To divide or not to divide: revisiting liver regeneration

The liver has a remarkable capacity to regenerate. Even with surgical removal (partial hepatectomy) of 70% of liver mass, the remnant tissue grows to recover the original mass and functions. Liver regeneration after partial hepatectomy has been studied extensively since the 19th century, establishing the long-standing model that hepatocytes, which account for most of the liver weight, proliferate to recover the original mass of the liver. The basis of this model is the fact that almost all hepatocytes undergo S phase, as shown by the incorporation of radioactive nucleotides during liver regeneration. However, DNA replication does not necessarily indicate the execution of cell division, and a possible change in hepatocyte size is not considered in the model. In addition, as 15–30% of hepatocytes in adult liver are binuclear, the difference in nuclear number may affect the mode of cell division during regeneration. Thus, the traditional model seems to be oversimplified. Recently, we developed new techniques to investigate the process of liver regeneration, and revealed interesting features of hepatocytes. In this review, we first provide a historical overview of how the widely accepted model of liver regeneration was established and then discuss some overlooked observations together with our recent findings. Finally, we describe the revised model and perspectives on liver regeneration research.     

Cell density-dependent regulation of albumin synthesis and DNA synthesis in rat hepatocytes by hepatocyte growth factor.

Hepatocyte growth factor (HGF), a potent mitogen for mature hepatocytes, has been considered to act as a hepatotropic factor for liver regeneration. We examined the effect of HGF on albumin synthesis and DNA synthesis of adult rat hepatocytes cultured at various cell densities. HGF stimulated albumin synthesis of hepatocytes by 40-60% when they were cultured at higher cell densities such that there was tight cell-cell contact. But at lower cell densities HGF failed to stimulate albumin synthesis. In contrast, the stimulatory effect of HGF on DNA synthesis of hepatocytes was more potent at lower than at higher cell densities: HGF did not stimulate DNA synthesis of hepatocytes cultured at confluent cell density. Thus, HGF seems to stimulate both albumin synthesis and DNA synthesis of hepatocytes, in a reciprocal relationship depending on cell density. When the effects of various cytokines were examined, epidermal growth factor, transforming growth factor-alpha, and acidic fibroblast growth factor also stimulated albumin synthesis by 20-30%. However, transforming growth factor-beta 1, basic fibroblast growth factor, and interleukin-1 beta had no effect on albumin synthesis, while interleukin-6 inhibited it by 42%. Thus HGF was the most potent in stimulating albumin synthesis in these cytokines. Since HGF is markedly increased in the liver or plasma following various liver insults, HGF may be involved in liver regeneration through the potential to stimulate both cell growth and liver-specific functions such as albumin synthesis in a cell density-dependent manner.     

An integrative model links multiple inputs and signaling pathways to the onset of DNA synthesis in hepatocytes

During liver regeneration, quiescent hepatocytes re-enter the cell cycle to proliferate and compensate for lost tissue. Multiple signals including hepatocyte growth factor, epidermal growth factor, tumor necrosis factor?α, interleukin-6, insulin and transforming growth factor?β orchestrate these responses and are integrated during the G(1) phase of the cell cycle. To investigate how these inputs influence DNA synthesis as a measure for proliferation, we established a large-scale integrated logical model connecting multiple signaling pathways and the cell cycle. We constructed our model based upon established literature knowledge, and successively improved and validated its structure using hepatocyte-specific literature as well as experimental DNA synthesis data. Model analyses showed that activation of the mitogen-activated protein kinase and phosphatidylinositol?3-kinase pathways was sufficient and necessary for triggering DNA synthesis. In addition, we identified key species in these pathways that mediate DNA replication. Our model predicted oncogenic mutations that were compared with the COSMIC database, and proposed intervention targets to block hepatocyte growth factor-induced DNA synthesis, which we validated experimentally. Our integrative approach demonstrates that, despite the complexity and size of the underlying interlaced network, logical modeling enables an integrative understanding of signaling-controlled proliferation at the cellular level, and thus can provide intervention strategies for distinct perturbation scenarios at various regulatory levels.     

