Loss and Recovery of Liver Regeneration in Rats with Fulminant Hepatic Failure

JOURNAL OF SURGICAL RESEARCH ARTICLE NO.

72, 112–122 (1997)

JR975175

Loss and Recovery of Liver Regeneration in Rats with Fulminant Hepatic Failure Susumu Eguchi, M.D., Helene Lilja, M.D., Winston R. Hewitt, M.D., Yvette Middleton, M.S., Achilles A. Demetriou, M.D., Ph.D., and Jacek Rozga, M.D., Ph.D.1 Liver Support Research Laboratory, Department of Surgery, Burns and Allen Research Institute, Cedars-Sinai Medical Center, UCLA School of Medicine, Los Angeles, California 90048 Submitted for publication April 14, 1997

We earlier described a model of fulminant hepatic failure (FHF) in the rat where partial hepatectomy is combined with induction of right liver lobes necrosis. After this procedure, lack of regenerative response in the residual viable liver tissue (omental lobes) was associated with elevated plasma hepatocyte growth factor (HGF) and transforming growth factor b (TGF-b1) levels and delayed expression of HGF and c-met mRNA in the remnant liver. Here, we investigated whether syngeneic isolated hepatocytes transplanted in the spleen will prolong survival and facilitate liver regeneration in FHF rats. Inbred male Lewis rats were used. Group I rats (n Å 46) received intrasplenic injection of 2 1 107 hepatocytes and 2 days later FHF was induced. Group II FHF rats (n Å 46) received intrasplenic injection of saline. Rats undergoing partial hepatectomy of 68% (PH; n Å 30) and a sham operation (SO; n Å 30) served as controls. In 20 FHF rats (10 rats/group), survival time was determined. The remaining 72 FHF rats (36 rats/group) were used for physiologic studies (liver function and regeneration and plasma growth factor levels). In Group I rats survival was longer than that of Group II controls (73 { 22 hr vs. 33 { 9 hr; P õ 0.01). During the first 36 hr, Group I rats had lower blood ammonia, lactate, total bilirubin, PT, and PTT values, lower activity of liver enzymes, and higher monoethylglycinexylidide (MEGX) production than Group II rats. In Group I rats, livers increased in weight at a rate similar to that seen in PH controls and showed distinct mitotic and DNA synthetic activity (incorporation of bromodeoxyuridine and proliferation cell nuclear antigen expression). Plasma HGF and TGF-b1 levels in these rats decreased and followed the pattern seen in PH rats; additionally, c-met expression in the remnant liver was accelerated. Hepatocyte transplantation prolonged survival in FHF rats and facilitated liver re1

To whom correspondence should be addressed at Liver Support Research Laboratory, Department of Surgery, Cedars-Sinai Medical Center, 8700 Beverly Blv., D-4018, Los Angeles, CA 90048. Fax: (310) 652-7168.

0022-4804/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

JSR 5175

/

6n26$$$$41

generation. Even though the remnant liver increased in weight four times reaching 30% of the original liver mass, the transplant-bearing rats expired due to inability of the regenerating liver to support the rat. q 1997 Academic Press

INTRODUCTION

Management of severe acute liver failure continues to be one of the most challenging problems in clinical medicine [1]. Orthotopic liver transplantation has emerged as an effective treatment for acute liver failure and end-stage liver disease [2, 3]. However, wider application of this therapeutic modality is limited primarily by lack of donors, inability to procure organs on short notice, and high cost. Additionally, existing criteria of poor prognosis are inadequate to predict who will recover from acute liver failure without liver transplantation [4]. In searching for solutions to these problems, investigators have focused on transplantation of isolated hepatocytes. Potential advantages of this strategy include: (i) cells from a single donor can be used to treat multiple recipients, (ii) cell cryopreservation allows repeated treatments of a single recipient, (iii) in vitro cell manipulation could potentially decrease cell immunogenicity and/or correct genetic defects, and (iv) hepatocellular transplantation does not deprive recipients of their native liver. The latter may be important in patients with inherited defects of liver metabolism, in whom only a small number of normal hepatocytes is needed to replace a single missing function, and in patients with acute liver failure, in whom the native liver has potential for regeneration and full recovery. In experimental animals, the ability of transplanted hepatocytes to ameliorate, at least partially, a specific enzymatic defect [5–8] and provide temporary support to animals with toxic and surgically induced hepatic failure [9–11] has been demonstrated by many investigators, including ourselves [12–14]. Hence, it is now generally accepted that transplanted hepatocytes can

