The First Department of Surgery, Nagoya City University Medical School, Nagoya, Japan

Runing Title: liver regeneration and the exocrine pancreas

Key Words: Hepatocyte growth factor (HGF), Liver regeneration, Partial hepatectomy, Subtotal pancreatic duct ligation

Address correspondence and reprint requests to: Takashi Hashimoto, M.D., First Department of Surgery, Nagoya City University Medical School, 1 Kawasumi Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Japan
Phone: 81-52-853-8226, Fax: 81-52-842-3906

To further investigate roles of the exocrine pancreas in liver regeneration, we conducted a study with a subtotal pancreatic duct ligation model in rats. [Materials and methods] Eighty SD rats were divided into four groups receiving the following treatments: control group (sham laparotomy); subtotal pancreatic duct ligation (PL); 68% heptectomy (Hx); and PL plus Hx (PL+Hx). The surgical schedule was divided into two stages: stage one, subtotal ligation of pancreatic duct was performed for Groups 2 (PL) and 4 (PL+Hx); stage two, 68% hepatectomy was performed for Groups 3 (Hx) and 4 (PL+Hx) one week later. On postoperation days 1, 2, 3 and 7, five rats of each group were sacrificed. Liver wet weight (LWW) was measured and liver regeneration rate (LRR) was calculated. BrdU-labeling and mitotic indices in liver were examined, and insulin and HGF concentrations in serum were also measured. [Results] Compared with the Hx group, the LRR in the PL+Hx group markedly decreased on postoperation days 1 and 3. The peak of hepatocyte DNA synthesis was also suppressed with mitosis showing significant reduction and delay, the insulin concentration on postoperation day 1 was decreased and the peak of HGF showed delay and depression. [Conclusions] Subtotal pancreatic duct ligation suppresses liver regeneration in the initial period. This is related to changes in HGF and insulin serum concentrations and suggests that the exocrine pancreas provides factor supporting early liver regeneration.

After partial hepatectomy the remnant liver rapidly proliferates and returns to the original mass within 7 to 10 days in rats (1, 2, 3). This process involves a variety of growth factors and cytokines (1, 4), and the pancreas plays an important role as an origin of these (5). It is known that the endocrine pancreas supports liver regeneration by its production of insulin (6). We previously reported that the exocrine pancreas may be necessary to support liver regeneration (7). In addition, HGF (scatter factor), a key factor for liver regeneration (8), is presented in pancreatic exocrine portion in rats at high levels (9). However, it remains unclear as to what relation exists between blood HGF concentration and the exocrine pancreas during liver regeneration.
It is sometimes necessary to perform major hepatectomy in cases with chronic damage of the exocrine pancreas, chronic pancreatitis, or with pancreatoduodenectomy because of advanced pancreatobiliary malignancies. These clinic conditions contribute to hepatic failure from unsuccessful regeneration of remnant liver.
Therefore, we conducted a study of the liver regeneration after hepatectomy with subtotal ligation of the pancreatic duct in rats to further investigate the role of exocrine pancreas in liver regeneration.

