocyte ecto-ATP synthase, with the F1 sector protruding towards the extracellular side, has an orientation opposite that of mitochondrial ATP synthase, whose F1 sector is located in the mitochondrial matrix. Such orientation implies that ATP synthesis and/or hydrolysis occurs in the extracellular milieu of the hepatocytes, thus potentially affecting extracellular ADP/ATP levels and/or proton influx/efflux trafficking.
The present study represents the first concerning the presence of ATP synthase on the plasma membrane of regenerating liver cells. We investigated, for the first time, the protein level and ATP synthase/hydrolase catalytic activities of the ecto- and mitochondrial ATP synthase during the LR process. We found: (i) the depression of the content and protonmotive activities of the ecto-ATP synthase, with the specific inhibition of the extracellular ATP synthase activity, in the LR priming phase, and (ii) their full recovery in the LR termination phase. Additionally, we report a similar lowering of the ecto- and mitochondrial ATP synthase level indicating the presence of a common expression system. Our results provide novel clues as to the possible role of the ecto-ATP synthase enzyme in the LR process.
MATERIALS AND METHODS
Oligomycin, CCCP (carbonyl cyanide m-chlorophenylhydra-zone), Ap5A [P1,P5-di(adenosine-5′)pentaphosphate], collagenase Type I and BSA (fraction V fatty-acid-free) were from Sigma–Aldrich; NADH, NADP, pyruvate kinase, lactate dehydrogenase, glucose-6-phosphate dehydrogenase, hexokinase, ATP, ADP and PEP (phosphoenolpyruvate) were from Roche; acrylamide, N,N′-methylenebisacrylamide, SDS and Bradford protein assay were from Bio-Rad Laboratories; ACMA (9-amino-6-chloro-2-methoxyacridine), Alexa Fluor® 488-conjugated goat anti-rabbit antibody and HRP (horseradish peroxidase)-conjugated goat anti-rabbit were from Molecular Probes/Invitrogen; antibody against mitochondrial VDAC (voltage-dependent anion channel) and anti-pan-cadherin were from Abcam. All other chemical reagents were of analytical grade.
Animal and experimental protocol
Male Wistar rats (200–250 g) (Harlan Laboratories) were allowed free access to food and water. Rats were maintained in a temperature-controlled room with a 12 h light/12 h dark illumination cycle. Animals under general anaesthesia, underwent PH by removing 70% of the liver (the right median, left lateral and median liver lobes) or only opening the abdominal skin in the case of sham-operated animals . Surgery was performed at 08.30–09.30 h. The control (non-operated), sham-operated and PH rats were killed 3 h or 7 days later. Completeness of LR was evaluated by monitoring liver weight recovery at the time of killing. Animal handling was carried out in agreement with the standards stated in the principles expressed in the Declaration of Helsinki, and under the authorization of the ‘Comitato Etico per la sperimentazione animale’ (CESA) of University of Bari “Aldo Moro”.
Preparation of rat liver mitochondria and plasma membrane
Livers were rapidly processed and RLM (rat liver mitochondria) were isolated by differential centrifugation essentially as reported previously . The mitochondrial suspension, deprived of peroxisomes as judged from measurements of catalase activity (results not shown), was immediately used for the subsequent experiments. Plasma membranes were isolated from rat livers as reported in . Briefly, livers were weighed, minced in four volumes of 0.25 M sucrose, 5 mM Tris/HCl (pH 7.4) and 1 mM MgCl2 (buffer A) and homogenized in a 10 ml Potter–Elvehjem homogenizer using a tight-fitting Teflon-coated pestle. The homogenate was filtered through four layers of gauze and the filtrate was centrifuged at 280 g for 5 min. The supernatant was saved and the pellet was resuspended in buffer A and again centrifuged as above. The resulting supernatant was added to the first one and centrifuged at 1500 g for 10 min. The pellet was resuspended in buffer A and 2 M sucrose in buffer A was added, obtaining a final sucrose concentration of 1.42 M. Then, 4 ml of the sample was added to polyallomer tubes and overlaid with 0.8 ml of 0.25 M buffer A. After ultracentrifugation at 82000 g for 60 min with a swing-out rotor, the pellicle at the interface was collected, resuspended in buffer A and centrifuged at 12000 g for 10 min. The final pellet was used as the plasma membrane fraction.