112

10-08-97 13:13:10

srgas

EGUCHI ET AL.: LIVER REGENERATION

assume most of the intact whole liver functions. However, while in the case of a genetic liver defect there is no doubt that the appearance of a missing gene product can be attributed solely to transplanted cells, the mechanism by which transplanted hepatocytes provide liver-specific support in recipients treated for physical liver injury is obscure. The observed beneficial effects can be attributed to transplanted cell function, to improved function of the native liver, or both. This dilemma was noted, but not resolved, nearly two decades ago when various investigators showed that D-galactosamine-induced acute liver failure is reversed by liver cell supernatants and cytosol fractions of homogenized liver [15, 16]. We have developed and characterized a novel experimental model of FHF in rats that has a number of physiological and biochemical features seen clinically in FHF [17]. In this preparation, resection of the two anterior liver lobes (median and left lateral; 68% liver mass) is combined with ligation of the common right liver lobes pedicle (24% liver mass). As a result, the functional liver mass is greatly reduced, there is a significant amount of liver tissue undergoing necrosis leading to toxemia, and animal survival is dependent on the ability of the residual omental liver lobes (8% liver mass) to function and regenerate. After this procedure, progressive liver failure was associated with markedly elevated plasma hepatocyte growth factor (HGF) levels, a finding commonly seen in FHF patients. At the same time, the residual liver showed no signs of cell proliferation, which seemed to be caused by a delayed expression of HGF and c-met messenger RNA and increased plasma transforming growth factor (TGF-b1) levels [17]. In the present study, we demonstrate that in FHF rats, ectopically (spleen) implanted syngeneic hepatocytes had significant beneficial effects on animal survival, blood chemistry, and regeneration of the native remnant liver. We also show that restoration of liver regeneration was associated with lowering of elevated plasma HGF and TGF-b1 levels and accelerated expression of HGF and c-met mRNA in the native liver. This study suggests that in the medical management of FHF, agents neutralizing hepatocyte growth inhibitors and enhancing/accelerating hepatocyte responsiveness to mitogens may be of value in facilitating repair of the injured liver. METHODS Animal studies were performed in compliance with institutional and National Research Council guidelines for humane care of experimental animals. Animals. Adult male Lewis (syngeneic) rats weighing 150 to 300 g were obtained from Harlan Sprague-Dawley, Inc. (Indianapolis, IN). Rats were acclimatized to our laboratory conditions for 1 week prior to use in the experiments. They were housed in a climatecontrolled (217C) room under a 12-hr light/dark cycle. Rats were given tap water and commercial rat chow (Rodent Chow 5001, Ralston Purina, St. Louis, MO) ad libitum. All operations were per-

AID

JSR 5175

/

6n26$$$$42

10-08-97 13:13:10

113

FIG. 1. Schematic diagram depicting the technique used to transplant hepatocytes into the rat spleen. Two anterior lobes (68%; median lobe and left anterior lobe) were resected and right lobes (24%) were rendered necrotic. Only the omental lobes (8%) were left intact. During cell injection (2 1 107 syngeneic hepatocytes), the splenic vessels were occluded to avoid immediate migration of cells to the liver.

formed between 9 AM and noon, under general (metaphane) anesthesia using sterile surgical technique. After surgery, rats were warmed externally (heating lamp) and, to avoid dehydration, were given a single bolus of 15 ml of 5% dextrose in normal saline subcutaneously. Afterward, no glucose was added to the drinking water. We previously reported that with this regimen, FHF rats developed moderate hypothermia and maintained blood glucose at 50 mg/dl or higher until death [17]. Chemicals. Unless otherwise noted, all chemicals were obtained from Sigma Chemical Co. (St. Louis, MO). Hepatocyte isolation. Donor hepatocytes were harvested from the livers of inbred Lewis rats (Ç150 g) by in situ two-step ethylenediaminetetraacetic acid [EDTA]/collagenase digestion, as previously described [18]. After enrichment through a Percoll (Pharmacia Biotech., Piscataway, NJ) gradient, hepatocyte viability was always greater than 95%, as judged by trypan blue exclusion.

Surgical Animal Experimental Models Induction of FHF. FHF was induced as earlier described [17]. Briefly, the abdomen was entered through a midline incision. The common pedicle to the right liver lobes (24% of the liver) was ligated, and the two anterior liver lobes (68% of the liver) were removed using the standard Higgins and Anderson technique [19]. The two omental liver lobes (8% of the liver) were left intact (Fig. 1). Special care was taken to fully mobilize the anterior liver lobes and to place a ligature around their common pedicle high, so that there would be no interference with the arterial blood supply to the liver remnant or impairment of venous outflow. The suggested contributions of the affected liver lobes to the total liver mass are based on earlier data we derived from Sprague-Dawley rats subjected to selective portal branch ligation and hepatic resection [20]. Partial hepatectomy (PH). For partial hepatectomy of 68% (PH), the two anterior liver lobes were removed as above [19]. Sham operation (SO). SO consisted of laparotomy and mobilization of the liver. Hepatocyte transplantation. A small left subcostal incision was made and the spleen was exposed. Hepatocytes (2 1 107 cells suspended in 1 ml of physiologic saline) were injected into the spleen using a 24-gauge Venocath (Becton Dickinson Vascular Access, Sandy, UT). During cell injection, the blood in- and out-flow in the splenic vessels was occluded to avoid immediate passage of cells into the liver. Spleen blood perfusion was reestablished after 15 min and the abdomen was closed in two layers using a continuous 4-0 prolene (D&G Monofil Inc., Manati, PR).

srgas

114

JOURNAL OF SURGICAL RESEARCH: VOL. 72, NO. 2, OCTOBER 1997

Experimental design. Animals received either intrasplenic injection of 2 1 107 syngeneic hepatocytes (Group I; n Å 46) or intrasplenic injection of physiologic saline (Group II; n Å 46). After 2 days, to allow hepatocyte engraftment, FHF was induced in all rats. All animals were monitored until death and autopsied to confirm the presence of intact omental lobes and necrotic right lobes. Survival Studies (n Å 20). Ten rats from each group were monitored until death. At autopsy, the omental liver lobes were removed, weighed, and fixed in 10% neutral buffered formalin. Effect of hepatocyte transplantation on blood chemistry and liver regeneration (n Å 60). Thirty Group I rats and 30 Group II rats were killed in batches of 6 rats at 12, 24, 36, 48, and 72 hr. One hour prior to euthanasia, 5-bromo-2-deoxyuridine (BrdU) was injected intraperitoneally (50 mg/kg body weight). When the animals were killed, blood was collected by aortic puncture and the omental liver lobes were removed, weighed, and processed for morphologic examinations. Effect of hepatocyte transplantation on lidocaine metabolism (n Å 18). In six Group I rats, six Group II rats, and six SO controls, at 24 hr after induction of FHF the jugular vein was cannulated with silastic tubing and a bolus of lidocaine (2 mg/kg body weight) was administered. Blood samples were collected at 0, 15, 30, and 45 min and analyzed (triplicates) for monoethylglycinexylidide (MEGX) content using a fluorescent polarization immunoassay (TDX analyzer, Abbott laboratories, Abbott Park, IL). Control animals (PH, SO; n Å 30/group). Control animals were also prepared in batches of six rats each, killed at the same time points, and evaluated post mortem in an identical manner.

