Transplanted hepatocytes proliferate differently after CCl4 treatment and hepatocyte growth factor infusion

Sanjeev Gupta, Pankaj Rajvanshi, Emma Aragona, Chang-Don Lee, Purnachandra R. Yerneni, Robert D. Burk


To understand regulation of transplanted hepatocyte proliferation in the normal liver, we used genetically marked rat or mouse cells. Hosts were subjected to liver injury by carbon tetrachloride (CCl4), to liver regeneration by a two-thirds partial hepatectomy, and to hepatocellular DNA synthesis by infusion of hepatocyte growth factor for comparative analysis. Transplanted hepatocytes were documented to integrate in periportal areas of the liver. In response to CCl4 treatments after cell transplantation, the transplanted hepatocyte mass increased incrementally, with the kinetics and magnitude of DNA synthesis being similar to those of host hepatocytes. In contrast, when cells were transplanted 24 h after CCl4 administration, transplanted hepatocytes appeared to be injured and most cells were rapidly cleared. When hepatocyte growth factor was infused into the portal circulation either subsequent to or before cell transplantation and engraftment, transplanted cell mass did not increase, although DNA synthesis rates increased in cultured primary hepatocytes as well as in intact mouse and rat livers. These data suggested that procedures causing selective ablation of host hepatocytes will be most effective in inducing transplanted cell proliferation in the normal liver. The number of transplanted hepatocytes was not increased in the liver by hepatocyte growth factor administration. Repopulation of the liver with genetically marked hepatocytes can provide effective reporters for studying liver growth control in the intact animal.

  • carbon tetrachloride
  • hepatocyte transplantation
  • liver regeneration

the regenerative potential of the liver has long been recognized (24), although there has been considerable debate concerning the relative role of hepatocytes and putative progenitor (or stem) cells in this process under various circumstances (1, 4, 22,35-37). Recent studies (2, 3, 31) indicated that progenitor liver cells can be isolated, with the potential to differentiate either completely or partially along the hepatocyte lineage in suitable microenvironments. On the other hand, when host hepatocytes are chronically and extensively depleted, such as in the albumin-urokinase-type plasminogen activator transgenic (Alb-uPA) or fumarylacetoacetate hydrolase (FAH)-deficient mice, transplanted hepatocytes can repopulate the entire liver without involving progenitor cell activation (27, 30). Indeed, serial cell transplantation studies with FAH-deficient mice proved that hepatocytes isolated from adult mouse donors can replicate indefinitely during repopulation of the diseased liver (27). These findings suggest that, in appropriate pathobiological settings, such as fulminant hepatic failure due to extensive loss of parenchymal liver cells, transplanted hepatocytes should proliferate significantly if at a selective advantage to do so and, consequently, help improve survival (13). However, whether similar concepts concerning liver repopulation could be applied to proliferation of transplanted hepatocytes in the normal liver has not been well established.

To establish mechanisms concerning proliferation of transplanted hepatocytes in the normal liver, we performed a series of studies in rodent models. Our general hypothesis was that in the presence of suitable mitogenic signals, transplanted hepatocytes will proliferate in the normal liver. We further considered that, on integration into the hepatic parenchyma, transplanted cells would participate physiologically in ongoing processes in the liver similar to those affecting host hepatocytes. We utilized recently developed genetic reporter systems for our studies, including a transgenic hepatitis B virus (HBV) cell transplantation system in which transgenic cells expressing HBV surface antigen (HBsAg) are transplanted into congenic mice (10, 11, 14) and the dipeptidyl peptidase IV-deficient (DPPIV−) F-344 rat system in which F-344 rat hepatocytes are transplanted into syngeneic mutant animals (12, 13, 28, 29). The former system allows estimates of the transplanted hepatocyte mass by serum HBsAg measurements (11), whereas the latter system offered convenient strategies for in situ localization and dual-label studies to analyze proliferation in transplanted cells (12, 28). Previous work by Rajvanshi et al. (28) established that transplanted hepatocytes are distributed in periportal areas (zone 1) of the rodent liver, allowing consideration of strategies to spare transplanted cells from ablative injury.