Interleukin-1 beta is a potent growth inhibitor of adult rat hepatocytes in primary culture

Interleukin-1 beta (IL-1 beta) strongly inhibited DNA synthesis of adult rat hepatocytes in primary culture stimulated by insulin and epidermal growth factor (EGF). Its effect was dose-dependent and was maximal at 2 ng/ml. IL-1 beta had no cytotoxic effect but changed the cells from a flat to a spindle shape as shown by phase-contrast microscopy. The inhibition of DNA synthesis by IL-1 beta was closely correlated with a decrease in the labeling index. This inhibitory effect was observed only when IL-1 beta was added for 10 h to cultured hepatocytes in the G1 phase within 12 h after addition of insulin and EGF: it was not observed in the S phase, which starts about 24 h after addition of the mitogens. Exposure of the hepatocytes to IL-1 beta for two 1-h periods, one at an early stage (0-6 h) and one at a late stage (6-12 h) of the G1 phase, resulted in the same marked inhibition of DNA synthesis as exposure to IL-1 beta for 10 h in the G1 phase. This requirement of IL-1 beta at two stages in the G1 phase for inhibition of DNA synthesis of hepatocytes is different from that with transforming growth factor-beta, which is required for only 1 h in the early G1 phase for a similar inhibition. These findings suggest that IL-1 beta acts at two distinct stages in the G1 phase and that its cooperative actions are necessary to inhibit growth of adult rat hepatocytes in primary culture. Other cytokines, such as IL-6/B-cell stimulating factor-2, were less potent, but caused significant inhibition of DNA synthesis of adult rat hepatocytes at 2 ng/ml, whereas IL-2 and tumor necrosis factor did not affect hepatocyte growth. From these results it is suggested that Kupffer cells in liver lobules and macrophages in the blood may play important roles, mainly via IL-1, in repair of liver damage and regeneration.     

Blocking of G1/S transition and cell death in the regenerating liver of Hepatitis B virus X protein transgenic mice

The Hepatitis B virus X (HBx) protein has been strongly implicated in the carcinogenesis of hepatocellular carcinoma (HCC). However, effects of the HBx protein on cell proliferation and cell death are controversial. This study investigates the effects of the HBx protein on liver regeneration in two independent lines of HBx transgenic mice, which developed HCC at around 14 to 16 months of age. High mortality, lower liver mass restoration, and impaired liver regeneration were found in the HBx transgenic mice post-hepatectomy. The levels of alanine aminotransferase and alpha-fetoprotein detected post-hepatectomy increased significantly in the HBx transgenic livers, indicating that they were more susceptible to damage during the regenerative process. Prolonged activation of the immediate-early genes in the HBx transgenic livers suggested that the HBx protein creates a strong effect by promoting the transition of the quiescent hepatocytes from G0 to G1 phase. However, impaired DNA synthesis and mitosis, as well as inhibited activation of G1, S, and G2/M markers, were detected. These results indicated that HBx protein exerted strong growth arrest on hepatocytes and imbalanced cell-cycle progression resulting in the abnormal cell death; this was accompanied by severe fat accumulation and impaired glycogen storage in the HBx transgenic livers. In conclusion, this study provides the first physiological evidence that HBx protein blocks G1/S transition of the hepatocyte cell-cycle progression and causes both a failure of liver functionality and cell death in the regenerating liver of the HBx transgenic mice.     

Role of cell-surface modulator of DNA synthesis in liver regeneration

A cell-surface modulator of DNA synthesis by cultured rat hepatocytes was studied in relation to the liver regeneration process. When rat hepatocytes isolated 24 h after partial hepatectomy were cultured, the first burst of DNA synthesis peaked at 5-8 h and declined until 24 h, followed by the second burst. Rat liver plasma membranes added 2 h after plating inhibited only the second burst, while in the case of the normal hepatocytes where the DNA synthesis began to increase after 5 h, this inhibition was observed at 16 h but not at 8 h. The inhibition did not differ when the membranes obtained from regenerating livers 12 h after partial hepatectomy were used. Epidermal growth factor binding to the cultured hepatocytes was not hindered by the membranes. These results suggest that the modulator inhibits hepatocyte proliferation at the early G1-phase of the cell cycle and that its action might be controlled by other factors in the process of liver regeneration.     

