Partial purification and characterization of hepatocyte growth factor from serum of hepatectomized rats

Hepatocyte growth factor in transgenic mice: effects on hepatocyte growth, liver regeneration and gene expression

Attention has recently been focused on hepatocyte growth factor as a major candidate factor in liver regeneration because it is the most potent known mitogen for hepatocytes in vitro. However, hepatocyte growth factor also displays diverse activities in vitro as scatter factor, as an epithelial morphogen, as a pluripotent mitogen and as a growth inhibitor. Consequently, we developed transgenic mice that expressed hepatocyte growth factor under the control of albumin regulatory sequences to examine its in vivo role in hepatocyte growth. Hepatocytes of these mice expressed increased levels of hepatocyte growth factor as an autocrine growth factor. Hepatocyte growth factor was a potent stimulus for liver repair; the livers of hepatocyte growth factor-transgenic mice recovered completely in half the time needed for their normal siblings after partial hepatectomy. This transgenic model also enabled us to study the chronic effects of hepatocyte growth factor expression. During several months of observation, the labeling index of hepatocytes in albumin-hepatocyte growth factor mice was doubled, and liver DNA content was increased compared with that in wild-type mice. To identify intermediate signaling pathways for hepatocyte growth factor that might regulate this increased growth response, we examined transgenic mice for changes in expression of genes that are known to be regulated during liver regeneration. We found that levels of c-myc and c-jun mRNA were increased in the hepatocyte growth factor-transgenic mice. In additional experiments the increased c-myc expression was the consequence of increased transcription rates as seen in nuclear run-on and myc-CAT reporter gene experiments. We conclude that hepatocyte growth factor increases growth and repair processes when expressed for long periods in the liver and that c-myc and c-jun may be important intermediaries in the hepatocyte growth response caused by hepatocyte growth factor.

Stimulation of hepatocyte DNA synthesis by neurotensin

Epidermal growth factor and transforming growth factor alpha stimulated DNA synthesis in primary cultures of adult rat hepatocytes. Neurotensin amplified epidermal growth factor-stimulated or transforming growth factor alpha-stimulated DNA synthesis by three- to eightfold. Neurotensin by itself did not stimulate DNA synthesis. Amplification of DNA synthesis by neurotensin was observed as low as 10(-10) M, and it was increased in a dose-dependent manner with maximal effects at 10(-8) M. These results were obtained when hepatocytes were cultured in Williams' medium E, but not in Leibovitz L-15 medium, suggesting that a minor component(s) in the medium is required for hepatocytes to fully respond to neurotensin. Neurotensin effect on DNA synthesis was observed not only in normal rat hepatocytes but also in partially hepatectomized rat hepatocytes, although its effect was stronger in normal hepatocytes. Amplified DNA synthesis was inhibited by transforming growth factor beta. Secondary mitogens (co-mitogens) such as insulin, vasopressin, or angiotensin II interacted additively with low concentrations of epidermal growth factor as well as with neurotensin. Neurotensin-related peptides such as kinetensin or neuromedin-N, which was released from blood plasma by pepsin digestion, did not have this amplifying effect on DNA synthesis at any concentrations tested. Neurotensin mRNA was found in several organs including brain and intestine, but not liver. These results suggest that neurotensin can be regarded as a new secondary mitogen and that it may be involved in cell proliferation, including regenerating liver as a gastrointestinal hormone and/or a neurotransmitter.

Expression and localization of hepatocyte growth factor in rat submandibular gland

By combination of in situ hybridization and immunohistochemical techniques, the expression of hepatocyte growth factor (HGF) was demonstrated in the submandibular gland of rats. Both the mRNA signal and immunoreactivity for HGF were localized exclusively to the epithelial cells of granular convoluted tubules, whereas they were absent from the other components of the submandibular gland. In the granular convoluted tubule cells, HGF-immunoreactivity was localized to the apical secretory granules, which was further substantiated by immunoelectron microscopy. These results added HGF to the list of many growth factors that are produced in the rat submandibular gland and secreted into the saliva.

c-met mRNA overexpression in human hepatocellular carcinoma

This study was aimed at assessing the presence of c-met overexpression in human hepatocellular carcinoma and at determining whether this feature is associated with a definite clinical or pathological characteristic. Expression of c-met was determined by Northern-blot hybridization of a specific probe (human met proto-oncogene) in 18 tumoral and nontumoral liver samples obtained in 18 cirrhotic patients with hepatocellular carcinoma submitted to surgical treatment. Eight of the 18 hepatocellular carcinomas exhibited c-met overexpression, with an increase ranging between 2-fold and 10-fold when compared by densitometry with the surrounding liver. By contrast, in the remaining 10 cases c-met expression was almost identical to that of the surrounding nontumoral liver tissue. Overexpression of c-met was not related to either the age, sex, etiology or functional status of the underlying liver disease, or to the size of the tumor, to its differentiation degree or to the presence of pseudocapsule invasion and existence of additional neoplastic nodules. These data indicate that almost half of the human hepatocellular carcinomas exhibit c-met overexpression. Nevertheless, the biological relevance of this characteristic is not known.

Effects of mitogens and co-mitogens on the formation of small-cell colonies in primary cultures of rat hepatocytes

Colonies of small hepatocytes appeared after the culture of primary adult rat hepatocytes for 4 days in serum-free Dulbecco's modified Eagle's medium containing 10 mM nicotinamide and 10 ng/ml of epidermal growth factor (EGF), acidic and basic fibroblast growth factors (FGF), hepatocyte growth factor (HGF), or transforming growth factor-alpha (TGF-alpha). Every colony consisted of cells that each had a single nucleus and a higher nucleus/cytoplasm ratio than surrounding hepatocytes, and immunocytochemically the cells induced by any mitogen were stained with albumin, transferrin, cytokeratin-8 and -18. But these cells expressed neither cytokeratin-7 nor -19. When 6 x 10(5) cells were plated on 35-mm dishes, about 15 colonies per 1,000 attached cells were observed in the cultures treated with EGF, HGF, and TGF-alpha. Although FGFs could also induce colonies, their number was less than half of the number induced by EGF. Furthermore, the numbers of colonies induced by the combinations of EGF+HGF, EGF+TGF-alpha, and HGF+TGF-alpha were not different from those of the colonies induced by each mitogen alone. To examine the ability of co-mitogenic factors to induce small-cell colonies, angiotensin-II, insulin-like growth factor-I, norepinephrine, tumor necrosis factor, and vasopressin were used. In the cells cultured without EGF, these co-mitogens neither stimulated DNA synthesis nor induced colonies. On the other hand, in cells cultured with both EGF and each co-mitogen, although the DNA synthesis of the hepatocytes was enhanced, the number of colonies detected was not significantly different from the number which EGF alone could induce. These results showed that the small-cell colonies in primary cultures of rat hepatocytes were inducible by EGF, HGF, TGF-alpha, or FGFs and that the co-mitogens did not influence the formation of the small-cell colonies.