Our major goal here was to determine how transplanted hepatocytes responded to proliferative stimuli in the normal liver. The experimental perturbations were to ablate liver cells in a targeted fashion by carbon tetrachloride (CCl4), which depletes perivenous hepatocytes (zone 3 of the liver lobule) (23), to perform a partial hepatectomy, which induces highly synchronous DNA synthesis in the remnant liver (16, 24, 38), and to administer hepatocyte growth factor (HGF), a potent hepatic mitogen (24, 25, 32) that is released after both partial hepatectomy and toxic liver injury such as by CCl4 (20, 21).


Animals. Male F-344 rats serving as hepatocyte donors were obtained from the National Institutes of Health (Bethesda, MD) and weighed 250–300 g (12–16 wk old) when used. C57BL/6J mouse recipients weighing 15–20 g (6–8 wk old) were from Jackson Laboratories (Bar Harbor, ME). Syngeneic, DPPIV− F-344 recipient rats (160–180 g) and congenic G26 HBV transgenic donor mice (25–30 g) were provided by the Special Animals Core of the Marion Bessin Liver Research Center (Bronx, NY). The studies used animals in multiple groups composed of at least three to four animals each under protocols approved by the institutional Animal Care and Use Committee in accordance with theGuide for the Care and Use of Laboratory Animals [DHEW Publication No. (NIH) 85-23, revised 1985, Office of Science and Health Reports, DRR/NIH, Bethesda, MD 20505].

Cell transplantation by intrasplenic injection in rodents was performed as previously described (14, 28). The animals were anesthetized with inhaled ether (Fisher Scientific, Fairlawn, NJ). Rats received 2 × 107 cells and mice received 2 × 106 cells suspended in RPMI 1640 medium. A two-thirds partial hepatectomy in rats was performed according to Higgins and Anderson (16). A partial hepatectomy in mice, which required leaving a collar of tissue adjacent to the gallbladder bed and bile duct, was performed as previously described (38). After the abdomen was closed, the animals were monitored closely until recovery. To induce hepatic injury, 1.45 ml/kg of CCl4 was administered intramuscularly once after dilution to either 1:1 or 1:2.5 (vol/vol) in mineral oil for rats and mice, respectively (Sigma, St. Louis, MO). To analyze changes in individual animals, serial wedge liver biopsies were performed via a midline laparotomy in recipients of CCl4 as previously described (29).

For growth factor-induced hepatic DNA synthesis, we used recombinant human (h) HGF (lot 18181-25, Genentech, South San Francisco, CA). The hHGF was >90% two chain as shown by densitometry of reduced, purified protein by SDS-PAGE (25). To administer HGF, Alzet osmotic pumps (Alza, Palo Alto, CA) were primed overnight, and injection ports were inserted into the spleen of animals with P-60 tubing (VWR Scientific, West Chester, PA). The tubing was secured by an encircling ligature around the splenic lower pole while the proximal end was tunneled subcutaneously to connect with the pump implanted on the back of the animals. For mice, Alzet 2001 pumps were used, with infusion rates of 1 μl/h, and for rats, Alzet 2ML1 pumps were used, with infusion rates of 10 μl/h. The hHGF was diluted in 0.5 M NaCl and mixed with an equal amount of dextran sulfate as previously described (32). The dose of hHGF administered was 2.4 mg ⋅ kg−1 ⋅ day−1for 7 days.

Hepatocyte isolation and culture.Hepatocytes were isolated from donor animals with a two-step perfusion with 0.025% (wt/vol) collagenase D (Boehringer Mannheim, Indianapolis, IN) as previously described (14, 28). The cell viability was tested by trypan blue exclusion as well as by attachment to tissue culture plastic in RPMI 1640 medium containing 100 U/ml of penicillin, 100 μg/ml of streptomycin, and 10% heat-inactivated fetal bovine serum (GIBCO, Grand Island, NY). To document mitogenic activity of hHGF, hepatocytes were cultured on rat tail collagen-coated dishes at 2.5 × 105cells/cm2 with and without 20 ng/ml of hHGF for 48 h. DNA synthesis was measured by incubating hepatocytes with [3H]thymidine (specific activity 50–70 mCi/mmol; ICN, Irvine, CA) for 1 h. The cells were then lysed in 0.3 M sodium hydroxide, TCA-precipitable DNA was extracted, and [3H]thymidine incorporation was determined along with microfluorometric quantitation of DNA in aliquots as previously described (9).