Freshly isolated hepatocyte (referred to hereinafter as hepatocytes) suspensions were isolated by liver perfusion with collagenase buffer and purified essentially according to Seglen , as modified in . The integrity and viability of the cells were confirmed with a Trypan Blue-exclusion test and at least 95% of the cells remained viable at the end of the preparation of which 80–90% were in the form of single cells.
Spectrophotometric determination of ATP synthase activity
The rate of mitochondrial ATP synthesis was determined in freshly isolated RLM, as reported in , and spectrophotometric assay of extracellular ATP synthesis in hepatocyte suspensions was performed as described previously . In detail, hepatocytes (∼106) were incubated at 37°C in a mixture containing 10 mM Hepes (pH 7.4), 150 mM NaCl, 2 mM MgCl2, 1 mM EDTA, 20 mM glucose, 4 units/ml hexokinase and 300 μM Ap5A, an adenylate kinase inhibitor. After 4 min, 100 μM ADP alone or with 5 mM Pi was added, and the reaction was stopped by centrifugation. The supernatant was added to a mixture containing 1 mM MgCl2, 150 mM Tris/HCl (pH 7.4) and 7 units/ml of glucose-6-phosphate dehydrogenase. NADP at 1 mM was added and glucose 6-phosphate levels were determined following NADP reduction using a dual-wavelength spectrophotometer (Beckman Coulter DU 800, wavelength couple 360/374 nm with ε=2.01 mM−1).
All enzymatic analyses were performed in the presence and/or absence of oligomycin, the specific inhibitor of ATP synthase .
SDS/PAGE and Western blot procedure
SDS/PAGE was performed on a slab gel with a 12–20% linear gradient of polyacrylamide. Amounts of 30 μg of mitochondrial and 30 μg of plasma membrane proteins were subjected to SDS/PAGE, electrotransferred on to nitrocellulose and incubated overnight at 4°C with rabbit polyclonal antibody against anti F1-α/β subunits as described in . Antibody directed against VDAC was used as a loading control for mitochondria, whereas antibody directed against cadherin was used as a plasma membrane marker and loading control. Densitometric analysis was performed using a Bio-Rad Laboratories VersaDoc Imaging System.
Confocal microscopy imaging
Confocal analysis was performed as reported previously , with minor modifications. Briefly, hepatocytes were kept at 4°C, fixed in 4% (w/v) formaldehyde in PBS for 15 min, washed twice with ice-cold PBS and incubated for 15 min with 2% (v/v) goat serum in PBS (pH 7.4). Cells were washed three times and incubated for 1 h at 4°C with antibody against the F1-α/β ATP synthase (1:300 dilution). Hepatocytes were then washed in PBS and incubated with goat anti-rabbit IgG conjugated to Alexa Fluor® 488 (diluted 1:300 in 2% goat serum in PBS) for 1 h at 4°C in the dark; finally, cells were washed three times with PBS. Appropriate controls, with permeabilized cells incubated either with the primary or the secondary antibody alone, were included. Immunolabelled hepatocytes were analysed under a confocal laser-scanning microscope (Leica TCS SP8) and images were recorded using a ×40 objective (numerical aperture 1.3).
Three-dimensional analysis of rat hepatocytes was performed with 1024 pixels×1024 pixels and scan speed of 200 Hz and Z-stacks were acquired with a z-step size of 1 μm for a total z-volume of 50 μm to cover the entire thickness of the cells, and 3D reconstruction was obtained with the 3D View module of LasAF software.
Flow cytometry analysis (FACS)
FACS analysis was performed using as primary the antibody against the F1-α/β subunit of ATP-synthase as described in . Briefly, after excitation at a wavelength of 488 nm, MFI (median fluorescence intensity) was determined for each sample on a FACSCanto II instrument (BD Biosciences) and analysed using CellQuest software (Becton Dickinson).
Measurement of hepatocyte intracellular acidification
Proton translocation across the hepatocyte plasma membrane was monitored using ACMA fluorescence quenching (excitation at 410 nm, emission at 490 nm) using a Jasco FP6200 spectrofluorimeter as reported in . A buffer containing 10 mM Hepes (pH 7.4), 150 mM NaCl, 2 mM MgCl2, 1 mM EDTA and 300 μM Ap5A was employed, and the instrument was set to zero. Then hepatocytes, with and without the addition of 20 mM glucose, were added for 4 min at 37°C and after the addition of 2 μM of ACMA. The recorded trace of hepatocytes alone was subtracted from that measured after ACMA addition. The addition of the ionophore CCCP, dissipating the transmembrane ΔpH, restored the fluorescence.