Postoperative Evaluation Blood chemistry. Blood samples were analyzed for glucose, lactate, ammonia, and total bilirubin levels; activity of alanine aminotransferase (ALT), aspartate aminotransferase (AST), lactate dehydrogenase (LDH), prothrombin time (PT), and partial thromboplastin time (PTT) in the clinical laboratory. In addition, plasma HGF and TGF-b1 levels were determined, as described below. Morphological evaluation. Liver and spleen sections were stained with hematoxylin-eosin. In addition, sections of transplant-bearing spleens were immunostained for albumin using rabbit anti-rat albumin IgG antibodies and immunoperoxidase avidin–biotin complex (Vectastain, Vector Laboratories, Burlingame, CA), as earlier described [18]. Restitution of liver mass. Growth of the residual liver lobes (omental in FHF and SO rats; right and omental in PH controls) was assessed using

RÅ

RLs 0 RLo 1 100, RLo

where R is residual liver lobes growth (%), RLs is absolute weight of the residual liver lobes at the time of sacrifice, and RLo is estimated weight of the residual liver lobes at the time of surgery (omental liver lobes, 8% of the estimated original liver; original liver weight, 4% of the total body weight). Hepatocyte proliferative activity. Mitotic activity in the liver was evaluated in 5-mm-thick sections stained with hematoxylin-eosin. Mitosing hepatocytes were large polygonal cells forming hepatic cell plates. Cells in prophase before the dissolution of nuclear membranes and late telophases were excluded. Mitosing hepatocytes were counted in 100 consecutive high-power fields (hpf; 4001). The mitotic index was expressed per 1000 hepatic nuclei. For BrdU and PCNA indices, immunohistochemical staining was performed using an indirect, two-step labeling technique with peroxidase-conjugated immunoglobulin (Ig) G (Amersham NA 931), as previously described [17]. Monoclonal anti-BrdU and anti-PCNA antibodies were purchased from Sigma (BMC 9318) and Amersham (PC10), respectively. Proliferative activity in the remnant liver was expressed as the labeling index. The labeling index was the ratio of

AID

JSR 5175

/

6n26$$$$42

10-08-97 13:13:10

the number of either BrdU- or PCNA-positive hepatocytes to the total number of hepatocytes counted. In each liver, hepatocytes in 100 consecutive hpfs were counted. Plasma HGF level. Plasma HGF levels were determined using an HGF rat enzyme-linked immunosorbent assay (ELISA) kit (Institute of Immunology, Tokyo, Japan). Arterial blood samples were drawn into a tube containing EDTA and centrifuged at 3000 rpm for 10 min. After centrifugation, the samples were stored at 0707C until assay. A three-step specific sandwich method using mouse monoclonal antibodies to recombinant rat HGF and mouse monoclonal antibodies labeled by peroxidase was used in the ELISA system [22]. Standard curves for quantification of purified rat HGF were obtained using the rat HGF standard solution provided with the kit. The standard curve showed linearity from 0.4 to 10 ng/ml on a logarithmic scale. Plasma TGF-b1. Plasma TGF-b1 levels were determined using a TGF-b1 ELISA kit (Genezyme Co., Cambridge, MA). EDTA-treated samples were activated with HCl/NaOH. The standard curves showed linearity from 0.1 to 4.0 ng/ml on a logarithmic scale. HGF and c-met mRNA expression. Relative levels of HGF and cmet RNA expression in liver tissue were detected using the reverse transcriptase polymerase chain reaction (RT-PCR) technique. When animals were killed, the resected livers were quickly frozen in OCT with liquid nitrogen until assay. Total RNA was extracted according to Trizol (Gibco-BRL) protocol. Single strand complementary DNA (cDNA) was synthesized by reverse-transcriptase (AMV Reverse Transcriptase; Promega) using the oligo-dT (20 mer) and 1 mg of mRNA in 10 ml of reaction mixture. PCR primers were synthesized for rat HGF, rat c-met, and rat glyceraldehyde-3-phosphate dehydrogenase (G3PDH) cDNA, as reported previously [17]. Relative levels of mRNA were detected using equal amounts of input single-stranded DNA (2 ml per 30 ml) and tested in parallel using polymerase chain reaction. Polymerase chain reaction consisted of denaturation at 947C, annealing at 607C, and extension at 727C for 30 cycles using a Thermal Cycler (Perkin Elmer, Norwalk, CT). Eighteen-microliter aliquots were electrophoresed on 1.5% agarose gel. After staining with ethidium bromide, gels were photographed using Polaroid film (Polaroid Co., Cambridge, MA). Statistical analysis. Data were analyzed statistically using generalized-Wilcoxon test, one-way ANOVA, and Mann-Whitney U test when they were appropriate. P values of £ 0.05 were considered significant. Data are presented as means { standard deviations (SD).