DNA synthesis measurements in tissues.To determine the magnitude and kinetics of liver regeneration with the procedures used, groups of three mice and three rats each were given 0.5 μCi [3H]thymidine/g body wt intraperitoneally 1 h before the animals were killed at various times after the interventions. The tissues were subjected to autoradiography with NTB-2 emulsion (Eastman Kodak, Rochester, NY) for 4 wk as previously described (11). To determine labeling indexes, two independent observers counted 10,000–15,000 hepatocytes/section in a blinded manner.

In some animals, H3 histone mRNA expression, which correlates with DNA synthesis, was analyzed with previously described specific probes (38). A 280-base pair fragment was used from the coding sequences of human albumin cDNA (plasmid pHSA1), which binds mouse albumin mRNAs (38). To generate antisense and sense RNA probes labeled with digoxigenin-11-UTP, commercial kits were used according to the manufacturer’s instructions (Boehringer Mannheim). In situ hybridization was on 5-μm-thick paraformaldehyde-fixed cryostat sections under stringent conditions (11, 14, 38). After hybridizations, probe binding was localized with an anti-digoxigenin-alkaline phosphatase conjugate according to the manufacturer’s instructions (Boehringer Mannheim).

For localization of DNA synthesis in transplanted cells, animals were given a 2-h pulse of 50 mg/kg of bromodeoxyuridine (BrdU; Boehringer Mannheim). To detect BrdU incorporation, cryostat sections were subjected to immunostaining with a commercially available antibody system (Amersham, North Chicago, IL). The tissues were blocked with 2% rabbit serum, and antibody binding was detected by a supersensitive multilink antibody system with the peroxidase reporter (BioGenex Laboratories, San Ramon, CA), followed by color development with a Vectastain kit (Vector Laboratories, Burlingame, CA).

Localization of transplanted cells. To localize transgenic G26 HBV hepatocytes, in situ hybridization was performed on cryostat tissue sections with35S-dUTP-labeled HBsAg probes as previously described (11, 14). The sections were autoradiographed at 4°C with NTB-2 emulsion (Eastman Kodak) for several days before being developed.

To localize F-344 rat hepatocytes in DPPIV− animals, histochemical staining was performed as previously reported (12, 28,29). Integrations of transplanted cells in the liver parenchyma were demonstrated by colocalization of DPPIV activity and bile canalicular ATPase activity. To colocalize DNA synthesis in transplanted cells, tissues were first subjected to DPPIV staining, and this was followed by immunostaining for BrdU incorporation. Tissues were counterstained with methyl green or hematoxylin as appropriate.

Morphometric analysis. To determine changes in cell number in the tissues, wedge biopsies taken at 4 wk after cell transplantation from three individual rats before initiation of CCl4 treatment and final tissue samples obtained at 36 wk after cell transplantation were analyzed with histochemical staining for DPPIV activity. The number of cells per portal area was determined in individual tissue sections for quantitative morphometry. A minimum of 50 portal areas were analyzed in each tissue. For additional analysis, changes in cluster sizes were determined by analyzing a minimum of 250 consecutive areas with transplanted hepatocytes in each tissue. For this purpose, transplanted cells in consecutive areas were scored for the number of cells per cluster: 1–3, 4–6, 7–9, 10–12, 13–15, and >16 cells each. Similar analysis was conducted to determine whether the cell number changed in animals subjected to other mitogenic treatments.

Serological assays. Blood was sampled at 3- to 7-day intervals from the tails of animals, and the sera were stored at −20°C for analysis. Serum HBsAg was measured with a commercially available radioimmunoassay (AUSRIA II, Abbott Laboratories, North Chicago, IL). HBsAg standards were included in a known range of concentrations as previously described (11, 14). The baseline serum HBsAg levels were used to normalize subsequent serum HBsAg levels in individual animals for comparisons.