Immunohistochemical detection of hepatocyte growth factor/scatter factor in human cancerous and inflammatory lesions of various organs

Hepatocyte growth factor (HGF)/scatter factor (SF) is a multifunctional factor considered to be potentially involved in tissue regeneration, wound healing, embryogenesis, angiogenesis and cancer invasion. Here we examined immunohistochemically the distribution of HGF/SF in human tissues, including cancerous and inflammatory tissues, using anti-HGF antibody. HGF/SF accumulation was clearly detected in the extracellular matrix, particularly along the basement membrane, in cancerous and inflammatory tissues, but only a little was detected in normal tissues. HGF/SF is well known to have a strong affinity for heparin in vitro, and from the results of our immunohistochemical assay, we considered that HGF/SF was bound to heparin or heparan sulfate of the extracellular matrix and basement membrane. HGF/SF was well localized in cancerous and inflammatory lesions of human lung, liver and pancreas, and in apparently normal tissues of kidney, adrenal gland and pancreas obtained at autopsy. In lung, HGF/SF was localized along the basement membranes of cancer cell nests, in the extracellular matrix of the cancer cell surface, cancer stroma and tissues invaded by cancer, and the basement membranes of bronchial epithelium and capillary vessels in inflammatory stroma. Since HGF/SF makes some cancer cells more invasive in vitro, the accumulation of HGF/SF in cancerous tissue suggests that the invasiveness of some cancer cells may be increased by HGF/SF in vivo.

Intrahepatic distant metastasis and metachronous multicentric occurrence in solitary hepatocellular carcinoma of less than five centimeters in diameter

During the 7 years from 1984 to 1990, 36 patients underwent liver resection for solitary hepatocellular carcinoma (HCC) measuring less than 5 cm in diameter, with no intrahepatic vascular invasion on imaging diagnoses and no macroscopic infiltration into the tumor capsule or surrounding tissues. Although HCC is less likely to cause intrahepatic adjacent metastasis to the cut liver surface, an analysis revealed the possibility of intrahepatic distant metastasis and metachronous multicentric occurrences, even after complete removal of the primary tumor. The 5-year cumulative survival rate was 53%, while the 5-year cumulative recurrence-free survival rate was 19%. Of the 36 patients, 18 (50%) had suffered a recurrence by April, 1992, one with extrahepatic metastasis. Recurrence of intrahepatic metastasis was multifocal in 5 patients, single and adjacent in 1, and single (or a few) and distant in 11. Multifocal recurrence was observed within 1 year after liver resection. The sole single and adjacent metastatic case occurred in one of eight patients in the recurrent group in whom distance of the surgical margin was less than 1 cm [TW(+)]. Multicentric occurrence was found in 6 of 13 patients (46%) whose recurrent tumors were examined histologically, and all belonged to the "single (or a few) and distant" type of recurrence. In this report, we also present two typical cases of metastasis, one being multifocal metastasis occurring within 3 months after liver resection and the other being intrahepatic metastasis occurring after a 4-year-dormant state, to demonstrate the complicated nature of the intrahepatic metastatic pattern.

Hepatocyte growth factor/scatter factor, liver regeneration and cancer metastasis

Hepatocyte growth factor (HGF) is the most potent stimulator of hepatocyte growth and DNA synthesis identified; it is now known to be the same molecule as scatter factor, which increases the motility of a variety of cell types. HGF is becoming recognized as one of the most important factors in the regulation of liver regeneration after surgical resection or chemical damage. HGF is produced by several tissues, including neoplasms; it can therefore provide a stimulus for increased motility of malignant cells by both a paracrine and autocrine mechanism. The receptor for HGF has been identified as the product of the oncogene c-met, raising the possibility that this gene plays a key role in facilitating cellular invasion. HGF may therefore be important not only for liver cell growth but also in metastasis. This article summarizes the current position of research on HGF, and presents both clinical and scientific evidence that strongly implicates this factor in liver regeneration and cancer invasion and metastasis.

Hepatocyte growth factor-like protein is identical to macrophage stimulating protein

Although the hepatocyte growth factor-like protein (HLP) shares a 50% homology with the hepatocyte growth factor, the biological function of HLP has remained unknown. Addition of conditioned medium of COS-7 cells transfected with the expression plasmid for HLP cDNA to cultures of resident peritoneal macrophages induced specific activation of macrophages, and the factor which stimulates macrophages was purified from the conditioned medium. The purified protein showed M(r) of 85 kDa on SDS-PAGE, and this M(r) is in agreement with that of macrophage-stimulating protein (MSP) previously purified from human serum, as well as with the predicted M(r) of HLP. Amino acid composition of the purified protein coincided with the compositions of human HLP and MSP. Together with the finding that the partial amino acid sequences of MSP are highly homologous to that of HLP, we conclude that the biological function of HLP is to activate macrophages and that HLP and MSP are identical molecules.

K252a: a new blocker of the cell-cycle at G1 phase in a human hepatoma cell line

The administration of 200 nM K252a to HuH7 suppressed the proliferation of the cells almost completely. The uptake of [3H]thymidine was inhibited, and flow cytometry revealed only one peak at 2C on day 3 after treatment with 100 nM K252a. The expression of proto-oncogene c-myc was not reduced. Despite the blockage at G1, both the size of the cells and the amount of cell protein had increased by 4 times by day 3 after treatment with K252a, while the cells secreted albumin and alpha-fetoprotein into the medium as usual. These results show that K252a can increase the cell size of HuH7 without losing its function by blocking the cell cycle at G1 phase.