Deciliation interferes with cell-cycle progression in Tetrahymena

The impact of ciliary regeneration upon cell-cycle progression of the ciliate Tetrahymena was studied. It was found that cell division ceases during ciliary regeneration, and starts again about 4 h after deciliation. Deciliation of an asynchronously multiplying culture results in a rapid interruption of DNA synthesis, followed by resumption 1 h later. This was shown by pulse-labelling the cells with [3H]thymidine at various times after deciliation. Cytophotometric determinations of the macronuclear DNA content substantiated these observations, since the average DNA content per cell remained constant within the first hour of regeneration, confirming the labelling experiments, after which it rose. At its maximum, the average DNA content was more than doubled as compared with the beginning of the experiment. This indicates that a substantial proportion of the regenerating cells performed two rounds of DNA replication prior to cell division. The massive drop in the average DNA content during the fifth hour after deciliation indicates that the culture becomes partly synchronized for cell division by the deciliation procedure. The division synchrony results from a greater delay of the next cell division when G2 cells are deciliated than occurs in G1 cells. This was shown by deciliating cultures of Tetrahymena thermophila cells in the respective stages of the cell cycle, which had been partly synchronized by elutriator centrifugation. Thus, deciliation followed by ciliary regeneration causes a varying degree of retardation in progression through the cell cycle, being greatest for G2 cells and least for G1 cells.     

Tumor necrosis factor stimulates DNA synthesis of mouse hepatocytes in primary culture and is suppressed by transforming growth factor beta and interleukin 6.

In a previous study, we revealed that tumor necrosis factor (TNF) was secreted in mouse liver at an early phase of liver regeneration after partial hepatectomy. Here, we investigated direct actions of TNF on the in vitro DNA synthesis of adult mouse hepatocytes in primary culture. TNF enhanced both 3H-TdR uptake and the number of 3H-TdR-labeled nuclei of hepatocytes. Their time courses were similar to those by epidermal growth factor (EGF) with about a 15 h lag period and a peak period of 24-48 h. This action of TNF was abrogated by DNA polymerase alpha inhibitor, aphidicolin and blocked specifically by anti-TNF antibody. The actions of rmTNF and rhTNF were not distinguishable; ED50 was about 7.5U/ml (5ng/ml) and 30U/ml (20ng/ml) for maximal response (about 2-fold or more of control). Other inflammatory monokines showed differential effects on in vitro DNA synthesis of hepatocyte. Neither type of interleukin 1 affected hepatocyte DNA synthesis in the range examined (up to 50 ng/ml). IL-6 markedly inhibited the hepatocyte DNA synthesis stimulated by TNF and EGF. The action of TNF was completely suppressed by transforming growth factor beta, which is known as a potent inhibitor of hepatocyte growth. Interferon gamma also blocked this TNF action when added simultaneously. These results indicate that the activation of tissue macrophages and local secretion of TNF in liver after partial hepatectomy is of physiological importance in liver regeneration, in part by a direct stimulation of hepatocyte DNA synthesis. Cytokines induced by TNF may also participate in the later termination of liver regeneration.     

Basic fibroblast growth factor and transforming growth factor-alpha are hepatotrophic mitogens in vitro

Basic fibroblast growth factor (bFGF) and transforming growth factor-alpha (TGF alpha) have been identified as potent hepatotrophic mitogens. bFGF and TGF alpha induce DNA synthesis in fetal and adult rat hepatocytes in primary culture and support fetal rat hepatocyte multiplication in chemically defined medium. No additional exogenous growth or progression factors are required by the cells for traversing the cell cycle or for cell division. These mitogenic polypeptides, previously identified in various cell types including liver and endothelial cells, platelets, and macrophages may act locally in a paracrine mode in controlling hepatocyte multiplication in the liver during development and regeneration.     