Eighty adult male SD rats weighting from 200 to 300g were acclimatized to our laboratory conditions for 1 week. They were housed in a temperature controlled (23Ž‘C) room under a 12-hr light/dark cycle. The animals were given tap water and standard rat chow, and randomly divided into the following four groups: Group 1 (n=20), sham laparotomy (Lapa.); Group 2 (n=20), pan-ceatic duct subtotal ligation (PL); Group 3 (n=20), 68% hepatectomy (Hx); and Group 4 (n=20), subtotal pancreatic duct ligation plus 68% hepatectomy (PL+Hx). After fasting overnight but with access to water, operations were performed with sterile surgical techniques under anesthesia induced with intraperitoneal thiopental sodium (30mg/kg of body weight) and ether inhalation. The experimental schedule was divided into two stages: stage one, laparotomy was performed for groups 1 (Lapa.) and 3 (Hx), and subtotal pancreatic duct ligation for groups 2 (PL) and 4 (PL+Hx). Stage two, a second laparotomy was performed for groups 1 (Lapa.) and 2 (PL), and 68% heptectomy was performed for groups 3 (Hx) and 4 (PL+Hx), one week later.
Subtotal ligation of the pancreatic duct was achieved according to Hultguist's procedure (10). In brief, the duodenal portion of the pancreas was first carefully freed from the mesentery of the colon. After this, the parenchyma of the pancreas nearest the bile duct was divided into 3 or 4 parts, and double ligatures were performed around each of these, using 3-0 nonabsorbable silk thread. Thereafter the parenchyma between the ligatures was cut off. With this procedure, a careful check was made to ensure that no bridges of pancreas tissue remained between the lineal part and the area around the common bile duct. Hepatectomy (68%) was performed according to Higgin's procedure (2). The medial and left lateral lobes were excised after placement of a suture ligature around the vascular pedicle. In the sham group, the pancreatic parenchyma around the bile duct was exposed and rubbed between the fingertips.
On postoperation days 1, 2, 3 and 7 of stage two, 5 rats of each group were sacrificed by exsanguination through the vena cava. Serum samples were obtained and stored at -80C for the measurement of HGF, insulin, albumin (Alb), total bilirubin (TB), aspartate aminotransferase (GOT), and amylase. Liver and pancreas were removed and immediately weighed. The resected liver and pancreas specimens were excised and fixed in 10% formaldehyle for histologic investigation. One hour before sacrifice, 5-bromo2-deoxyuridine (20mg/kg of body weight) was injected intraperitoneally to allow determination of the BrdU-libeling index (BrdU LIs) in liver.
Liver regeneration rates (LRR) were derived from the resected liver weight and the liver weight at autopsy using Child's formula (11).
(Liver weight at autopsy - estimated residual liver weight

at the time of hepatectomy) x 100
resected liver weight

For histologic examination, liver specimens routinely processed for embedding in paraffin, sectioned and stained with hematoxylin-eosin for light microscopic examination. Counts of mitotic figures in 20 fields for each group were made under high-power magnification, the results being expressed as mean numbers per 1000 parenchymal cells (mitotic index, MI).
Immunohistochemical staining of incorporated bromodeoxyuridine was performed to generate BrdU-labeling indices (BrdU LIs). The staining procedure was accomplished as previously described (12) and positive cells were counted in 10 randomly selected high-power fields (x200) and the results expressed as the numbers of hepatocytes labeled per 100 cells.
Data are presented as means ‘ή standard errors. Statistical analysis of the results was conducted with the paired Student's t-test.

Histologic changes in the ligated pancreas
Histopathological findings revealed that, at 1 week after the pancreatic duct ligation, the acinar parenchyma was almost completely lost, whereas the islet parenchyma was relatively normal. Ducts were markedly dilated and interstitial tissue demonstrated slight edema, fibrosis and inflammation. Islets progressively proliferated but acinar cells did not essentially reappear during the experimental period.

Liver Wet Weight (LWW)
The liver weight in the PL group was significantly decreased on the postoperation day 2 (8.17Ž±2.16, 11.05Ž±0.96, P < 0.05) of stage two, this continuing on the postoperation days 3 and 4 compared with the sham group (Fig. 1).

Liver regeneration rate (LRR)
The liver regeneration rate in the PL+Hx group was significantly lower on postoperation day 1 (6.66Ž±4.32%, 14.63Ž±4.59%, P < 0.05), and markedly higher on postoperation day 3 (49.93Ž±11.62%, 38.14Ž±7.11%, P < 0.05) than the Hx group (Fig. 2).

BrdU labeling index (BrdU LI)
BrdU positive nuclei were almost absent in the sham and PL groups. In the Hx and PL+Hx groups, peaks were found on postoperation day 1, that for the PL+Hx group (23.46Ž±14.94%) being significantly lower than the Hx group value (54.92Ž±13.28%, P <0.01) (Fig. 3).

Mitotic index (MI)
Mitotic cells were essentially absent in sham and PL groups. In the PL+Hx group, the mitotic cell peak (28.10Ž±12.79 / 1000 cells) after hepatectomy was delayed 24 hours and significantly decreased compared with the Hx group(43.93Ž±7.73 / 1000 cells, P < 0.05) (Fig. 4).