Structure, genetic mapping, and expression of the mouse Hgf/scatter factor gene

The cytokine termed hepatocyte growth factor or scatter factor (HGF/SF) has been implicated in embryonic development and liver regeneration. Mouse HGF/SF cDNA clones were obtained by screening a mouse liver cDNA library with synthetic oligonucleotides, as well as by using the method of reverse transcription-polymerase chain reaction (RT-PCR) with mRNA isolated from adult mouse brain. Sequence analysis of mouse HGF/SF cDNA clones spanning the entire coding region revealed an overall amino acid identity of 90% and 98% with human and rat HGF, respectively. Using mouse HGF antisense RNA as a probe, we observed the expression of HGF/SF mRNA in the liver, kidney, whole brain and cerebellum of adult mice as well as in day 9 and day 10 mouse embryos. The mouse HGF/SF cDNA was utilized for mapping Hgf to the centromeric region of mouse Chromosome 5 in apparent close proximity to the reeler mutation by the analysis of two multilocus crosses. The relationship of Hgf to the reeler mutation is discussed. The availability of mouse HGF/SF cDNA clones should facilitate further analysis of HGF/SF function during development.

A possible prototype of multifocal recurrence after liver resection of hepatocellular carcinoma: report of a case

The clinical course of a 68-year-old male with HCC is herein reported. In addition to two tumor nodules detected preoperatively, the resected surgical specimen disclosed macroscopically invisible tumor-cell clusters as well as intrahepatic metastatic foci. These clusters had no clear border and were more basophilic with small-sized cells, high cellularity, and a higher nuclear/cytoplasm ratio, which was suggestive of early HCC with a form of de novo occurrence. The postoperative course, characterized by an early recurrence of small nodules in the remnant liver suggested the presence of a similar invisible tumor mixture at the time of operation. This case can thus be regarded as a prototype of the multifocal recurrence type.

Iron-saturated lactoferrin as a co-mitogenic substance for neonatal rat hepatocytes in primary culture

We studied the effect of lactoferrin on DNA synthesis in neonatal rat hepatocytes in primary culture to determine if this agent acts as a mitogen in human milk. Thymidine incorporation into the DNA of cultured hepatocytes stimulated by lactoferrin in the presence of insulin and human epidermal growth factor was examined. Iron-saturated lactoferrin increased DNA synthesis of neonatal hepatocytes by 1.5 times and this potency was the same as that of insulin. It significantly enhanced the stimulatory effect of human epidermal growth factor plus insulin; DNA synthesis under these conditions was seven times that of control. Iron-free lactoferrin did not affect DNA synthesis, nor did the exogenous addition of ferric ions. The enhancement of DNA synthesis by iron-saturated lactoferrin was significant for neonatal hepatocytes, but not for adult hepatocytes. These results suggest that iron-saturated lactoferrin, which itself had low mitogenic activity, is a co-mitogenic substance for neonatal hepatocytes in vitro.

Expression of the hepatocyte growth factor receptor in the regenerating rat liver

The c-met oncogene product is a cell-surface receptor, which ligand is believed to be the hepatocyte growth factor. We studied the expression of c-met oncogene in the regenerating rat liver after either partial hepatectomy or CCl4-induced liver injury. Northern blot analysis showed that after partial hepatectomy the transcripts of c-met decreased at 8 h, reached the minimum at 36 h, and returned to the steady level on the seventh day. In contrast with the hepatectomized liver, the transcripts of c-met increased after CCl4 treatment. These observations suggest that c-met transcription may be regulated differently depending on regeneration signals.

Expression of hepatocyte growth factor in growing and regenerating rat skeletal muscle

The expression of hepatocyte growth factor (HGF) was studied in rat skeletal muscle during postnatal growth and during regeneration after ischemic injury. By Northern blot analysis two RNA transcripts with estimated sizes of 6.0 and 3.1 kb, respectively, could be detected in immature skeletal muscle at 2, 4, and 10 days after birth, whereas no signal could be detected in adult skeletal muscle. In regenerating muscle, HGF mRNA was reexpressed, and a distinct signal was evident during the first days after the injury. One week after the injury only a weak signal was obtained. By in situ hybridization staining, HGF mRNA could be demonstrated in the immature muscle tissue of newborn rats but not in adult skeletal muscle. In regenerating muscle, the staining for HGF mRNA could be demonstrated in the regenerating muscle cells during the early phase of the regeneration. One week after the injury, only faint staining for HGF mRNA persisted in the regenerated fibers. It is concluded that HGF expression is developmentally regulated in skeletal muscle and that HGF is reexpressed in a transient manner during postischemic muscle regeneration.

Regulation of spreading and growth of colon cancer cells by hepatocyte growth factor

Hepatocyte growth factor (HGF), also known as scatter factor, regulates both cell motility and the growth of some cell types. We have determined the effects of HGF on the motility and growth of human colon cancer cell lines (HT115, HT29, HRT18 and HT55). Cell motility, as measured by dissociation from carrier beads or by scattering of cell colonies, was greatly increased in all cell lines. The effects were completely blocked by anti-HGF antibody. In contrast, cell growth of HT115, HT29 and HRT18 cells was inhibited by a wide range of concentrations of HGF. HT55 cell growth was also inhibited but needed a prolonged culture period (> 5 days). The HGF receptor/Met protein is highly expressed in the membrane fraction of these cells as determined by Western blotting. It is concluded that HGF has an effect on both colon cancer cell motility and growth, which may be important in the control of the spread of colon cancer.