Statistics. Data are expressed as means ± SD. SigmaStat 2.0 software was used for data analysis (Jandel, San Rafael, CA). The significance of variances was analyzed as appropriate by Student’s t-test, χ2 test, Mann-Whitney rank sum test, and ANOVA.


Characteristics of isolated hepatocytes. The viability of isolated rat and mouse hepatocytes ranged from 85 to 90% as assessed by trypan blue dye exclusion. The cells promptly attached to tissue culture plastic (>60% attachment after 1 h) and assumed normal parenchymal morphology in culture. In addition, both mouse and rat hepatocytes responded to hHGF, with DNA synthesis rates increasing by four- to sixfold in comparison with unstimulated control cells at 48 h (P < 0.001). These data verified the integrity of isolated cell preparations and indicated that our hHGF was bioactive.

Magnitude of liver regeneration induced. To document that various maneuvers induced liver regeneration, we studied C57Bl/6J mouse and DPPIV− F-344 rat hosts before proceeding with subsequent experiments.

The kinetics of DNA synthesis after a two-thirds partial hepatectomy in rats was similar to a previous experience, with an early peak at 24 h, followed by rapid attenuation of the regenerative response (Fig.1 A). In contrast, after CCl4administration, the initial peak was delayed and DNA synthesis was characterized by a relatively lower magnitude as well as a longer persistence compared with partial hepatectomy in both mice and rats. Also, after partial hepatectomy, DNA synthesis was apparent in hepatocytes essentially throughout the liver lobule, whereas CCl4 treatment resulted in DNA synthesis in the periportal areas because perivenous hepatocytes were destroyed by the toxin (Fig. 1 B). The distribution of hepatocytes containing either [3H]thymidine or BrdU was similar to the distribution of hepatocytes expressing histone H3 mRNA (Fig. 1 C) as also previously observed (38). In the dose used, CCl4 depleted hepatocytes from only ∼30 to 50% of the liver lobule (Fig. 1,DF), which most likely accounted for the difference in the magnitude of DNA synthesis compared with that resulting from a two-thirds partial hepatectomy. The hepatic susceptibility to CCl4 is determined by the activity of specific cytochrome P-450 isoforms in individual cells, with perivenous hepatocytes being most sensitive (21).

Fig. 1.

Induction of DNA synthesis by mitogenic stimuli.A: comparison of DNA synthesis induced by two-thirds partial hepatectomy or carbon tetrachloride (CCl4) treatment. DNA synthesis rates were measured by analyzing incorporation of [3H]thymidine into hepatocyte nuclei after autoradiography of liver from 3 rats/group. Kinetics and magnitude of hepatic DNA synthesis after CCl4 treatment were different from those after partial hepatectomy. B: immunostaining for bromodeoxyuridine (BrdU) incorporation in a rat liver 48 h after CCl4 treatment. BrdU was incorporated by cells situated mostly in periportal areas (p) compared with perivenous areas (c). In contrast, after partial hepatectomy, BrdU was incorporated by hepatocytes throughout liver lobule (data not shown). C: in situ hybridization showing H3 histone mRNA expression in liver from a C57BL/6J mouse 48 h after CCl4treatment. Note that pattern of DNA synthesis is similar to that in rat liver in B.D: in situ hybridization for albumin mRNA expression showing ablation of significant portions of liver lobule in a mouse 24 h after CCl4treatment. Albumin mRNA was expressed throughout liver in normal mice (E), and no hybridization signal was apparent in negative control animals (F).G: BrdU incorporation in mouse liver (arrow) after initiation of human hepatocyte growth factor (hHGF) infusion 60 h previously. H: in an untreated control mouse, only an occasional hepatocyte showed BrdU incorporation.