Ecto-ATP synthesis and hydrolysis activities coupled with plasmatic proton flux
Extracellular ATP synthesis and/or hydrolysis associated with proton conduction across the plasma membrane was monitored following ACMA fluorescence measurements (excitation at 410 nm, emission at 490 nm) using a Jasco FP6200 spectrofluorimeter as reported in . The reaction mixture used for ecto-ATP synthesis analysis was the same reported for the spectrophotometric determination with the exception of hexokinase, glucose-6-phosphate dehydrogenase and NADP. The reaction buffer used for ATP hydrolysis was 100 mM NaCl, 20 mM KCl, 5 mM MgCl2, 0.5 mM EGTA and 40 mM Tris/HCl (pH 7.4). Proton conduction during ATP synthesis and hydrolysis was determined by the addition to the reaction buffer of ADP+Pi or ATP respectively, and the instrument was set to zero. Hepatocytes were added and incubated for 4 min at 37°C in the ATP synthesizing and/or hydrolysing buffer with or without oligomycin, and then 2 μM ACMA was pulsed. The recorded trace of hepatocytes alone was subtracted from that measured after ACMA addition. The addition of CCCP, dissipating the transmembrane ΔpH, restored the fluorescence.
Lactate dehydrogenase activity
Hepatocytes were frozen in liquid nitrogen three times, the cell lysate obtained was centrifuged and the supernatant was used immediately for the assay. LDH (lactate dehydrogenase) activity was measured at 340 nm and 30°C using 100 μg of cell supernatant in a 50 mM Tris/HCl buffer (pH 7.4), with 1.5 μg/ml rotenone, 2 mM pyruvate and 0.1mM NADH. LDH activity is expressed as nmol of NADH oxidized/min per mg of protein.
Protein concentration was determined by using the Bradford protein assay (Bio-Rad Laboratories) with BSA as a standard.
Results were computed using Microsoft Excel. Comparison was made using one-way ANOVA followed by a post-hoc Turkey's B test. All statistical analyses were performed using SPSS software. Differences were considered statistically significant at P<0.05.
Mitochondrial ATP synthase level and catalytic activities are down-regulated 3 h after partial hepatectomy
We first looked at mitochondrial ATP synthase, and found a marked decrease (∼37%) of the F1-α/β subunit content, associated with a lowering of the ATP synthase (∼40%) and hydrolase (∼35%) oligomycin-sensitive enzymatic activities in 3 h PH RLM, compared with control and sham-operated samples (Figure 1). On the other hand, 7 days after PH, mitochondria, during the LR quiescent phase, were restored to control and sham levels of the ATP synthase catalytic activities and protein level (Figures 1A and 1C).
Mitochondrial ATP synthase catalytic activities and content in the early (3 h) and termination (7 days) phases of liver regeneration
Localization and reduced content of the ecto-ATP synthase in 3 h PH hepatocytes
In the present study, we employed isolated hepatocytes obtained after in situ liver perfusion with a collagenase solution, which represents the most appropriate procedure to obtain excellent yields (∼95%) of viable cells [22,27]. Following isolation, hepatocytes from sham-operated and 3 h PH rats were incubated first with anti-F1-α/β antibody then with fluorochrome-conjugated secondary antibody, and visualized by laser-scanning confocal microscopy. Ecto-ATP synthase is clearly localized on the plasma membrane of both sham-operated and 3 h PH hepatocytes (Figures 2A–2F), confirming our previous finding in control hepatocytes . In particular, confocal images showed extensive fluorescence at the surface of sham hepatocytes (Figures 2A–2C), whereas, even though not quantified, it is strongly reduced at the 3 h PH hepatocytes plasma membrane (Figures 2D–2F). No fluorescence was present in confocal sections inside the cells. It should be noted that cells were not permeabilized and only the cell surface was exposed to antibodies. In digitonin-permeabilized 3 h PH hepatocytes, a strong fluorescence increase, due to the mitochondrial F1-α/β ATP synthase, was observed (Figures 2G–2I), confirming the surface localization of the antibody-linked fluorescence in intact hepatocytes. The presence of the ecto-ATP synthase on the 3 h PH hepatocyte plasma membrane was confirmed by 3D reconstruction of serial optical sections (Supplementary Video S1).