Stimulation of DNA synthesis and protooncogene expression in primary rat hepatocytes in long-term DMSO culture

We have previously described the use of a chemically defined medium (CDM) supplemented with epidermal growth factor (EGF) and dimethylsulfoxide (DMSO) to maintain long-term cultures of rat hepatocytes in a highly differentiated state. In this study, conditions necessary to stimulate high levels of DNA synthesis in hepatocytes in long-term DMSO culture were defined. Hepatocytes were maintained in culture for 20 days in CDM containing DMSO and EGF, insulin, and glucagon. EGF, insulin, and glucagon were then removed for 7 days. Readdition of EGF, insulin, and glucagon at day 27 (shiftup) was accompanied by a three- to sixfold increase in labeling index. If DMSO or dexamethasone (dex) + DMSO were removed at time of shiftup, the labeling index increased by 18- to 54-fold. TGF beta inhibited DNA synthesis stimulated by EGF shiftup, TGF alpha shiftup, or EGF shiftup in combination with removal of dex + DMSO. Stimulation of DNA synthesis was accompanied by a specific, sequential induction of protooncogene mRNA levels; c-fos mRNA was induced 23-fold at 0.5 h after readdition of EGF; c-myc mRNA was induced three- to four-fold by 0.5 h; TGF alpha mRNA was induced sevenfold by 8 h; K-ras mRNA was induced fourfold by 26 h. Changes in protooncogene expression paralleled changes seen in regenerating liver. When DMSO was removed for greater than 48 h, the cells flattened and spread out, chords of cells were no longer well defined, albumin mRNA levels decreased, and fibronectin, beta 1 integrin, and TGF beta transcripts increased.

Immunohistochemical localization of hepatocyte growth factor protein in pancreas islet A-cells of man and rats

Hepatocyte growth factor (HGF), a potent mitogen for adult rat hepatocytes in primary culture, has previously been shown to be primarily expressed in the nonparenchymal cells of the liver. Using polyclonal antisera against human and rat HGFs we studied the tissue distribution of HGF immunohistochemically and found the most intense staining in the pancreas islet cells in both man (autopsy cases) and the rat. Differential localization of 4 pancreas islet hormones, glucagon, insulin, somatostatin and pancreatic polypeptide, revealed HGF to be preferentially expressed within the glucagon‐positive cells. The results indicate that HGF is primarily produced or stored in A‐cells and may act as a growth factor in a paracrine and an endocrine fashion, like various other hormones.

Apoptosis of serum-free C2.8 mouse embryo hepatocytic cells caused by hepatocyte growth factor deprivation

C2.8 mouse embryo hepatocytic cells, acutely required exogenous hepatocyte growth factor (HGF) to survive and proliferate in serum-free Dulbecco's modified Eagle's medium supplemented with insulin, transferrin and Na-selenite. Greater than 90% of cultured C2.8 cells died within 48 hours from plating in the absence of HGF. Conversely, HGF prolonged maintenance of life and stimulated cell proliferation. Removal of HGF from the medium of cultures that had grown to confluency, also resulted in a rapid decreased cell survival. In the last circumstance, light microscopic observations revealed, with high frequency, morphological features characteristic of apoptosis. DNA within the affected cells underwent rapid fragmentation, revealed as a ladder of DNA fragments in multiples of about 200 base pairs. HGF prevented loss of cell viability, morphological damages and retarded DNA fragmentation in confluent C2.8 cells. Cycloheximide delayed cell death caused by HGF deprivation.

Hepatocyte growth factor stimulates phosphoinositide hydrolysis and mitogenesis in cultured renal epithelial cells

Hepatocyte growth factor (HGF), a novel heparin-binding peptide growth factor of MW 97-kDa, is a potent mitogen for parenchymal hepatocytes. HGF is present in normal serum and increases following liver injury or partial hepatectomy. In addition to liver, HGF mRNA has been detected in kidney. In cultured rabbit proximal tubule cells, recombinant human HGF (10(-10) M) increased DNA synthesis, measured as [3H] thymidine incorporation, from 1345 +/- 213 to 2931 +/- 636 cpm/10(6) cells; n = 9; p < 0.005). HGF was found to exert mitogenic effects at lower concentrations than epidermal growth factor (EGF), with half maximal effects seen at 6 x 10(-11) M compared to 7 x 10(-10) M for EGF. HGF was additive with EGF in stimulating [3H] thymidine incorporation. In addition to rabbit proximal tubule cells, HGF increased proliferation in a cultured mouse proximal tubule cell line, MCT, and in rat glomerular epithelial cells. In contrast, HGF did not stimulate proliferation of either rat mesangial cells or a rat aortic smooth muscle cell line, A7r5. The HGF receptor is the product of the c-met proto-oncogene. C-met mRNA was detected in total kidney and in cultured proximal tubule cells but was not detected in cultured mesangial cells. In contrast, HGF mRNA was detected in mesangial cells but not in cultured proximal tubule cells. Preincubation of rabbit proximal tubule cells with the tyrosine kinase inhibitor, genistein (50 microM), prevented HGF-stimulation of [3H] thymidine incorporation. In LiCl pretreated rabbit proximal tubule cells loaded with [3H] myoinositol, HGF increased total inositol phosphate release, measured by anion exchange chromatography (control: 2181 +/- 414 vs HGF: 2609 +/- 478 cpm/10(6) cells; n = 6; p < 0.05). Although genistein did not affect baseline phosphoinositide hydrolysis, it inhibited the HGF stimulation. Thus, HGF is mitogenic for cultured proximal tubule cells as well as glomerular epithelial cells. Inhibition of proliferation and PI turnover by genistein suggests that HGF's actions are mediated in part by tyrosine kinase activity. In mammalian kidney, HGF released from mesangial cells may serve as a paracrine activator of the adjacent epithelial cells.

Proteolytic activation of a single-chain precursor of hepatocyte growth factor by extracellular serine-protease

Hepatocyte growth factor (HGF) is biosynthesized as a single-chain precursor (pro-HGF) and is proteolytically processed to a two-chain mature form. When MRC-5 fibroblasts were pulse-radiolabeled under serum-free conditions, pro-HGF was the predominant molecular form of HGF in the culture medium. CHO cells transfected with an expression plasmid containing a full-size human HGF cDNA produced pro-HGF when these cells were cultured in serum-free medium. These findings suggest that HGF is secreted as a pro-form, which is then converted to a two-chain form by extracellular protease. Single-chain HGF exhibited mitogenic activity on cultured hepatocytes, with a potency similar to that of mature HGF, but this activity was remarkably inhibited by leupeptin. We postulate that inactive pro-HGF is converted to an active two-chain form by a leupeptin-sensitive serine-protease expressed by hepatocytes. Neither plasminogen activators nor plasmin showed any processing activity of pro-HGF in vitro.