After hHGF infusion in the animals, we analyzed DNA synthesis rates at 60 h after commencing the infusion, which has previously been shown to be effective in intact animals (32). Studies were performed in three mice per group and showed significantly increased BrdU incorporation compared with untreated control animals (Fig. 1,G andH). The overall hepatocyte labeling index approached 42 ± 6/1,000 cells, which was less than that observed with either a two-thirds partial hepatectomy or CCl4 treatment, although this compared favorably with DNA synthesis after a 35% partial hepatectomy in a previous study (38).

Effect of CCl4-induced liver regeneration on transplanted cell proliferation.

The strategy was based on the fact that intrasplenic transplantation deposits cells in periportal locations (28). The hypothesis was that the susceptibility of perivenous hepatocytes to CCl4 would not extend to transplanted cells located in periportal areas and that transplanted cells would participate in the ensuing proliferative response of host hepatocytes in nonablated areas.

The experimental strategy was to let transplanted cells become fully integrated into the liver parenchyma over a period of several weeks (Fig. 2). Although cell engraftment may require less time, we wished to ensure that transplanted cells, which were not fractionated to exclude cytochromeP-450-expressing cells, were resistant to CCl4 injury, and the time required for this was unknown. The animals were then given three doses of CCl4 at 4-wk intervals to permit complete recovery from the preceding injury. Subsequently, the animals were observed for a total of 36 wk from the time of cell transplantation.

Fig. 2.

Strategy for analyzing effect of CCl4 treatments on transplanted cell proliferation. Experimental design was to transplant cells into young adult recipients for an 8-wk baseline period. CCl4 was given at 8, 12, and 16 wk in test animals. Control animals did not receive CCl4. Animals were studied for up to a total of 36 wk. Mice underwent blood sampling throughout at 3- to 5-day intervals. For tissue analysis, 2 mice in the test group were killed before CCl4 treatments were begun and tissues from 2 mice in CCl4-treated group were obtained at completion of experiments. At least 6 mice in each group survived for 36-wk duration of experiment. In several rats, wedge biopsies were performed before each dose of CCl4for documenting histological changes serially.

The first experiment utilized the transgenic G26 HBV system because serum HBsAg levels provide a simple and convenient measure of the transplanted hepatocyte mass (11). The hepatocytes were transplanted into 15 congenic C57BL/6J mice, of which 9 mice underwent CCl4 treatment and 6 mice served as untreated controls. The data showed that serum HBsAg levels increased incrementally after CCl4 treatment (Fig.3 A). In situ hybridization showed that HBsAg mRNA containing transplanted cells were present in the liver of the recipients, with an increase in the overall number of cells at the end of the 36-wk period (Fig. 3,B andC). When data were further analyzed with respect to the various time periods defined above, it became clear that serum HBsAg levels increased in each period after CCl4 treatment and were 2.7-fold greater in the final period after all three CCl4 treatments (Table1). Interestingly, serum HBsAg levels showed a gradual increase from baseline to the final period in control normal mice, which was in agreement with some spontaneous increase in the transplanted cell number without CCl4 treatment, although tissues were not analyzed to verify this.

Fig. 3.

Hepatocyte proliferation after CCl4 treatment in mice transplanted with transgenic G26 hepatitis B virus (HBV) hepatocytes.A: cumulative changes in serum HBV surface antigen (HBsAg) levels in CCl4-treated (arrows) vs. control mice (n = 6/group). Note that serum HBsAg levels increased markedly in recipients of transplanted cells after completion of CCl4treatments. B: in situ hybridization showing transplanted cells containing HBsAg mRNA in a mouse before CCl4 treatments. Transplanted cells covered with black grains are situated adjacent to a portal area.C: after completion of all 3 CCl4 treatments, there was a significant increase in the number of transplanted cells. In addition, transplanted cells were situated further away from portal areas.

View this table:
Table 1.