Insulin concentration
The serum insulin level in the sham group was higher than in PL group and there were significant differences on postoperation day 2 (10.60Ž±4.26, 3.40Ž±2.49, P < 0.05). In the PL+Hx group, on postoperation day 1 insulin levels was significantly lowered, then showed a tendency for increase on postoperation day 2 after hepatectomy compared to the Hx group (Table 1).

HGF concentration
The HGF concentration in was increased in the Hx group on the first day after hepatectomy compared with the PL+Hx group; whereas it was delayed until the third day (Table 2).

Hepatic and pancreatic functions
Data for serum Alb, TB, GOT and amylase concentrations are summarized in Table 1. In the sham, PL, and Hx groups the Alb concentrations showed relative increase on the postoperation day 1 in stage two compared with postoperation day 2. The Alb levels progressively increased in the sham group, but fluctuated in the PL group. Both PL and PL+Hx groups later showed low Alb concentrations. TB and GOT concentrations in serum did not exhibit any significant intergroup differences. The amylase concentration in the PL group had a tendency for decrease on postoperation days 1, 3, and 7 in stage two compared with the sham group; similar changes were observed in the PL+Hx group on postoperation days 1 and 2 compared with the Hx group (Table 3).

Liver regeneration is controlled by various growth factors and cytokines such as HGF, EGF, insulin, transforming growth factor-a(TGF-a), interleukin-6, tumor necrosis factor and norepinephrine (1, 4). Among these, HGF, EGF and TGF-a are complete mitogens for hepatocytes in culture (1). HGF displays a 5-10 fold higher potency than epidermal growth factor (EGF) and tumor growth factor-a (TGF-a) on a molar basis (13, 14), and its concentration in plasma rapidly increases 15-17 fold following hepatectomy , above the mitogenic range for hepatocytes (15). EGF and TGF-a plasma levels show only a small increase after partial hepatectomy (16, 17, 18). The kinetics of hepatocytic DNA synthesis and cell division are similar to the HGF levels in plasma (1, 7). Thus it is considered that HGF is a definitive factor for liver regeneration.
Previous studies have indicated that HGF rapid rise in plasma after hepatectomy occurs within 1 to 2 hours (15), whereas its mRNA expression in hepatic Ito cells increases 3-6 hours after partial hepatectomy (19, 20). The elimination rate in the liver, as the main organ for clearing circulating HGF (21), does not appreciably affect the magnitude of the HGF rise in plasma after partial hepatectomy (22), suggesting the rapid elevation is mainly due to the release of preexisting stores from different tissues.
Recent studies demonstrated that matrix break down in the regenerating liver itself is one source of HGF (23, 24). Immunohistochemical studies have shown in fact that HGF is widely distributed in most surface epithelia, lung, kidney, brain, and at high levels in the exocrine pancreas of rats (8). The pancreas plays important supporting roles not only for normal liver but also for liver regeneration (5, 9). We previously showed that 50% pancreatectomy, that has minimal affects on pancreas endocrine function, induced a delay in hepatocytic DNA synthesis and a decrease in the liver regeneration rate (9). Form these results we speculated that the exocrine pancreas may support liver regeneration due to HGF release.
In the present study, almost complete loss of pancreatic acinar cells, with maintenance of islets 1 week after pancreatic duct ligation was associated with a decrease in the HGF concentration in plasma in the PL+Hx on the first day, but increase on the third and seventh days after hepatectomy. This indicats that the exocrine pancreas contributes to the early rapid rise in plasma HGF after hepatectomy. The increase after the second day may be due to endocrine regulatory responses.
Hepatectomy was here found to result in low blood insulin levels in normal pancreas animals compared with sham group after hepatectomy. This may be due to endocrine regulation for maintenance of euglycemia. In the PL+Hx group, the significant decrease in insulin levels on postoperation day 1, despite maintenance of islet structure and normal glucose tolerance tests (25, 26), is in line with the literature (27, 28).
The present study showed that the regenerating liver DNA synthesis peak to be significantly decreased and hepatocytic mitosis markedly reduced and delayed 24 hours in PL+Hx compared with Hx group. Depression of early rapid rise in HGF is the most likely main cause, with low serum insulin levels of secondary importance. HGF is a distinct and potent mitogen for hepatoctes (13, 14, 29) whereas insulin only enhances hepatocyte proliferation and has no mitogenic effect itself (1, 30).
The LRR for the PL+Hx group was markedly depressed on the postoperation day 1 after hepatectomy followed by significant increase on the third day compared with that in the Hx group. These data conform with changes in HGF and insulin concentrations in the plasma. The reduction of HGF and insulin here would be responsible for the LRR reduction on first day; and the elevation of HGF might induce the increase of LRR on the third day. HGF not only stimulates hepatocyte proliferation but also is an essential factor for hepatocytic maturity and develpoment (31). It enhances hepatocytic protein and lipid syntheses to cause hepatocytic enlargement (32, 33, 34). Insulin is the most important trophic factor for liver.
Laparotomy one week after subtotal pancreatic duct ligation was associated with a reduction in the normal liver weight compared with the sham group, in agreement with the insulin and Alb levels. These may be related to changes in endocrine and exocrine pancreas induced by the subtotal pancreatic duct ligation.
In this study, the hepatocyte mitosis peak was early at 24 hours in both Hx and PL+Hx groups compared with previous studies. This may have been due to primary effects of the first operation but this remains to be confirmed.
Our present results suggest an important role of exocrine pancreas in liver regeneration. In the clinical field, the major hepatectomy is attempted for cases of hepatocellular cancinoma with liver cirrhosis, and advanced biliary malignancy (35, 36), and postoperative liver decompensation due to insufficient liver mass is a major complication. Sakai T reported that liver cirrhosis is usually linked with reduced pancreas exocrine function (37) and our results are very interesting in this context suggesting that the early administration of HGF and insulin might be useful for increasing liver volume after major hepatectomy.
In conclusion, subtotal pancreatic duct ligation model partial hepatectomy is associated with decrease in the peak of DNA synthesis, reduction and delay in hepatocyte mitosis, and depression in early liver weight. Change in HGF blood concentrations related to pancreatic duct ligation and depression of early insulin blood levels appear responsible, suggesting that exocrine pancreas plays a supporting role in early liver regeneration.