Human lung cancer cell line producing hepatocyte growth factor/scatter factor

Hepatocyte growth factor (HGF)/scatter factor (SF) is a cytokine which is produced by mesenchymal cells and stimulates the motility of some epithelial cells, including cancer cells and vascular endothelial cells. Two human lung cancer cell lines, PC-1 and PC-13, were found to produce a protein which was indistinguishable from HGF/SF with regard to biological activities and immunological characteristics, although they were derived from epithelial cells. In general, highly aggressive cancer cells often show some mesenchymal characteristics, and production of HGF/SF by cancer cells is also considered as a phenomenon of acquisition of mesenchymal phenotype, which may be involved in cancer invasion and progression. These cell lines showed no apparent response to exogenous HGF/SF. In addition, no c-met proto-oncogene product was detectable in these cells by Western blot analysis. Although the function of HGF/SF produced by cancer cells, either autocrine or paracrine stimulation, remains to be studied, this is the first report to describe cancer cells producing HGF/SF.

Rapid and marked induction of hepatocyte growth factor during liver regeneration after ischemic or crush injury

Liver injuries induced by ischemia or physical trauma are characterized by noninflammatory damage frequently observed in a clinical setting. When the liver of rats was injured by ischemic treatment or physical crushing, necrotic tissue degeneration occurred in several sites of lobulus within 24 hr. Hepatocyte growth factor, a potent mitogen for adult rat hepatocytes in primary culture, was markedly induced in the livers of rats injured by ischemia or physical trauma. In both cases, the hepatocyte growth factor messenger RNA level in the injured liver reached about 10 to 20 times that of the normal level during 12 to 24 hr after liver injury. The increase in hepatocyte growth factor messenger RNA correlated well with the degree of liver damage as evaluated by serum ALT activity in the sera of rats. In situ hybridization showed that hepatocyte growth factor messenger RNA expression occurs in nonparenchymal liver cells, primarily in Kupffer cells of the ischemic liver. After the increase of hepatocyte growth factor messenger RNA in the injured liver, a marked compensatory hepatocyte DNA synthesis occurred 48 to 72 hr after these treatments. These results suggest that hepatocyte growth factor acts as a hepatotropic factor for liver regeneration after noninflammatory liver damage caused by ischemia and physical crush, probably through a paracrine mechanism.

Extracellular proteolytic cleavage by urokinase is required for activation of hepatocyte growth factor/scatter factor

The extracellular protease urokinase is known to be crucially involved in morphogenesis, tissue repair and tumor invasion by mediating matrix degradation and cell migration. Hepatocyte growth factor/scatter factor (HGF/SF) is a secretory product of stromal fibroblasts, sharing structural motifs with enzymes of the blood clotting cascade, including a zymogen cleavage site. HGF/SF promotes motility, invasion and growth of epithelial and endothelial cells. Here we show that HGF/SF is secreted as a single-chain biologically inactive precursor (pro-HGF/SF), mostly found in a matrix-associated form. Maturation of the precursor into the active alpha beta heterodimer takes place in the extracellular environment and results from a serum-dependent proteolytic cleavage. In vitro, pro-HGF/SF was cleaved at a single site by nanomolar concentrations of pure urokinase, generating the active mature HGF/SF heterodimer. This cleavage was prevented by specific urokinase inhibitors, such as plasminogen activator inhibitor type-1 and protease nexin-1, and by antibodies directed against the urokinase catalytic domain. Addition of these inhibitors to HGF/SF responsive cells prevented activation of the HGF/SF precursor. These data show that urokinase acts as a pro-HGF/SF convertase, and suggest that some of the growth and invasive cellular responses mediated by this enzyme may involve activation of HGF/SF.

Tumorigenicity of the met proto-oncogene and the gene for hepatocyte growth factor

The met proto-oncogene is the tyrosine kinase growth factor receptor for hepatocyte growth factor/scatter factor (HGF/SF). It was previously shown that, like the oncogenic tpr-met, the mouse met proto-oncogene transforms NIH 3T3 cells. We have established NIH 3T3 cells stably expressing both human (Methu) and mouse (Metmu) met proto-oncogene products. The protein products are properly processed and appear on the cell surface. NIH 3T3 cells express endogenous mouse HGF/SF mRNA, suggesting an autocrine activation mechanism for transformation by Metmu. However, the tumor-forming activity of Methu in NIH 3T3 cells is very low compared with that of Metmu, but efficient tumorigenesis occurs when Methu and HGF/SFhu are coexpressed. These results are consistent with an autocrine transformation mechanism and suggest further that the endogenous murine factor inefficiently activates the tumorigenic potential of Methu. The tumorigenicity observed with reciprocal chimeric human and mouse receptors that exchange external ligand-binding domains supports this conclusion. We also show that HGF/SFhu expressed in NIH 3T3 cells produces tumors in nude mice.

Developmental changes in hepatocyte growth factor mRNA and its receptor in rat liver, kidney and lung

Hepatocyte growth factor (HGF) is a mesenchymal-derived factor which induces mitosis, cell movement and morphogenesis of tissue-like structure. We analyzed changes in HGF mRNA and its receptor, the c-met proto-oncogene product, in the liver, kidney and lung during late fetal and postnatal development in rats. In the liver, the HGF-mRNA level was very low during late gestation and in neonates, it increased remarkably and reached a maximum two weeks postnatally, to be followed by a decrease to 33% of the maximum. HGF mRNA in the kidney and lung was either undetectable or very low during late gestation and the neonatal period and increased markedly to reach a maximum, respectively, 3-4 weeks postnatally. HGF-mRNA level in the adult rat lung was fivefold higher than that in the liver and kidney. The number of HGF receptors on plasma membranes of these tissues was low in neonates but there was a rapid increase after birth and a maximum was reached within three weeks. The number of HGF receptors/ng plasma membrane protein at the maximal level was highest in the liver and lowest in the lung. c-met/HGF-receptor mRNA in the liver was also low during late-gestation or in early neonatal periods and increased postnatally. Since HGF-mRNA and HGF-receptor levels changed differently in liver, kidney and lung, the expression of HGF and its receptor may be independently regulated in each organ. However, in these organs, HGF mRNA and the HGF receptor increased within a few weeks of birth, HGF may play roles in organ growth, organ maturation and the maintenance of tissue homeostasis during the postnatal period, presumably through its potential to act as mitogen, motogen and morphogen.