Changes in serum HBsAg levels in hepatocyte recipients after CCl4 treatments

In further studies, we performed experiments in DPPIV− F-344 rats for additional analysis of CCl4-induced increases in cell proliferation. The studies verified that transplanted cells integrated into the liver parenchyma during the baseline period of 8 wk and formed hybrid bile canaliculi by joining adjacent host hepatocytes (Fig.4, A andB). Moreover, when tissues were analyzed serially with multiple individual animals (3 rats/time point) undergoing wedge liver biopsies before each CCl4 treatment, it became clear that the number of transplanted cells in the liver increased significantly after CCl4 treatment (Fig. 4,CE). To determine whether transplanted cells underwent DNA synthesis in this situation, three additional rats were treated with CCl4 after 8 wk of cell transplantation and were given BrdU before being killed 48 h later. BrdU incorporation was apparent in both transplanted and host hepatocytes at this time (Fig. 4 F). The proportion of cells undergoing DNA synthesis was the same in both host and transplanted hepatocytes (15 ± 4 vs. 13 ± 8%;P = not significant), indicating that the proliferative activity of transplanted cells was similar to that of the host cells.

Fig. 4.

Effect of CCl4 treatments on transplanted cells in rats. In dipeptidyl peptidase IV-deficient (DPPIV−) rats, transplanted cells were located in periportal areas similar to those in mice. A: transplanted cells in an animal from control group at 8 wk.B: same control animal again at 36 wk after cell transplantation showing no significant change in transplanted cell number. C: transplanted cells in an animal at 8 wk immediately before 1st CCl4 treatment.D: same animal as inC after completion of all 3 CCl4 treatments. Number of transplanted cells was increased after CCl4 treatment.E: colocalization of bile canalicular DPPIV activity (short arrow, transplanted cells) and ATPase activity (long arrows in A and E, host cells) showed that transplanted cells were integrated in host liver parenchyma and formed networks of hybrid bile canaliculi. F: DNA synthesis in transplanted hepatocytes with BrdU incorporation (arrows) 48 h after CCl4treatment. Tissues inAEwere reacted first for DPPIV activity and then for biliary ATPase activity.

Cumulative analysis of tissues showed 6 ± 3 transplanted cells/portal area before CCl4treatment and 15 ± 7 cells/portal area at 36 wk, subsequent to all three CCl4treatments, which represented a 2.5-fold mean increase (P < 0.001 byt-test). The analysis of cell cluster sizes was further indicative of changes in the overall cell number (Table 2). Before CCl4 treatments, the overall number of cells per cluster was limited, whereas, subsequent to CCl4 treatments, the number of cells per cluster increased markedly and large clusters containing >50–100 transplanted cells were also observed, albeit less frequently (∼1%). In addition, analysis of the position of transplanted cells in the liver lobule showed that, after CCl4 treatments, cells progressively migrated toward perivenous areas (zone 3) such that after all three CCl4 treatments, a number of transplanted cell clusters were clearly adjacent to the perivenous areas of the hepatic lobule. However, in normal control rats that did not receive CCl4, such a change in the position of transplanted cells was not observed. Also, there was no significant change in the number of transplanted cells between early or late times in control rats that did not receive CCl4. In this situation, the number of transplanted cells and cluster sizes were similar throughout to those observed at 4 wk before CCl4 administration in treated rats.

View this table:
Table 2.

Changes in cell cluster size distributions at various times after CCl4 treatments in DPPIV− rats

Induction of cell proliferation immediately after hepatocyte transplantation. If mechanisms could be identified for rapid amplification of transplanted hepatocyte mass, clinical applications should be facilitated. To determine this, studies were performed by transplanting cells in DPPIV− rats subsequent to CCl4 or hHGF administration. In view of preliminary data derived from pilot studies, these experiments were restricted to 2-wk durations.

Hepatocytes were transplanted into 6 DPPIV− rats (2 × 107 cells; control group), and, in 12 additional rats, CCl4 was administered 24 h before transplantation of 2 × 107 cells. Two control animals and three animals that were subjected to CCl4 treatment were each killed 1, 2, and 14 days after cell transplantation. At 24 or 48 h after transplantation in normal rats, transplanted cells appeared healthy, whereas in animals treated with CCl4 transplanted cells showed significant fatty change and vacuolization (Fig.5 A). At 14 days after transplantation, the number of transplanted cells was less in animals treated with CCl4(1 ± 3 cells/portal area) compared with normal control animals (4 ± 5 cells/portal area; P < 0.05 by ANOVA; Fig. 5, B andC). These findings suggested that residual CCl4 was harmful to transplanted cells. CCl4-induced injury to transplanted hepatocytes in portal and periportal areas was different from the absence of such an injury in fully engrafted hepatocytes. This may have been due to variable CCl4 utilization by cells because hepatocytes were not fractionated before transplantation to exclude cytochrome P-450-expressing perivenous cells.