The authors are grateful to Dr. Tian X. Tang for assistance with the operative procedures for the experimental animals

1. Michalopoulos GK, Defrances MC. Liver regeneration. Science 1997; 2376: 60-66.
2. Higgins GW, Anderson RM. Experimental pathology of the liver: Restoration of the liver the white rat following partial surgical removal. Arch Pathol 1931; 12: 186-202.
3. Bucher NR, Swaffeld MN. The rate of incorporation of labeled thymdine into the doxyrib-onucleic acid of regenerating rat liver in relation to the amount of liver excised. Cancer Res 1964; 24: 1611-1625.
4. Fausto N, Laird AD, Webber EW. Role of growth factors and cytokine in hepatic regenera-tion. FASEB 1995; 9: 1527-1536.
5. Kyprianidis KG, Mkoniatis MG, Papadimitriou DG, Valsamidou A. Effects of subtotal pa-ncreatectomy on liver regeneration of the rate: the role of hepatic stimulator substance. J Surg Res 1996; 62 (2): 267-272.
6. Caruara JA, Gage AA. Increased uptake of insulin and glucagon by the liver as a signal for liver regeneration. Surgery Gynecology Obstetrics 1980; 150: 391-394.
7. Tan TX, Hashimoto T, Chao LY, Itoh K, Manabe T. Effects of partial pancreatectomy on liver regeneration in rats. J Surg Res 1997; 72: 8-14.
8. Michalopoulos G, Zarnegar R. Hepatocyte growth factor. Hepatology 1992; 15: 149-156.
9. Wolf HK, Zarnegar R, George K. Localization of hepatocyte growth factor in human and rat tissues: An immunohistochemical study. Hepatology 1991; 14: 488-494.
10. Hultguist GT, Jonsson LE. Ligation of the pancreatic duct in rats. Soc Acta Med Upsal 196 5; 70: 82-88.
11. Child CG, Barr D, Halswarde GK. Liver regeneration following portal-caval transposition in dogs. Ann Surg 1953; 136: 600-608.
12. Anna D, Leary JA, Hedly DW, Tattersall MH. Immunohistochemical detection of p-roliferating cells in vivo. J Histochem Cytochem 1987; 35: 571-356.
13. Strain AJ, Ismail T, Tsubouchi H, Arakaki N, Hishidr T, kitamula N, Daikuhara Y. Native and recombinant human hepatocyte growth factors are highly potent promoters of DNA sy-nthesis in both human and rat hepatocytes. J Clin Invest 1991; 87: 1853-1857.
14. Zarnegar R, Michalopoulos G. Purification and biological characterization of human hepa-topoeitin A: a polypepide growth factor for hepatocytes. Cancer Res 1989; 49: 3314-332 0.
15. Lindroos PM, Zarnegar R, Michalopoulos GK. Hepatocyte growth factor (hepatoproeintin A) rapidly increases in plasma before DNA synthesis and liver regeneration stimulated by partial hepatectomy and carbon tetrachlioride administration. Hepatology 1991;13: 743-75 0.
16. Peter SO, Steen B, Preben KG, Kim TS, Steen SP. Influence of epidermal growth factor on liver regeneration after partial hepatectomy in rats. Hepatology 1998; 8: 992-996.
17. Tomiya T, Fujiwara K. Serum transforming growth factor-a as a marker of hepatocellular carcinoma complicating cirrhosis. Cancer 1996; 77: 1056-1059.
18. Tomiya T, Fujiwara K. Liver regeneration in fulminant hepatitis as evaluated by serum tra-nsforming growth factor-a levels. Hepatology 1996; 23: 253-257.
19. Zarnegar R, Defrance MC, Kost DP, Michalopoulos GK. Expression of hepatocyte growth factor mRNA in regenerating rat liver after partial hepatectomy. Biochem Biophys Res Co-mmun 1991; 177: 559-565.