RESULTS

All rats tolerated the operative procedures well and recovered uneventfully from anesthesia. Survival. Group I rats had significantly prolonged survival times when compared to Group II rats (73 { 22 hr vs. 33 { 9 hr, P õ 0.01; Fig. 2). All PH and SO control rats survived. Laboratory data. As reported earlier [17], changes in blood chemistry reflected rapid development of liver failure. Group I FHF rats showed statistically significant improvements in all liver function tests, including conversion of lidocaine into MEGX (Table 1, Fig. 3). In our opinion, necrosis of the right liver lobes contributed to the increased activity of serum transaminases and LDH. Since these enzymes are being primarily metabolized by the liver, decreased activity in transplant-bearing FHF rats might have resulted from improved metabolism by a larger hepatocellular mass (intrasplenic hepatocytes, regenerating liver). Liver remnant. In Group II nontransplanted FHF rats, there was only a 20–40% increase in the weight

srgas

115

EGUCHI ET AL.: LIVER REGENERATION

FIG. 2. Survival rate in rats with FHF (*P õ 0.01 vs. Group II).

FIG. 3. Formation of MEGX in FHF rats and SO controls. For the indicated time, data are plotted as means { SD for each group (*P õ 0.05 vs. Groups I and II, #P õ 0.05 vs. Group II).

of the remnant liver lobes (Fig. 4). At the same time, the omental liver lobes in Group I transplant-bearing FHF rats increased in weight over a 72-hr period to the same or higher degree (220 { 34%; n Å 4) as the remnant liver in PH controls (162 { 13%; n Å 6) (Fig. 4). No changes in liver weight were observed in SO controls. On light microscopy, the omental lobes in Group II nontransplanted FHF rats showed interstitial edema and extensive lipid vacuolization of hepatocytes. Similar changes, although of much lesser degree, were seen in the omental liver lobes of Group I FHF rats. At the same time, those lobes showed no evidence of embolism caused by intrasplenically transplanted hepatocytes. In PH controls the remnant liver showed a typical posthepatectomy steatosis. The livers of SO rats showed no abnormal changes. In all spleens recovered from Group I FHF rats, there were numerous clusters of cells showing reactivity to anti-rat albumin IgG (Fig. 5). All PH controls showed signs of vigorous liver regeneration (Figs. 6 and 7). In contrast, no signs of cell proliferation were detected in the SO rat livers.

In Group II FHF rats, the BrdU and PCNA labeling indices were at the level of 0.1% or lower and no hepatocytes showed mitotic figures (Fig. 6). Although in Group I FHF rats, the BrdU labeling index was also low (0.1 { 0.1%) at 24 hr postoperatively, it was markedly increased at 48 hr (1.2 { 0.4%; n Å 6) and 72 hr (3.3 { 1.6%; n Å 4). The changes in PCNA labeling followed a similar pattern (Figs. 6 and 8). In Group I FHF rats, mitotic activity was conspicuous at all time points studied (Figs. 7 and 9). After PH, plasma HGF levels were elevated from 24 hr posthepatectomy. In Group II FHF rats, plasma HGF levels were markedly elevated at all time points studied. Group I FHF rats had lower plasma HGF levels than Group II FHF rats and, at the same time, very similar to those found in PH controls (Table 2). After PH, plasma TGF-b1 levels were mildly elevated at all time points studied. Group II FHF rats had

TABLE 1 Biochemical Changes 24 hr Baseline Glucose (mg/dl) ALT (IU/l 1 10) ASL (IU/l 1 10) LDH (IU/l 1 10) T. bil (mg/dl) Ammonia (mmole/dl) Lactate (mEq/L) PT (sec) PTT (sec)

110 5.5 4.0 19 0.3 48 1.2 12.0 24.0

{ { { { { { { { {

18 2.0 1.7 2 0.1 10 0.2 1.4 2.0

Group I 52 377 490 1701 1.5 550 2.0 29.5 36.2

{ { { { { { { { {

36 hr Group II

8 81* 237* 831 0.8* 118* 0.3* 5.9 9.5

70 1117 948 2620 2.7 770 6.2 25.0 36.7

{ { { { { { { { {

37 148 46 829 0.6 131 3.0 1.6 12.4

Group I 34 472 458 580 2.9 383 2.3 30.0 31.5

{ { { { { { { { {

18 41* 62* 67* 1.0* 40* 0.2* 1.0* 2.2*

48 hr

Group II 26 686 747 2920 5.8 1743 6.6 39.0 50.5

{ { { { { { { { {

14 243 130 927 1.2 454 2.8 5.0 8.4

Group I 40 281 371 478 5.0 499 3.3 43.2 33.4

Note. Each baseline value represents mean { SD for entire group (Group I and II). * P õ .05 vs. Group II. All differences between Group I, II, and baseline are at P õ 0.05.

AID

JSR 5175

/

6n26$$$$43

10-08-97 13:13:10

srgas

{ { { { { { { { {

20 113 114 140 0.6 152 1.1 12.6 9.1

72 hr Group II — — — — — — — — —

Group I 64 118 236 618 6.6 485 4.0 50.0 42.7

{ { { { { { { { {

40 61 161 689 2.8 235 2.7 9.2 9.6

Group II — — — — — — — — —

116

JOURNAL OF SURGICAL RESEARCH: VOL. 72, NO. 2, OCTOBER 1997

In SO rats, plasma HGF and TGF-b1 levels remained at or below baseline levels throughout the study period. After PH, HGF was expressed in the remnant liver at 12 hr and later. In Group II FHF rats, HGF was weakly expressed at 24 hr postoperatively only, whereas in Group I FHF rats, it was expressed at 24 hr and later. In both Group I FHF rats and PH controls, HGF receptor c-met was expressed in the remnant liver at all time points, whereas in Group II FHF rats it was not expressed until 24 hr postoperatively (Fig. 10). As expected, SO controls demonstrated no HGF mRNA signal and low c-met mRNA signals at all time points studied. DISCUSSION

FIG. 4. Restitution of liver mass following FHF and PH. For the indicated time, data are plotted as means { SD for each group (*P õ 0.05 vs. Groups I and II, #P õ 0.05 vs. Group II, †P õ 0.05 vs. PH).

higher plasma TGF-b1 levels than Group I FHF rats and virtually identical to those found in PH controls (Table 3).