Fig. 5.

Induction of cell proliferation immediately after hepatocyte transplantation in CCl4recipients. All tissues were subjected to both DPPIV and ATPase stainings. Dark brown staining represents bile canalicular ATPase activity in host liver. Transplanted cells show red staining with DPPIV activity. A: hepatocytes in portal spaces 24 h after transplantation into an animal that had received CCl4 48 h previously. Note vacuoles in transplanted cells (arrow), which were not observed in normal animals (data not shown). B andC: transplanted hepatocytes in DPPIV− rats 14 days after cell transplantation into a control rat and a rat with prior CCl4treatment, respectively. Note that there are fewer transplanted cells after CCl4 treatment compared with transplanted cells in control rat liver.

Because a partial hepatectomy and CCl4 cause a coordinated release of a variety of growth factors including HGF (20, 21), we also investigated whether growth factor infusion at the time of cell transplantation could increase the transplanted hepatocyte mass. For this purpose, we first implanted primed osmotic pumps into three DPPIV− rats to infuse hHGF and into three additional rats to infuse saline as a control. After 24 h, F-344 rat hepatocytes were transplanted via the spleen followed by determination of transplanted cell numbers 14 days later. The animals were given 50 mg/kg body wt of BrdU intraperitoneally daily to determine whether hHGF infusion induced hepatic DNA synthesis. Although hepatocytes showed BrdU incorporation throughout the liver lobule in recipients of hHGF, morphometric analysis showed that the number of transplanted cells was not greater than that in saline-treated control animals (P = not significant).

To determine whether transplanted cells could be induced to proliferate by exogenous administration of growth factors, we infused hHGF into three mice, with a group of four control animals. The infusion was begun 24 h before transplantation of G26 HBV hepatocytes and continued for 7 days. However, there was no significant increase in the transplanted hepatocyte mass (Fig. 6).

Fig. 6.

Transplanted cell proliferation in response to direct growth factor stimulation. Serum HBsAg levels in mice (n = 3) subjected to hHGF infusion 24 h before transplantation of transgenic G26 HBV hepatocytes were similar to those in control animals (n = 4). There was a significant increase in BrdU incorporation in hHGF recipients at 60 h after infusion.


Our findings provide novel insights into mechanisms regulating proliferation of transplanted hepatocytes in the healthy liver. The ability of transplanted hepatocytes to integrate in the liver parenchyma and assume normal polarity in the liver plate suggested to us that the cells should respond appropriately to physiological stimuli (12). We were impressed by the capacity of transplanted hepatocytes to undergo repeated proliferation in the setting of recurrent liver injury, although cell proliferation appeared to be highly regulated each time and occurred in a manner sufficient to replace the lost hepatocyte mass.

The results of our CCl4 treatment studies were particularly instructive and revealed several important features. Previous studies (27, 30) documented that transplanted hepatocytes can proliferate extensively in the setting of chronic depletion of host hepatocytes, such as in the Alb-uPA transgenic and FAH-deficient mice. Serial transplantation studies in FAH-deficient mice showed that transplanted hepatocytes can undergo multiple cell divisions while repopulating the chronically diseased liver (27). However, because transplanted cells did not previously undergo significant proliferation in the normal liver (10), it was unclear as to what mechanisms could be useful in amplifying the transplanted hepatocyte mass in the healthy liver. This consideration is especially relevant in cell therapy strategies concerning metabolic deficiency states such as familial hypercholesterolemia and Criggler-Najjar syndrome, which cause little or no liver injury and will benefit from amplification of the transplanted hepatocyte mass for superior results (1a, 7, 15, 17, 26).