20. Schirmacher P, Geerts A, Pietranglo A, Rogler CE. Hepatocyte growth factor/ hepa topoie-tin A is expressed in fat-storing cell from rat liver but not myofibroblast-like cells derived from fat-storing cells. Hepatology 1992; 15: 5-11.
21. Appsamy R, Tanabe M, Mulase N Zarnegar R. Hepatocyte growth factor, blood clearance, organ uptake, and biliary excretion in normal and partially hepatectomized rats. Lab Invest 1993; 68: 270-275.
22. Liu KX, Kato YK, Narukawa M, Sugiyama Y. Importance of the liver in plasma clearance of hepatocyte growth factor in rats. Am J Phys 1992; 263: G462-469.
23. Liu ML, Mars WM, Zarnegar R, Michalopoulos GK. Uptake and distribution of hepatocyte growth factor in normal and regenerating adult rat liver. Am J Pathology 1994; 14: 129-14 0.
24. Kim TH, Mars WM, Stolz DB, Petersen BE, Michalopoulos GK. Extracellular matrix rem-odeling at the early stage of liver regeneration in the rat. Hepatology 1997; 26: 896-904.
25. Edstrom C. Glucose tolerance of rats at various interval following ligation of pancreatic du-cts. Acta Soc Med Upsal 1971; 76: 39-48.
26. Edstrom C, Falkmer S. Pancreatic morphology and blood glucose level in rats at various in-tervals after duct ligation. Virchows Arch Abt Apath Anat 1968; 354: 139-153.
27. Nishiwaka H, Sakazaki S, Shin K, Umeyama K. Experimental studies on pancreatic endoc-rine function of dogs after pancreatic duct ligation (in Japanese). Nippon Shokakibyo Gakk-ai Zasshi (Jap J Gastroenterology). 1979; 76: 934-941.
28. Kanami Y, Miyazaki I, Sugii M, Mula T. An experimental study on the function of pancre-as endocine cells in acute pancreatitis (in Japanese). Nippon Shokakibyo Gakkai Zasshi (Jap J Gastroenterology). 1981; 78: 56-64.
29. Nakamula T, Nishizawa T, Hagiya M, Seki T, Shimanishi M, Sugimula A, Tashilo K, Shi- mizu S. Molecular cloning and expression of human hepatocyte growth factor. Nature 198 9; 342: 440-443.
30. Junge U, Creutzfeldt W. Hepatotrophic effects of pancreatic and gastrointestinal hormones in the rate in vivo and vitro. Ciba Found Simp 1978; 72: 1157-1160.
31. Schmidt C, Bladt F, Goedecke S, Brinkmann V, Zschiesche W. Scatter factor/hepacyte gro-wth factor is essential for liver development. Nature 1995; 373: 699-703.
32. Okano K, Tsubouchi T, Yamashita Y, Wakabayashi H, tanaka S. Hepatic protein synthesis in the regenerating rat live after hepatopancreatectomy. Surgery Today 1997; 27: 511-517.
33. Kaiboli M, Kwon AH, Oda M, Kitamula N, Okumula T. Hepatocyte growth factor stimula-tes synthesis of lipids and secretion of lipoproteins in rat hepatocytes. Hepatology 1998; 27: 1354-1361.
34. Michalopoulos GK. Liver regeneration: molecular mechanisms of growth control. FASEB 1990; 4: 176-187.
35. Myagawa Y, Makuuchi M, Kawazaki S, Hayashi K, Harada H, Seki H. Out of major hepat-ectomy with pancreatoduodenectomy for advanced biliary malignancies. J world surgery 19 96; 20: 77-80.
36. Chijiwa K, Tanaka M. Carcinoma of gallbladder: Appraisal of surgical resection. Surgery 1 994; 115: 751-756.
37. Sakai T. Pancreatic exocrine function in patients with chronic liver disease. Kurume Medic-al J 1998; 45(2):181-185.