Fulminant hepatic failure is a clinical syndrome resulting from severe hepatocellular injury (dysfunction) and necrosis. The majority of patients will rapidly deteriorate and die unless they urgently undergo liver transplantation; only a small percentage of these patients will recover and subsequently have normal liver function [1–3]. Due to our limited knowledge of the pathophysiology of acute liver failure and factors responsible for healing of the injured liver, we are unable to identify patients which have a potential for recovery

FIG. 5. Microscopic section of spleen from FHF rat, treated by hepatocyte transplantation (hematoxylin-eosin; original magnification 160). On subsequent immunohistochemical staining for albumin, the epithelioid cells occupying splenic parenchyma were found to be albumin positive (Vectastain ABC method).

AID

JSR 5175

/

6n26$$$$43

10-08-97 13:13:10

srgas

EGUCHI ET AL.: LIVER REGENERATION

117

FIG. 6. Labeling index (PCNA) in the liver of FHF rats. For the indicated times, data are plotted as means { SD for each group (*P õ 0.05 vs. Group I and II, #P õ 0.05 vs. Group II).

without liver transplantation [4]. Also, we do not know how to stimulate the residual viable hepatocytes to proliferate which should, at least in theory, increase the effectiveness of medical therapy. Hepatocyte transplantation has been used by many

FIG. 7. Mitotic index in the liver of Group I FHF rats and PH controls. For the indicated times, data are plotted as means { SD for each group (*P õ 0.05 vs. Group I).

AID

JSR 5175

/

6n26$$$$43

10-08-97 13:13:10

investigators to demonstrate metabolic support and improve survival in rats with hepatic failure [5–16, 18]. However, none of those reports have examined the impact of cell therapy on the regenerative response in the native liver. Our hypothesis was that in rats with FHF, hepatocyte transplantation can not only provide metabolic support by increasing the available functional liver mass but also facilitate regenerative response in the remnant liver. In order to test this hypothesis, we used a small animal (rat) model of FHF, which has a number of physiological and biochemical features seen clinically in FHF, including severely impaired ability of the residual liver tissue to regenerate [17]. As in the previous study, FHF rats became comatose and died within 24–48 hr with signs of rapidly developing liver failure. Despite high plasma HGF (hepatocyte growth stimulator) levels, the residual livers displayed no signs of hepatocyte proliferation. This could be explained, at least partially, by a significant delay in tissue expression of the HGF receptor c-met and elevated plasma TGF-b1 (hepatocyte growth inhibitor) levels. Group I rats, which received intrasplenic injection of isolated syngeneic hepatocytes 2 days prior to induction of FHF, behaved differently. Not only survival time more than doubled in these rats and all liver function tests improved but also, more importantly, a vigorous regenerative response occurred in the residual liver. The latter, as judged by mitotic activity and labeling indices (BrdU, PCNA), was associated with positive tissue c-met expression and normalization of plasma

srgas

118

JOURNAL OF SURGICAL RESEARCH: VOL. 72, NO. 2, OCTOBER 1997

FIG. 8. Group I FHF rat liver at 72 hr stained with anti-PCNA antibody. Dark nuclei represent positive staining, indicating hepatocytes in S through M phase of the cell cycle (magnification 1100).

growth factor (HGF, TGF-b1) levels. Here, ‘‘normalization’’ refers to data obtained in PH controls, which are known to respond to partial hepatectomy of 68% with maximal liver regenerative response [23]. Although some of the intrasplenically injected hepatocytes could have migrated into the liver, the observed fourfold increase in liver remnant weight did not result from proliferation of translocated hepatocytes. First, whatever the number of translocated hepatocytes, more than 90% of them were eliminated during subsequent induction of FHF (partial hepatectomy / right lobe necrosis). Second, to account for the observed increase in residual liver weight, hepatocytes which translocated to the omental liver lobes would have to go through at least seven cell cycles in as little as 72 hr. Third, in all spleen sections we have seen an abundance of hepatocytes, which avidly incorporated bromodeoxyuridine (data not shown). HGF is widely recognized as the most potent stimulator of hepatocyte growth and DNA synthesis in vitro [24] as well as one of the key regulators of liver regeneration after partial hepatectomy or hepatic injury [25]. Less clear, however, is the role of HGF in the pathophysiology of fulminant and subfulminant hepatitis. In these settings, HGF activity and the HGF mRNA in the liver are markedly increased and so is the HGF level in the circulating blood [25, 26]. The latter is

AID

JSR 5175

/

6n26$$$$43

10-08-97 13:13:10

linked to a decrease in hepatic clearance [27] and concomitant liver regeneration [28]. In the present study, nontransplanted FHF rats showed no proliferative activity in their remnant livers and, therefore, an increase in blood HGF level reflected decreased hepatic clearance rather than continuing supply from extrahepatic sources. It seems reasonable to assume that in FHF rats transplanted with syngeneic hepatocytes, plasma levels of both HGF and TGF-b1 were much lower, because they were utilized by the regenerating liver and, possibly, by hepatocytes seeded in the spleen. We are unable to fully explain why the presence of additional 20 million syngeneic hepatocytes in an ectopic site had such a profound effect on FHF rats. While prolongation in survival time and improvement in liver function, as determined biochemically, could be attributed to the increase in the existing hepatocyte mass from 8 to 10%, we cannot rule out the possibility that transplanted hepatocytes enhanced, in some way, the metabolic performance of the native liver. In fact, the latter seems probable, because transplanted hepatocytes had only 2 days to engraft and undertake highly differentiated functions. It is even more difficult to understand the recovery of regenerative response in the native liver. In transplanted FHF rats, proliferation indices (PCNA, BrdU, mitosis) were lower at 24 and 48 hr when com-

srgas

119

EGUCHI ET AL.: LIVER REGENERATION

FIG. 9. Group I FHF rat liver at 72 hrs postinduction. Several hepatocytes show distinct mitotic figures (hematoxylin-eosin; magnification 1400).