Proliferation in transplanted cells commensurate with healing needed to restore liver mass after CCl4suggests that targeted hepatic ablation should be effective in applications of hepatocyte transplantation. Of course, CCl4 has extensive general toxicities as well as a relatively long half-life, precluding application, as also indicated in our studies, by injury to hepatocytes transplanted 24 h after CCl4administration. However, additional strategies might well become available in the future to ablate portions of the hepatic lobule safely. One such approach concerned transient expression ofAlb-uPA gene with an adenoviral vector in host hepatocytes before cell transplantation (39). These studies were reported subsequent to the initiation and completion of our experiments.

The migration of transplanted hepatocytes toward perivenous areas of the liver after CCl4 treatment was in agreement with shifts in host liver plates during recovery and offer further evidence for the potential of hepatocyte transplantation systems for addressing fundamental questions in liver physiology. For instance, such systems could be helpful in addressing mechanisms underlying position-specific regulation of gene expression and other events in the liver lobule that are incompletely resolved (8). We interpret migration of transplanted hepatocytes toward perivenous areas to represent a “wound repair” mechanism because such an event was not observed in the uninjured normal liver. The ability of transplanted hepatocytes to repopulate the entire liver in Alb-uPA and FAH-deficient mice, with transplanted hepatocytes shown to be present in perivenous areas (27), also illustrates replacement of injured hepatocytes and not necessarily “streaming” of hepatocytes from periportal to perivenous areas in the liver lobule, in agreement with similar interpretations by others using different systems (19).

Repeated proliferation in hepatocytes isolated from the unperturbed adult rat liver in our studies extend recent findings of extensive stem cell-type proliferative capacity in serially transplanted mouse hepatocytes (27). However, in contrast with our findings in CCl4-induced liver injury, we were intrigued to learn that transplanted hepatocytes did not proliferate in the liver after growth factor infusion, even though this induced significant hepatic DNA synthesis. Our data here are in agreement with a different proliferative response of transplanted hepatocytes after a partial hepatectomy compared with that after CCl4 treatment. Although hHGF infusion was ineffective in increasing the number of transplanted cells, we do not know whether more sustained hHGF treatment might have been effective in this regard. Similarly, whether simultaneous stimulation with additional growth factors such as transforming growth factor-α (40) or cell priming with other manipulations will improve results requires further analysis.

In conclusion, our data demonstrate that transplanted hepatocytes can be induced to proliferate significantly in the normal liver after selective injury to host hepatocytes, such as with CCl4. The role of hepatic ablative stimuli is particularly noteworthy in this respect compared with induction of DNA synthesis with growth factor stimulation. Finally, liver repopulation with genetically marked cells will permit the use of transplanted hepatocytes as powerful reporters for analyzing physiological mechanisms in liver regeneration.


We thank Dr. Ralph Schwall (Genentech, South San Francisco, CA) for the kind gift of human heptocyte growth factor.


  • Address for reprint requests and other correspondence: S. Gupta, Ullmann 625, Albert Einstein College of Medicine, 1300 Morris Park Ave., Bronx, NY 10461 (E-mail: sanjvgupta{at}

  • This work was supported in part by National Institute of Diabetes and Digestive and Kidney Diseases Grants K08-DK-01909 and R01-DK-46952 (to S. Gupta) and P33-DK-41296 (to D. A. Shafritz, Marion Bessin Liver Research Center, Albert Einstein College of Medicine, Bronx, NY) and the Irma T. Hirschl Trust.

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.


  1. 1.
  2. 1a.
  3. 2.
  4. 3.
  5. 4.
  6. 5.
  7. 6.
  8. 7.
  9. 8.
  10. 9.
  11. 10.
  12. 11.
  13. 12.
  14. 13.
  15. 14.
  16. 15.
  17. 16.
  18. 17.
  19. 18.
  20. 19.
  21. 20.
  22. 21.
  23. 22.
  24. 23.
  25. 24.
  26. 25.
  27. 26.
  28. 27.
  29. 28.
  30. 29.
  31. 30.
  32. 31.
  33. 32.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
View Abstract