Density-dependent growth control of adult rat hepatocytes in primary culture

Adult rat hepatocytes in primary culture, which show various liver functions, did not show any mitosis at confluent cell density, although they entered the S phase and remained in the G2 phase, judging by cytofluorometry, when insulin and epidermal growth factor (EGF) were added to 2-day cultures (Tomita, Y., Nakamura, T., & Ichihara, A. (1981) Exp. Cell Res. 135, 363-371). However, when the cell density was decreased by half or one third, the number of nuclei and cell number increased to 1.5-2.0 times that after culture for 35 h with insulin and EGF. Moreover, at these lower densities, DNA synthesis started much earlier, although at the usual high density DNA synthesis with these two hormones did not start until the hepatocytes had been cultured for over 40 h. These results suggest that proliferation of mature rat hepatocytes is regulated by the cell density. First, cells in G0 enter the G1 phase density-dependently; then cells in the G1 phase seem to be stimulated to enter the S phase by insulin and EGF, and a low cell density may permit cells after DNA synthesis to enter the M phase. DNA synthesis of rat hepatocyte cultures at low cell density was strongly inhibited by co-culture with a dense culture. Therefore, the density-dependent mechanism of hepatocyte proliferation seems to involve regulation by a soluble inhibitor(s) secreted by the hepatocytes into the culture medium.     

Kinetics of DNA synthesis in feeding-dependent and independent hepatocyte populations of rats after partial hepatectomy

The effects of food consumption on the kinetics of hepatic DNA synthesis after partial hepatectomy (PH) have been studied in rats. Short-term (4-24 hr) fasting before or after PH resulted in depression and/or delay of DNA synthesis on days 1, 2 and 3 of regeneration. This depression was found in hepatocytes and, to a lesser extent, in littoral cells. Re-feeding resulted in an increase of DNA synthesis within 3-8 hr. The results suggest that two different hepatocyte subpopulations exist in regenerating rat liver: one which proceeds to DNA synthesis without apparent exogenous signals, and another one which needs, in addition to the specific mitogenic action of PH, food intake as a secondary permissive signal in order to initiate DNA synthesis. In the latter population food consumption appears to be required at two different stages: (1) in G0 or the early pre-replicative phase (PRP); (2) in the late PRP 3-8 hr before initiation of DNA synthesis. In the latter stage dietary protein is needed, but no so in the former. The dependence on feeding in the late PRP increases relatively with time after PH. No evidence was found to suggest a different distribution of the two cell populations throughout the liver acinus. The findings support the hypothesis that the known effects of the light-dark rhythm on the timing of DNA synthesis after PH are mediated by the natural feeding rhythm of rats fed ad libitum. In addition they offer a means for improving the synchrony of hepatocyte proliferation in regenerating rat liver.     

Inhibitory effect of transforming growth factor-beta on DNA synthesis of adult rat hepatocytes in primary culture

A transforming growth factor-beta (TGF-beta) found in platelets strongly inhibited DNA synthesis of adult rat hepatocytes in primary culture stimulated by insulin plus EGF or by hepatocyte growth factor (HGF) from rat platelets, but not the syntheses of secretory and intracellular proteins by the cells. TGF-beta had no cytotoxic effect, as judged by phase-contrast microscopic examination of the cell morphology. The inhibition of DNA synthesis by TGF-beta was correlated with marked decrease in the labeling index. TGF-beta did not inhibit growth of hepatoma cell line. These findings indicate that TGF-beta is a strong growth inhibitor of adult rat hepatocytes and may block their shift from the G1 phase to the S phase. The physiological role of TGF-beta in inhibiting growth of adult hepatocytes during liver regeneration is discussed.     

Spatiotemporal changes in Ha-ras p21 expression through the hepatocyte cell cycle during liver regeneration.