pared to those in partially hepatectomized controls. However, these time points are known to be associated with maximal DNA-SA in normal rats subjected to partial (68%) hepatectomy [22]. Rapid liver growth which occurred in transplanted FHF rats between 48 and 72 hours postinduction suggests that regeneration response in FHF rats was delayed. In the normal rat liver, hepatocytes actively synthesizing DNA can be found only occasionally, perhaps 1/1000, and mitosis is virtually absent [20, 22]. Therefore, the observed BrdU and PCNA indices and mitotic activity of 3% at 72 hr postinduction were extremely significant. At present, we suggest that lowering of plasma

TGF-b1 levels, and not enhanced utilization of HGF, played a crucial role in the earlier c-met expression and initiation and maintenance of hepatocyte proliferation in the remnant liver. This is in agreement with our earlier suggestion [17] that agents neutralizing inhibitors of hepatocyte proliferation [29] (e.g., antibody against TGF-b1) may be of greater value in facilitating repair of the injured liver than hepatocyte mitogens such as HGF. The possible involvement of other growth factors (e.g., TGF-a) and modulators of liver regeneration (e.g., interleukin-6) needs to be elucidated [30]. All transplanted FHF rats died. It is of interest that

TABLE 2 Plasma HGF Levels Group (ng/ml)

0 hr

Group I (FHF w/Tx) Group II (FHF w/Saline) PH SO

0.13 0.10 0.10 0.08

{ { { {

24 hr

0.10 0.01 0.02 0.05

1.82 2.82 1.51 0.17

{ { { {

36 hr

0.39* 3.01* 0.70* 0.06

1.26 3.14 1.00 0.21

a

N/A; not available (no survivors). * P õ .05 vs. SO. # P õ .05 vs. Group II.

AID

JSR 5175

/

6n26$$$$43

10-08-97 13:13:10

srgas

{ { { {

0.30* 0.84*,# 1.12* 0.05

48 hr

72 hr

1.33 { 1.03* N/Aa 2.71 { 2.62* 0.14 { 0.07

1.74 { 1.03* N/A 1.83 { 1.63* 0.12 { 0.26

120

JOURNAL OF SURGICAL RESEARCH: VOL. 72, NO. 2, OCTOBER 1997

TABLE 3 Plasma TGF-b1 Levels Group (ng/ml)

0 hr

Group I (FHF w/Tx) Group II (FHF w/Saline) PH SO

17.7 22.8 21.0 23.1

{ { { {

4.9 5.2 6.3 5.5

24 hr 39.4 60.2 36.8 19.8

{ { { {

36 hr

8.0* 15.5*,# 25.8* 7.9

46.8 70.8 46.0 19.0

{ { { {

5.6* 14.3*,# 29.8* 9.9

48 hr

72 hr

53.3 { 10.6* N/Aa 55.2 { 35.1* 24.3 { 9.1

82.0 { 13.3* N/A 96.1 { 37.7* 18.9 { 2.6

a

N/A; not available (no survivors). * P õ .05 vs. SO. # P õ .05 vs. Group II.

at the time of death, the residual liver approached 30% of the estimated original liver mass. This amount of viable liver tissue is exactly what remains after PH, and yet partially hepatectomized rats display only mild and transient hepatic dysfunction postoperatively [3]. While this finding per se is not new [31], it suggests that in FHF, evidence of ongoing liver regeneration may be of limited prognostic value. As mentioned earlier, during the first 24–36 hr of FHF, increase in liver remnant weight in transplanted rats was small and proliferation indices were much lower than in PH rats. Afterward, growth of the native liver accelerated. However, by that time the animal was in severe liver failure. It is possible that during rapid liver growth, dividing hepatocytes and those preparing for mitosis did not perform differentiated functions. In our earlier studies utilizing this FHF rat model, blood glucose levels remained above 50 mg/dl until animal death and repeated administration of glucose had no effect on survival [17]. Therefore, we believe that in this study, hypoglycemia did not account for differences in survival between the two groups. Portal hypertension did not seem to be the cause of death either. Rats have extensive porto-sys-

temic communications, especially in the retroperitoneal and perisplenic areas. Additionally, they have the unique ability to rapidly form new collaterals and do not develop submucosal varices. In earlier experiments, we found that in rats with acute constriction of the portal vein to less than 20% of the cross-sectional area, there was only a modest (20–22 cm H2O) increase in portal pressure after 5 min, which persisted throughout the first postoperative week [32]. In this study, although the whole portal flow was directed through the omental liver lobes, none of the rats died of acute splanchnic venous stasis and sequestration of blood into the gut lumen [33]. Finally, in searching for the cause of death in transplanted rats, we found that Group 1 and 2 animals developed sepsis. Blood cultures obtained at 24 hrs postinduction were negative, but those at 30 hr and later were positive for several bacteria, including Escherichia coli and Staphylococcus aureus (data not shown). We have attempted to enhance survival in transplanted FHF rats using antibiotic therapy perioperatively, without success (data not shown). Further manipulations are currently being examined to result in long-term animal survival.