Differential expression of hepatocyte growth factor, transforming growth factor-alpha and transforming growth factor-beta 1 messenger RNAs in two experimental models of liver cell proliferation

Hepatocyte growth factor, a potent hepatocyte mitogen in vitro, appears to trigger hepatocyte regeneration after partial hepatectomy and after acute liver cell necrosis. Transforming growth factor-alpha and transforming growth factor-beta 1 may also be involved in the control of liver regeneration. In this study we assessed possible roles of hepatocyte growth factor, transforming growth factor-alpha and transforming growth factor-beta 1 on liver cell proliferation in vivo, using a model of choline deficiency that is associated with liver cell necrosis and a model of a hypolipidemic agent (4-chloro-6-(2,3 xylidino)-2-pyrimidinylthio (N-beta-hydroxyethyl) acetamide) without liver necrosis. Male F344 rats were fed a choline-deficient diet or 0.16% 4-chloro-6-(2,3 xylidino)-2-pyrimidinylthio (N-beta-hydroxyethyl) acetamide diet for 6 and 4 wk, respectively. Rats were killed periodically, and the expression of hepatocyte growth factor messenger RNA in the liver, lung and kidney was determined by Northern-blot analysis. The levels of transforming growth factor-alpha and transforming growth factor-beta 1 messenger RNAs in the liver were also determined. Feeding a choline-deficient diet for 1 to 6 wk led to gradual increases in the levels of hepatocyte growth factor, transforming growth factor-alpha and transforming growth factor-beta 1 messenger RNAs in the liver. Feeding a 4-chloro-6-(2,3 xylidino)-2-pyrimidinylthio (N-beta-hydroxyethyl) acetamide diet for 3 days and 2 wk induced marked enhancement of liver cell proliferation as judged by hepatocyte 5-bromo-2-deoxyuridine incorporation.(ABSTRACT TRUNCATED AT 250 WORDS)

Serum hepatocyte growth factor levels in hepatectomized and nonhepatectomized surgical patients

Serum hepatocyte growth factor levels were measured in hepatectomized and nonhepatectomized surgical patients. The levels were significantly increased and reached a maximum within 7 days after surgery in both groups, returning to preoperative levels 28 days after partial hepatectomy and 7 days after other operations. Multiple regression analysis showed that such maximal hepatocyte growth factor levels were significantly related to having liver cirrhosis and postoperative maximal serum total bilirubin and alanine aminotransferase levels and peripheral white blood cell counts in the hepatectomized group and to postoperative maximal peripheral white blood cell counts and serum C-reactive protein levels in the nonhepatectomized group. However, the levels showed no relation to the resected liver volume and increment of the remaining liver volume 28 days after partial hepatectomy. It is concluded that serum hepatocyte growth factor levels were increased after partial hepatectomy in association with hepatocellular dysfunction and necrosis and systemic inflammation. It is unlikely that the increase was related to liver regeneration.

Tumorigenicity of the met proto-oncogene and the gene for hepatocyte growth factor

The met proto-oncogene is the tyrosine kinase growth factor receptor for hepatocyte growth factor/scatter factor (HGF/SF). It was previously shown that, like the oncogenic tpr-met, the mouse met proto-oncogene transforms NIH 3T3 cells. We have established NIH 3T3 cells stably expressing both human (Methu) and mouse (Metmu) met proto-oncogene products. The protein products are properly processed and appear on the cell surface. NIH 3T3 cells express endogenous mouse HGF/SF mRNA, suggesting an autocrine activation mechanism for transformation by Metmu. However, the tumor-forming activity of Methu in NIH 3T3 cells is very low compared with that of Metmu, but efficient tumorigenesis occurs when Methu and HGF/SFhu are coexpressed. These results are consistent with an autocrine transformation mechanism and suggest further that the endogenous murine factor inefficiently activates the tumorigenic potential of Methu. The tumorigenicity observed with reciprocal chimeric human and mouse receptors that exchange external ligand-binding domains supports this conclusion. We also show that HGF/SFhu expressed in NIH 3T3 cells produces tumors in nude mice.

Hepatocyte growth factor is a potent angiogenic factor which stimulates endothelial cell motility and growth

Hepatocyte Growth Factor (HGF, also known as Scatter Factor) is a powerful mitogen or motility factor in different cells, acting through the tyrosine kinase receptor encoded by the MET protooncogene. Endothelial cells express the MET gene and expose at the cell surface the mature protein (p190MET) made of a 50 kD (alpha) subunit disulfide linked to a 145-kD (beta) subunit. HGF binding to endothelial cells identifies two sites with different affinities. The higher affinity binding site (Kd = 0.35 nM) corresponds to the p190MET receptor. Sub-nanomolar concentrations of HGF, but not of a recombinant inactive precursor, stimulate the receptor kinase activity, cell proliferation and motility. HGF induces repairs of a wound in endothelial cell monolayer. HGF stimulates the scatter of endothelial cells grown on three-dimensional collagen gels, inducing an elongated phenotype. In the rabbit cornea, highly purified HGF promotes neovascularization at sub-nanomolar concentrations. HGF lacks activities related to hemostasis-thrombosis, inflammation and endothelial cells accessory functions. These data show that HGF is an in vivo potent angiogenic factor and in vitro induces endothelial cells to proliferate and migrate.