ATP synthase localization in sham-operated and 3 h PH hepatocytes
Then, we performed quantitative flow cytometry analysis of hepatocytes incubated with anti-F1-α/β antibody. The MFI values obtained showed a decrease of nearly 40% in 3 h PH hepatocytes with respect to control and sham-operated hepatocytes, which presented similar values (Figure 3). These results were confirmed by Western blot analysis on plasma membrane isolated from whole liver, carried out with antibodies against F1-α/β and cadherin, a well-known plasma membrane marker that was used as loading control (Supplementary Figure S1). In detail, we found a decrease of nearly 38% in 3 h PH plasma membranes, compared with control and sham-operated hepatocytes that was restored in 7 days PH plasma membranes. It should be noted that these results could not be ascribed only to hepatocyte ecto-ATP synthase as the plasma membranes were prepared from whole liver. We used a monoclonal antibody against COX (cytochrome oxidase) IV as a marker of the mitochondrial inner membrane and no immunoreaction in the plasma membrane lanes was observed (Supplementary Figure S1), thus showing the absence of mitochondrial contamination in the plasma membrane fraction.
Ecto-ATP synthase expression is down-regulated in 3 h PH hepatocytes
Hepatocytes intracellular proton concentration is modulated during liver regeneration
Intracellular acidification was investigated by monitoring the fluorescence of ACMA added to the hepatocyte suspensions as described in . Intracellular pH decrease leads to intracellular accumulation of ACMA, which results in a proportional fluorescence quenching . The ACMA analysis was carried out on hepatocyte suspensions in the absence and/or in the presence of 20 mM glucose, whose addition induces a marked glycolytic-induced intracellular acidification . Incubation of control and sham-operated hepatocytes with glucose led to a sustained intracellular acidification, which was always reversed upon addition of the membrane-permeant CCCP protonophore (Figures 4A and 4B), confirming the integrity of the hepatocyte isolations. A large intracellular acidification took place in 3 h PH hepatocytes, compared with control and sham hepatocytes, which was not increased further by the addition of glucose (Figure 4C). In 7 days PH hepatocytes, the intracellular acidification returned to control and sham-operated levels and was again sensitive to glucose addition (Figure 4D). The sustained intracellular acidification during the LR priming phase was confirmed by doubled LDH activity in 3 h PH hepatocytes, compared with control and sham hepatocytes, followed by its normalization at 7 days after PH (Figure 4E).
Intracellular acidification during LR
Modulation of the ecto-ATP synthase protonmotive activities during liver regeneration
As we reported previously that hepatocyte ecto-ATP synthase works in ATP synthesizing and hydrolysing modes , pumping protons inside/out of the cell, we measured the ecto-ATP synthase and hydrolase activities coupled to proton translocation. The activities and vectoriality of the protonmotive ecto-ATP synthase were determined by following the time course of ACMA fluorescence changes induced by the external addition of ADP (plus Pi) or ATP to the hepatocyte suspension. It should be noted that both the ACMA fluorescence increase, caused by external addition of ADP (plus Pi) (Figure 5), and the ACMA fluorescence quenching, caused by external addition of ATP to the hepatocytes (Figure 6), were completely suppressed by the addition of oligomycin.
Hepatocyte intracellular alkalinization (proton efflux) driven by ecto-ATP synthase is suppressed in 3 h PH hepatocytes
The ACMA fluorescence quenching caused by intracellular acidification (i.e. proton influx) driven by hydrolysis of externally added ATP to ADP and Pi by the ecto-ATP synthase, observed in control and sham-operated hepatocytes (Figures 6A and 6B), was decreased in 3 h PH hepatocytes (Figure 6C). In control and sham-operated hepatocytes, we observed a marked ACMA fluorescence increase due to ATP synthesis from ADP plus Pi associated with proton efflux (Figures 5A and 5B), by the ecto-ATP synthase (i.e. intracellular alkalinization), that was completely inhibited in 3 h PH hepatocytes (Figure 5C).
Hepatocyte intracellular acidification (proton influx) driven by ecto-ATP hydrolase is reduced in 3 h PH hepatocytes
In order to confirm these relevant findings, the ecto-ATP synthase activity was also studied by direct spectrophotometric analysis of extracellular ATP synthesis. A sustained oligomycin-sensitive ecto-ATP synthase activity was found in control and sham-operated hepatocytes that was enhanced by glucose-induced intracellular acidification (Figure 7). The oligomycin-sensitive extracellular ATP synthesis was inhibited in 3 h PH hepatocytes, both in the absence and in the presence of glucose and restored in 7 days PH hepatocytes (Figure 7). The ecto-ATP synthesis observed in the presence of ADP alone was not influenced by oligomycin, and did not change under any of the experimental conditions investigated (Figure 7). This extracellular ATP could be ascribed to both ecto-AK (adenylate kinase) and/or ecto-NDPK (nucleoside diphosphokinase) activity that have been reported to be present in plasma membranes [5,30].