FIG. 10. HGF, c-met, G3PDH gene expression in the liver of Group I, Group II, PH, and SO rats. Animals were killed at 12, 24, 36, 48, and 72 hr after surgery. After extraction of RNA, 1 mg of total RNA was used for RT-PCR analysis, as described under Methods.

AID

JSR 5175

/

6n26$$$$43

10-08-97 13:13:10

srgas

EGUCHI ET AL.: LIVER REGENERATION

Whatever the mechanisms behind the observed effects of hepatocyte transplantation, this study suggests that in rats undergoing PH and with limited (24%) liver necrosis, there is a certain critical (Ç10%) functional liver mass needed for liver regeneration to occur. When such a critical liver mass is reached, even a small increase in the population of viable hepatocytes may have a profound effect on liver function and regenerative response in the residual liver. Clinical relevance of this notion seems to be highlighted by preliminary reports on hepatocyte transplantation in acute liver failure patients [31, 34] as well as by our own experience with hepatocyte-based artificial liver used in patients with FHF as a bridge to orthotopic liver transplantation [35]. In summary, we have demonstrated that rats with FHF and less than 10% viable liver mass died from liver failure with no signs of liver regeneration. Heterotopic transplantation of a small number of isolated syngeneic hepatocytes (2% liver mass) prolonged survival time, improved liver function, and triggered a regenerative response in the native liver. The latter effect was associated with accelerated tissue (liver) expression of c-met and lowering of plasma HGF and TGF-b1 levels. Even though the native liver increased in size by ú200% and showed distinct hepatocyte proliferation, the transplant-bearing rats expired due to an inability of the regenerating liver to support the animal. It is suggested, that in the medical management of FHF patients, agents neutralizing hepatocyte growth inhibitors may be of greater value in facilitating repair of the injured liver than hepatocyte mitogens.

121

munosuppression using transplanted microencapsulated hepatocytes. Hepatology 12: 1342, 1990. 9. Sutherland, D. E. R., Numata, M., Matas, A. J., Matas, A. J., Simmons, R. L., and Najarian, J. S. Hepatocellular transplantation in acute liver failure. Surgery 82: 124, 1977. 10. Ribeiro, J., Nordlinger, B., Ballet, F., Cynober, L., CoudrayLucas, C., Baudrimont, M., Legender, C., Delelo, R., and Panis, Y. Intrasplenic hepatocellular transplantation corrects hepatic encephalopathy in portacaval-shunted rats. Hepatology 15: 12, 1992. 11. Vogels, B. A. P. M., Mass, M. A. W., Bosma, A., and Chamuleau, A. F. M. Significant improvement of survival by intrasplenic hepatocyte transplantation in totally hepatectomized rats. Cell Transplant. 5: 369, 1996. 12. Moscioni, A. D., Rozga, J., Chen, S., Naim, A., Scott, H. S., and Demetriou, A. A. Long-term correction of albumin levels in the Nagase analbuminemic rat: Repopulation of the liver by transplanted normal hepatocytes under a regeneration response. Cell Transplant. 5: 499, 1996. 13. Eguchi, S., Rozga, J., Chen, S. C., Wang, C-C., Rosenthal, R., Hewitt, W. R., Middleton, Y., Lebow, L. T., and Demetriou, A. A. Treatment of hypercholesterolemia in the Watanabe rabbit using allogeneic hepatocellular transplantation under a regeneration stimulus. Transplantation 62: 588, 1996. 14. Demetriou, A. A., Reisner, A., Sanchez, A., Levenson, S. M., Moscioni, A. D., and Chowdhury, J. R. Transplantation of microcarrier-attached hepatocytes into 90% partial hepatectomized rats. Hepatology 8: 1006, 1988. 15. Baumgartner, D., LaPlante-O’Neill, P. M., Sutherland, D. E. R., and Najarian, J. S. Effects of intrasplenic injection of hepatocytes, hepatocyte fragments and hepatocyte culture supernatants on D-galactosamine-induced liver failure in rats. Eur. Surg. Res. 15: 129, 1983. 16. Makowka, L., Rotstein, L. E., Falk, R. E., Falk, J. A., Zuk, R., Langer, B., and Blendis, L. M. Studies into the mechanism of reversal of experimental acute hepatic failure by hepatocyte transplantation. Can. J. Surg. 24: 39, 1981. 17. Eguchi, S., Kamlot, A., Ljubimova, J., Hewitt, W., Demetriou, A. A., and Rozga, J. Fulminant hepatic failure in rats: Survival and effects on blood chemistry and liver regeneration. Hepatology 24: 1452, 1996.

REFERENCES 1. Trey, C. The fulminant hepatic failure surveillance study: Brief review of the effects of presumed etiology and age of survival. Can. Med. Assoc. J. 106(suppl): 522, 1992.

18. Rozga, J., Holzman, M., Fujioka, H., Morsiani, E., Moscioni, A. D., and Demetriou, A. A. Repeated intraportal hepatocyte transplantation in rats. Cell Transplant. 4: 237, 1994.

2. Ascher, N., Lake, J. R., Emond, J. C., and Roberts, J. P. Liver transplantation for fulminant hepatic failure. Arch. Surg. 128: 677, 1993.

19. Higgins, G. M., and Anderson, R. M. Restoration of the liver of the white rat following partial surgical removal. AMA Arch. Pathol. 12: 186, 1931.

3. Bismuth, H., Samuel, D., Castaing, D., Adam, R., Saliba, F., Johan, M., and Azoulay, D. Orthotopic liver transplantation in fulminant and subfulminant hepatitis. The Paul Brousse Experience. Ann. Surg. 222: 109, 1995.

20. Rozga, J., Jeppsson, B., and Bengmark, S. Portal branch ligation in the rat: reevaluation of a model. Am. J. Pathol. 125: 300, 1986.