The protein product of the ras oncogene, Ha-ras (p21), is thought to be an important regulator of cell growth. The cytoplasmic relocalization of p21 in the cell during the cell cycle suggests a transient signaling role for this protein in association with its signal transduction function. Because of the importance of this role we examined spatial patterns in vivo of p21 expression at the protein and mRNA levels in hepatocytes during compensatory growth in rat liver following partial hepatectomy. A low level of p21 was immunolocalized on the cytoplasmic membrane of nonregenerating hepatocytes. The level of hepatic p21 increased significantly and without spatial restriction within the liver from 36 to 60 hr after partial hepatectomy (PH). p21 was localized in the cytoplasm of dividing hepatocytes and on the hepatic cytoplasmic membrane. The elevated p21 level decreased and was found mainly on hepatocyte plasma membranes by 96 hr after PH. Immunogold electron microscopy showed p21 localized over mitochondrial membranes and nuclei in nondividing regenerating hepatocytes. Approximately 50% of nonregenerating hepatocytes show nuclear localization of p21. This percentage changes with time following PH. The decrease in nuclear localization was accompanied with an increase in the low number of hepatocytes which demonstrated cytoplasmic localization in nondividing hepatocytes in regenerating liver. Flow cytometric analysis revealed a significant increase of p21 at 36 hr after PH which was 12 hr after the initial induction of ras mRNA. ras mRNA level increased 1.5-fold at 24 hr after PH and a maximum twofold induction was observed at 48 hr. Cell-cycle analysis of regenerating hepatocytes indicated a synchronized first peak of cell division 36-40 hr after PH. Dual parameter flow cytometry revealed that the level of p21 in hepatocytes in S phase and G2/M phase of the cell cycle was significantly higher than that in G0/G1 phase during regeneration. These findings suggest that p21 is important for the progression of regenerating hepatocytes to S phase and then to G2/M phase.     

Inhibitors of Protein Biosynthesis Can Stimulate Proliferation of Mouse Hepatocytes in Vitro

Hepatocyte proliferation in the liver regenerating after partial hepatectomy ceases when the organ is restored, and the mechanism of this phenomenon is still unclear. In the experiments on fusing hepatocytes from the regenerated mouse liver (15 days after partial hepatectomy) with NIH 3T3 mouse fibroblasts, we revealed no DNA synthesis in the nuclei of stimulated fibroblasts in heterokaryons (in the presence of hepatocyte nuclei), whereas DNA synthesis in nonfused cells was undisturbed. In this work, our purpose was to find out whether the suppression of DNA synthesis in heterokaryons could be due to the appearance in hepatocytes of some endogenous factors having an inhibitory effect on proliferation. To this end, hepatocytes from the mouse liver regenerated after partial hepatectomy were treated with cycloheximide for 1–4 h and were then fused with stimulated fibroblasts. Such a short-term treatment of hepatocytes with cycloheximide proved to result in the loss of their ability to inhibit DNA synthesis in the nuclei of stimulated or quiescent fibroblasts in heterokaryons, but hepatocytes proper actively proliferated in the medium with a low serum content (0.2%). When the mice with the liver regenerated after partial hepatectomy were treated with a single sublethal dose of cycloheximide (3 mg/kg), their hepatocytes taken two days after this treatment had no inhibitory effect. Puromycin, another inhibitor of protein synthesis, had the same effect on hepatocytes. These results may be interpreted as evidence that the final stage of liver regeneration after damage is controlled by the factors having a negative effect on cell proliferation.     

The dependence of the cytokinetic effects of alkylating antitumor preparations on the phase of the hepatocyte cell cycle in regenerating liver]

Effects of alkylating antitumor drugs on resting (G0 phase of cell cycle) and proliferating (G1, S, G2 and M phases) hepatocytes were studied in regenerating mouse liver. Cell cycle kinetics (fraction of labeled mitoses, labeling and mitotic indices) were determined by 3H-thymidine autoradiography. Dipin and fotrin as a DNA-damaging agents attack mainly resting (G0) and proliferating (G1) cells. Effect of the damage results in the inhibition of DNA synthesis and G2 phase arrest in the following mitotic cycle. An alkylating drug phopurin as well as ara-C both suppress the mitotic progression in proliferating hepatocytes and do not influence the resting cells.