Extracellular ATP synthesis with or without glucose incubation
It should be noted that all the experiments related to control and sham-operated measurements always showed similar values, thus excluding any surgery-related effects.
LR is a crucial process, from both physiological and pathological perspectives, by which the organ responds to a loss of tissue mass, such as PH, or injury [16,31]. Despite the huge amount of papers published on LR, owing to the high complexity of the processes involved, the players and the mechanisms underlying LR have not yet been fully elucidated. The early phase of LR (0–4 h after PH), known as the priming phase, is critically related to the liver's extraordinary ability to accurately restore its original size as hepatocytes acquire the competencies to allow the transition from the quiescence state (G0) to the replicative phase (G1) [31,32].
To the best of our knowledge, no study has been performed on the ecto-ATP synthase presence and/or enzymatic functions during LR. In the present study, we focused on the LR priming phase since, as the interval between surgery and the earliest recorded events associated with LR shortens, the number of candidates responsible for inducing hepatocytes to enter the cell cycle decreases. Hepatocytes in suspension were employed as this experimental model represents the closest to in vivo conditions retaining the tissue cell phenotype, in particular preserving the properties of the hepatocytes plasma membrane surface enzymes . Moreover, the use of hepatocytes in suspension, avoiding any possible artefacts induced by the cell culture conditions, gave the possibility to perform robust enzymatic and/or expression analyses that would be unfeasible using liver sections.
The precise mechanism by which ATP synthase reaches the plasma membrane is still unknown ; however, preliminary evidence has been reported that the ectopic expression of ATP synthase is a consequence of translocation from mitochondria [4,34]. In the present study, we found that ecto-ATP synthase protein level mirrored the mitochondrial ATP synthase during the time points of LR investigated, so it is conceivable that the two enzymes share a common genetic and expression system. The reported ATP synthase protein decrease in 3 h PH mitochondria is consistent with the decrease in mitochondrial ATP synthase level and catalytic activities during other time points of the LR early phase [18,35,36]. It has been reported that the ATP synthase decrease could be ascribed to a lowered translation efficiency of the mRNA transcripts . Our findings suggest that the enhanced proteolytic digestion of the ATP synthase is less likely to occur, considering that we observed the protein decreases in two different membrane compartments, i.e. the matrix side of the inner mitochondrial membrane and the outer surface of the plasma membrane. The full recovery of the mitochondrial ATP synthase level and enzymatic activities has been reported in LR 4 days after PH (i.e. during the termination phase) [17,18], thus confirming our results at 7 days after PH.
It has been proposed that, depending on the cell type, the catalytic activities of the ecto-ATP synthase might promote and/or suppress various biological processes [5–8]. In the present study, we provide the first evidence that, under specific cellular events (i.e. the ones induced by LR priming phase), hepatocyte ecto-ATP synthase modulates its catalytic functions, especially extracellular ATP synthesis. In the liver, extracellular ATP and/or ATP-derived products, such as ADP, function as an autocrine/paracrine signal modulating a broad range of cellular functions such as cell shrinkage, stimulation of glucose release and protein catabolism through activation of plasma membrane purinergic receptors (i.e. P2Y and P2X receptors). Several lines of evidence highlight the importance of extracellular ATP level as an early regulator of LR process [37–39]. It has been reported that ATP was released in the extracellular milieu of hepatocytes, during the priming phase of LR, not in response to a change in cellular energy demand, but as an early stress signal contributing to LR . In addition, extracellular ATP sustains the entering of remnant quiescent hepatocytes into the cell cycle during LR priming phase, via activation of P2Y2 purinergic receptors, by stimulating cellular signalling pathways [e.g. p42/44 MAPK (mitogen-activated protein kinase) ERK (extracellular-signal-regulated kinase)] and increasing the cellular level of cytokines [e.g. TNFα (tumour necrosis factor α), IL-6 (interleukin 6) and NF-κB (nuclear factor κB)] and transcription factors [e.g. STAT3 (signal transducer and activator of transcription 3)] [38,39]. Interestingly, it has been proposed that an enhanced extracellular ATP hydrolysis by plasma ecto-diphosphohydrolase (EC 126.96.36.199) promotes the LR process by inhibiting extracellular ATP levels through rat NK (natural killer) cell inactivation .