4. Pauwels, A., Mostefa-Kara, N., Florent, C., and Levy, V. G. Emergency liver transplantation for acute liver failure. Evaluation of London and Clichy criteria. J. Hepatol. 17: 124, 1993.

21. Yamada, A., Matsumoto, K., Iwanari, H., Sekiguchi, K., Kawata, S., Matsuzawa, Y., and Nakamura, T. Rapid and sensitive enzyme-linked immunosorbent assay for measurement of HGF in rat and human tissue. Biochem. Res. 16: 105, 1995.

5. Groth, C. G., Arborgh, B., Bjorken, C., Sundberg, B., and Lundgren, G. Correction of hyperbilirubinemia in the glucuronyltransferase deficient rat by intraportal transplantation. Transplant. Proc. 9: 313, 1977.

22. Grisham, J. W. A morphologic study of deoxyribonucleic acid synthesis and cell proliferation in regenerating rat liver: Autoradiography with thymidine-3H. Cancer Res. 22: 842, 1962.

6. Matas, A., Sutherland, D., and Steffes, M. Hepatocellular transplantation for metabolic deficiences: Decrease of plasma bilirubin in Gunn rats. Science 192: 892, 1976. 7. Onodera, K., Kasai, S., Kato, K., Sawa, M., Mito, M., and Nozawa, M. Effects of intrasplenic hepatocyte transplantation in rats congenitally deficient in ascorbic acid biosynthetic enzyme. Cell Transplant. 3: S31, 1994. 8. Dixit, V., Darvasi, R., Arthur, M., Brezina, M., Lewin, K., Gitnick, G. Restoration of liver function in Gunn rats without im-

AID

JSR 5175

/

6n26$$$$44

10-08-97 13:13:10

23. Arakaki, N., Hirono, S., Ishii, T., Kimoto, M., Kawakami, S., Nakayama, H., and Tsubouchi, H. Identification and partial characterization of two classes of receptors for human hepatocyte growth factor on adult rat hepatocytes in primary culture. J. Biol. Chem. 267: 7107, 1992. 24. Tsubouchi, H., Niitani, Y., Hirono, S., Nakayama, H., Gohda, E., Arakaki, N., Sakiyama, O., Takahashi, K., Kimoto, M., Kawakami, S., Setoguchi, M., Tachikawa, T., Shin, S., Arima, T., and Daikuhara, Y. Levels of the human hepatocyte growth factor in serum of patients with various liver diseases determined

srgas

122

JOURNAL OF SURGICAL RESEARCH: VOL. 72, NO. 2, OCTOBER 1997 by an enzyme-linked immunosorbent assay. Hepatology 13: 1, 1991.

25. Tsubouchi, H., Hirono, S., Gohda, E., Nakayama, H., Takahashi, K., Sakiyama, O., Miyazaki, H., Sugihara, J., Tomita, E., Muto, Y., Daikuhara, Y., and Hashimoto, S. Clinical significance of human hepatocyte growth factor in blood from patients with fulminant hepatic failure. Hepatology 9: 875, 1989. 26. Liu, K-X., Kato, Y., Narukawa, M., Kim, D. C., Hanano, M., Higuchi, O., and Nakamura, T. Importance of the liver in plasma clearance of hepatocyte growth factor in rats. Am. J. Physiol. 263: G642, 1992. 27. Milandri, M., Gaub, J., and Ranek, L. Evidence for liver cell proliferation during fatal acute liver failure. Gut 21: 423, 1980. 28. Russell, W. E., Coffey, R. J., Ouellette, A. J., and Moses, H. L. Type b transforming growth factor reversibly inhibits the early proliferative response to partial hepatectomy in the rat. Proc. Natl. Acad. Sci. U.S.A. 85: 5126, 1988. 29. Cressman, D. E., Greenbaum, L. E., DeAngelis, R. A., Ciliberto, G., Furth, E. E., Poli, V., and Taub, R. Liver failure and defective hepatocyte regeneration in interleukin-6-deficient mice. Science 274: 1379, 1996.

AID

JSR 5175

/

6n26$$$$44

10-08-97 13:13:10

30. Moriyama, M., Mikayama, I., Shiota, M., Uesugi, K., Kannan, Y., Ohta, M., Kimura, K., and Sugano, T. Decreased ureagenesis from alanine, but not from ammonia and glutamine, in the perfused rat liver after partial hepatectomy. Hepatology 23: 1584, 1996. 31. Strom, S. C., Fisher, R. A., Thompson, M. T., Sanyal, A. J., Cole, P. E., Ham, J. M., and Posner, M. P. Hepatocyte transplantation as a bridge to orthotopic liver transplantation in terminal liver failure. Transplantation 63: 559, 1997. 32. Rozga, J., Jeppsson, B., Hagerstrand, I., Bengmark, S. Acute portal vein stenosis. An experimental study on portal circulation and hepatocplenic function. Acta Chir. Scand. 151: 125, 1985. 33. Mappes, G. Experimentelle erzeugung einer portalen hypertension bei der ratte. Arch. Klin. Chir. 295: 795, 1960. 34. Mito, M., and Kusano, M. Hepatocyte transplantation in man. Cell Transplant. 2: 65, 1993. 35. Watanabe, F. D., Mullon, J-P., Hewitt, W. R., Arkadopoulos, N., Kahaku, E., Eguchi, S., Khalili, T., Arnaout, W., Shackleton, C. R., Rozga, J., Solomon, B., and Demetriou, A. A. Clinical experience with a bioartificial liver in the treatment of severe liver failure. A phase I clinical trial. Ann. Surg. 225: 484, 1997.

srgas