FIG. 1. Changes in liver wet weight at various times after second sham laparotomy and subtotal pancreatic duct ligation. Values are means Ž± SEM. The asterisk represents statistically significa- nt differences from subtotal pancreatic duct ligation (P < 0.05).
FIG. 2. Changes in liver regneration rate at various times after 68% hepatectomy and subtotal pancreatic duct ligation + 68% hepatectomy. Values are mansŽ±SEM. The asterisk represents statistically significant differences from subtotal pancreatic duct ligation + 68% hepatectomy (P < 0.05).
FIG. 3. Bromodeoxyuridine (BrdU) labeling index in liver at various times after second sham laparotomy, subtotal pancreatic duct ligation, 68% hepaterctomy, and subtotal pancreatic duct ligation + 68% hepatectomy. Values are mans Ž± SEM. *Significantly different from subtotal pancreatic duct ligation + 68% hepatectomy (P < 0.01).
FIG. 4. Mitotic index (MI) in normal liver and in liver regeneration at various times after second laparotomy, subtotal pancreatic duct ligation, 68% hepatectomy, and subtotal pancreatic duct ligation + 68% hepatectomy. Values are means Ž± SEM. *Signicantly different from subtotal pancreatic duct ligation + 68% hepatectomy. (P < 0.05).
Table 1. Note. Values are means Ž± SEM. Lapa, sham laparotomy; PL, pancreatic duct subtotal ligation; Hx, 68% hepatectomy; PL+Hx, subtotal pancreatic duct ligation + 68% hepatectomy. The asterisk represents statistically significant differences from the control values. *P < 0.05, **P < 0.05.
Table 2. Note. Values are means Ž± SEM. Lapa, sham laparotomy; PL, subtotal pancreatic duct ligation; Hx, 68% hepatectomy; PL+ Hx, subtotal pancreatic duct ligation + 68% hepatectomy. No statistically significant differences were found.
Table 3. Note. Values are means Ž± SEM. Lapa, sham laparotomy; PL, subtotal pancreatic duct ligation; Hx, 68% hepatectomy; PL+Hx, subtotal pancreatic duct ligation + 68% hepatectomy. No statistically significant differences were found.

Nagoya City University
Medical School