At the same time, the balance between ecto-ATP synthase and hydrolase activity and the consequent changes in proton efflux and influx modulates the intra/extra-cellular hepatocyte pH, and it may be critical for the course of LR. The maintenance of normal cell metabolism and proper protein functions is closely related to tight regulation of intra-/extra-cellular pH within a narrow range, mainly by the activity of transporters located at the plasma membrane [41,42]. Changes in cellular pH have also been proposed to initiate cellular proliferation, to control and direct cell migration and to trigger apoptosis . Noteworthily, it has been reported that hepatocyte plasma membrane enzymes (e.g. Na+/K+-ATPase, Na+/H+ exchanger) modulate their catalytic activities and/or expression levels in the early phase of LR [44–47]. These changes, associated with an alteration of the hepatocyte plasma membrane potential, are thought to be related to the cell-mediated remodelling that occurs in the early phase of LR before the division of hepatocytes, involving the initiation of DNA synthesis and the triggering of cellular activation . In 3 h PH hepatocytes, we found an increase in both intracellular acidification, which was independent of glucose addition to the cells, and LDH enzymatic activity. This large intracellular pH decrease, characteristic of undifferentiated cells, is consistent with the reported 3-fold increase in the lactate/pyruvate ratios during the LR priming phase . It has been reported that hepatocytes, in the LR priming phase, undergo an early transitory retrodifferentiation shift towards a fetal phenotype .
The regulatory roles of the ecto-ATP synthase are impaired in 3 h PH hepatocytes, in which the activities of the protonmotive ecto-ATP synthase are severely reduced, with the specific loss of extracellular ATP synthesis. These findings are in good agreement with Chi and Pizzo  who observed in tumour cell lines that the ecto-ATP synthase inhibits its extracellular ATP synthesis activity in the presence of intracellular acidification, hampering cell survival. On the other hand, the observed lowering of the extracellular ATP hydrolysis during LR priming phase might be explained, similarly to what happens in mitochondria, by the decrease in the ecto-ATP synthase protein level.
Underlining the functional importance of our findings, multiple lines of evidence indicate that the mitochondrial ATP synthase specifically triggers its catalytic activities of ATP synthesis and/or hydrolysis, taking one route or another, in several pathophysiological conditions [50,51]. Noteworthily, more than 270 inhibitors of mitochondrial ATP synthase are known, proposing this key enzyme as a promising molecular target for drugs in the treatment of various pathologies and in the regulation of energy metabolism . However, there is still room for further research into additional time points of the LR process (e.g. replicative phase), and mechanistic investigations by silencing and/or overexpressing the ATP synthase in knockout/knockin rat models.
From the data of the present study and on the basis of the orientation of the ecto-ATP synthase, we propose a model (Figure 8) in which the inhibition of extracellular ATP synthesis might contribute to: (i) entrap protons inside the hepatocytes influencing intra/extra-cellular pH, and (ii) modulate the extracellular ATP and/or ATP-derived products levels with the related signalling pathways, thereby allowing hepatocytes to respond to specific cellular stimuli encountered in the LR priming phase. Connecting the multiple roles of the ecto-ATP synthase in various biological processes, in particular cell replication and modulation of the intracellular pH, the observed transitory depression of the protonmotive ecto-ATP synthase can have a definite impact on LR processes. Our results open new insights regarding the processes involved in LR and make the ecto-ATP synthase a candidate for a new potential target for therapeutic modulation of liver growth in health and disease.
Proposed model of the leading hypotheses regarding liver regeneration mediated by the ecto-ATP synthase complex
Antonio Gnoni designed the research. Franco Zanotti, Sergio Papa, Anna Maria Sardanelli and Antonio Gnoni supervised the research. Federica Taurino, Caterina Giannoccaro, Alessandro Cavallo, Elisa De Luca, Salvatore Santacroce, Anna Maria Sardanelli and Antonio Gnoni performed the research. Federica Taurino, Caterina Giannoccaro, Franco Zanotti and Antonio Gnoni analysed the data. Federica Taurino, Elisa De Luca and Antonio Gnoni wrote the paper. All authors read and approved the final paper.
These authors contributed equally to this work.