Up-regulation of hepatocyte growth factor gene expression by interleukin-1 in human skin fibroblasts

Hepatocyte growth factor (HGF) functions as a hepatotrophic and renotrophic factor for regeneration of the liver and kidney. When 1 ng/ml of interleukin-1 alpha (IL-1 alpha) or interleukin-1 beta (IL-1 beta) was added to cultures of human skin fibroblasts, the production of HGF was 5-6 fold higher than levels in the controls. HGF mRNA level in the cells was increased to 4-fold higher levels at 6 h after exposure to IL-1 alpha. Tumor necrosis factor-alpha and interferon-gamma but no other cytokine tested had slightly stimulatory effects on HGF production. The tumor promoter, tetradecanoylphorbol 13-acetate (TPA) markedly enhanced the stimulatory effect of IL-1 alpha and IL-1 beta on the production of HGF. The stimulatory effect of both IL-1 alpha and IL-1 beta and the synergistical stimulation with TPA were completely abrogated by 10 ng/ml TGF-beta 1 or 1 microM dexamethasone. These results suggest that IL-1 alpha and IL-1 beta are positive regulators for expression of the HGF gene and are likely have a role in regeneration of tissues following the occurrence of inflammatory diseases.

Differential proliferative response of cultured fetal and regenerating hepatocytes to growth factors and hormones

Upon epidermal growth factor (EGF) stimulation, fetal (20 days of gestation) and regenerating (44-48 h after partial hepatectomy) rat hepatocytes, isolated and cultured under identical conditions, increased DNA synthesis and entered into S-phase and mitosis, measured as [3H]thymidine incorporation and DNA content per nucleus in a flow cytometer, respectively. Fetal hepatocytes consisted of a homogeneous population of diploid (2C) cells. Two different populations of cells were present in regenerating liver, diploid (2C) and tetraploid (4C) cells, that responded to EGF. Glucagon or norepinephrine did not affect EGF stimulation of DNA synthesis in fetal liver cells, but they potentiated EGF response in regenerating hepatocyte cultures. Glucocorticoid hormones (dexamethasone) inhibited DNA synthesis in fetal hepatocyte cultures, an effect potentiated by the presence of glucagon or norepinephrine. In contrast, in regenerating hepatocytes, dexamethasone increased EGF-induced proliferation. EGF-dependent DNA synthesis was inhibited by TGF-beta in both fetal and regenerating cultured hepatocytes. TGF-beta action was partially suppressed by norepinephrine in regenerating hepatocytes, but was without effect in fetal hepatocyte cultures, whereas a synergistic action between TGF-beta and dexamethasone inhibiting growth in fetal but not in regenerating hepatocytes was found. Taken together, these results may suggest that there are significant differences between fetal and regenerating hepatocyte growth in their response to various hormones.

Human intrahepatic biliary epithelial cells proliferate in vitro in response to human hepatocyte growth factor

In previous studies, intrahepatic human biliary epithelial cells (BEC) were isolated in high purity. However, these cells demonstrated only limited growth responses. Here we report that human BEC proliferate in response to human hepatocyte growth factor (hHGF), retain BEC-specific phenotype, and can be serially passaged. BEC showed dose-dependent growth in response to 0.01-100 ng/ml hHGF. The maximum S-phase labeling index reached 40% with half-maximal stimulation at 1 ng/ml. The response of cells from normal and primary biliary cirrhotic liver to hHGF was similar. Cultures were immunostained with specific antibodies and then processed for [3H]thymidine autoradiography. Proliferating cells expressed BEC-specific markers (HEA125 and CK-19), but were negative for desmin and factor VIII-related antigen. Occasional vimentin-positive cells were observed, but these were nonproliferative. In conclusion, cells responding to hHGF were clearly BEC in origin. The observation that HGF is mitogenic for BEC as well as hepatocytes has important implications. First, greater yields of intrahepatic BEC are available for subsequent studies of the pathogenesis and etiology of diseases of the biliary epithelium. Secondly, some means of regulating the cellular response to HGF in vivo must operate, in that HGF levels rise early after partial hepatectomy and yet BEC proliferate 24 h later than hepatocytes.

Regulation of cell growth and motility by hepatocyte growth factor and receptor expression in various cell species

Hepatocyte growth factor (HGF), a humoral mediator for regeneration of liver and kidney, possesses multiple biological activities. To investigate target cell specificity and to examine whether multiple actions of HGF are related to properties of the HGF receptor on target cells, we examined the effects of HGF on cell growth and motility and analyzed the HGF receptor in various species of cells. HGF stimulated growth and DNA synthesis of PAM212 (naturally immortalized mouse keratinocytes), Mv1Lu (mink lung epithelia), and A431 (human epidermoid carcinoma) cells, as well as mature hepatocytes, but inhibited those of IM-9 (human B-lymphoblasts). Conversely, HGF had a marked stimulatory effect on cell motility of MDCK (Mardin-Darby canine kidney epithelia) cells, but not on their growth. Also, HGF enhanced the motility of various species of cells, including A431, PAM212, HepG2 (human hepatoma), KB (human epidermoid carcinoma), and J-111 (human monocytes) cells. Scatchard analysis of 125I-HGF binding to hepatocytes indicated that the cells expressed both high- and low-affinity binding sites for HGF with Kd values of 23 and 260 pM, respectively. High-affinity HGF receptor with Kd values of 20-25 pM was detected at 40-720 sites/cell in MDCK, A431, PAM212, Lu99, and IM-9 cells, but not in fibroblasts and hematopoietic cells. In contrast, low-affinity binding sites were detected in all cell lines examined, even in those not responsive to HGF. Northern blots revealed that cells possessing a high-affinity HGF receptor expressed c-MET/HGF receptor mRNA. Therefore, HGF probably regulates both cell growth and motility of various types of epithelial cells and some types of mesenchymal cells. The multiple biological activities of HGF may be exerted through a high-affinity HGF receptor linked to multiple distinct intracellular signaling pathways.

Hepatocyte growth factor rapidly induces the tyrosine phosphorylation of 41-kDa and 43-kDa proteins in mouse keratinocytes

We have examined the hepatocyte growth factor (HGF)-mediated changes in protein-tyrosine phosphorylation in mouse keratinocytes (PAM-212) and canine kidney epithelial cells (MDCK). In PAM-212 cells HGF and epidermal growth factor, both of which stimulated the DNA synthesis, rapidly induced the tyrosine phosphorylation of two 41-kDa and two 43-kDa proteins: increased tyrosine phosphorylation of those proteins has been commonly observed when quiescent fibroblasts are stimulated with a variety of mitogenic agents. In contrast, HGF did not stimulate the DNA synthesis but induced cell dissociation in MDCK cells; under this condition, increased tyrosine phosphorylation of the 41-kDa and 43-kDa protein was not observed. A possible role of the increased tyrosine phosphorylation of 41-kDa and 43-kDa protein in the signaling pathway of HGF is discussed.

Promotion of hepatic metastases by liver resection in the rat

In the early period following radical hepatectomy for hepatoma, recurrences in the remaining liver are frequently found. In regenerating liver, implantation and growth of tumour cells released into the portal system during surgical treatment might be promoted. We examined the relationship between liver regeneration and the formation of metastases following hepatic resection. Intraportal injections of rat ascites containing hepatoma AH130 cells at a concentration of 1 x 10(5) cells 0.2 ml-1 were made at various periods following two thirds liver resection in rats. Tumour cell injections immediately at 24 h after surgery resulted in an increased number of hepatic metastases compared with control animals. Tumour cell injections 2 weeks after hepatectomy, however, had no significant difference in effect compared with control rats. In contrast, tumour cells injected immediately after removal of half of the caudate lobe resulted in the same number of metastases as control animals. These results demonstrate that the number of artificially induced hepatic metastases was increased during an initial period of active liver regeneration and was proportional to the volume of hepatectomy. The effect of 5-fluorouracil (5FU) or mitomycin C (MMC) as inhibitors of hepatic regeneration on liver metastasis after hepatectomy was studied. The administration of 5FU (20 mg kg-1) or MMC (0.2 mg kg-1) immediately, 24 and 48 h after hepatectomy resulted in a marked reduction in metastatic lesions. The administration of 5FU caused delays in weight gain and decreases in the wet weight of remaining liver, while MMC had no effect on either. Accordingly, results of 5FU administration may be due to inhibitory effects on liver regeneration whilst that of MMC administration may be due to cytocidal antitumour effect. The effect of OK-432 as an immunoactivator on the implantation and growth of tumour cells in regenerating liver was also studied. Pretreatment with OK-432, 0.5 mg intraperitoneally on 7 consecutive days, had no effect on hepatic metastases. The pathophysiology of liver regeneration may enhance hematogenous hepatic metastasis and release of tumour cells during surgical manipulation may represent an important cause of recurrence following hepatic resection.

Partial purification and characterisation of an inhibitor of hepatocyte proliferation derived from nonparenchymal cells after partial hepatectomy

We have investigated the influences that nonparenchymal cells from regenerating rat liver exert on hepatocyte proliferation. When primary adult rat hepatocytes isolated from resting liver were co-cultured with nonparenchymal cells (NPCs) from resting liver of a different syngeneic animal, the proliferative response of hepatocytes to epidermal growth factor (EGF) was unaffected by the presence of NPCs. In the presence of NPCs taken from livers that had undergone partial hepatectomy 24 hours before (regen-NPCs), the response of hepatocytes from resting liver to EGF, TGF-alpha, and hepatocyte growth factor (HGF) was markedly inhibited. Inhibitory activity was not dependent on cell-to-cell contact, and conditioned-medium from regen-NPCs, but not normal NPCs, inhibited EGF-induced hepatocyte DNA synthesis by approximately 50%. After concentration by gel chromatography and lyophilisation, inhibition was 98%. The inhibitory activity migrated on SDS-PAGE gel electrophoresis with an apparent molecular weight of 14 to 17 kDa and was trypsin-sensitive but relatively heat-stable. The effects of blocking antibodies established that it was not TGF-beta 1, IL1-beta, or IL6. Investigations of regen-NPCs taken at different time points demonstrated that inhibitory activity was released into conditioned medium of cells harvested at 24 and 48 hours after partial hepatectomy, but not 10 or 72 hours. This powerful inhibitor of hepatocyte response to proliferogens is released by cultures of NPCs with a time course suggesting that it may be involved in terminating the surge of hepatocyte replication induced by partial hepatectomy.

Alteration of rat liver proteoglycans during regeneration

Sepharose CL-6B column chromatography of crude extracts from the slices of regenerating rat livers after partial hepatectomy and sham-operated controls labeled with [35S]sulfuric acid revealed an enhancement of [35S]sulfate incorporation into proteoglycan fractions during regeneration. The 35S-labeled proteoglycans contained heparan sulfate (more than 80% of the total) and chondroitin/dermatan sulfate. The 35S-incorporation into both glycosaminoglycans increased to maxima 3-5 days after partial hepatectomy and decreased thereafter toward the respective control levels. When [35S]sulfuric acid was replaced by [3H]glucosamine, similar results were obtained. These results suggest that the maximal stimulation of proteoglycan synthesis in regenerating rat liver follows the maximal mitosis of hepatic cells 1-2 days after partial hepatectomy. The 35S-labeled proteoglycans from regenerating liver 3 days after partial hepatectomy and control were analyzed further. They were similar in chromatographic behavior on a gel filtration or an anion-exchange column and in glycosaminoglycan composition. Their glycosaminoglycans were indistinguishable in electrophoretic mobility. However, these proteoglycans were slightly but significantly different in their affinity to octyl-Sepharose and in the molecular-weight distribution of their glycosaminoglycans.

Expression of c-met proto-oncogene in COS cells induces the signal transducing high-affinity receptor for hepatocyte growth factor

By transfection of the expression plasmid containing a human c-met cDNA into COS-7 cells, high-affinity binding sites specific for HGF with a Kd value of 30 pM were newly detected. Furthermore, only in the c-met transfected COS-7 cells, but not in the control COS-7 cells, DNA synthesis was markedly induced in response to HGF. Thus, transient expression of exogenous c-met cDNA resulted in the appearance of high-affinity receptor for HGF and conversion of the normally non-responsive COS-7 cells into the HGF-responsive cells. These results provide evidence for identifying the c-met product as a signal transducing high-affinity receptor for HGF.

The gene of hepatocyte growth factor is expressed in fat-storing cells of rat liver and is downregulated during cell growth and by transforming growth factor-beta

Hepatocyte growth factor (HGF) has been detected in non-parenchymal cells but not in hepatocytes. We performed Northern blot analysis of total RNA extracted from rat hepatocytes, Kupffer cells, endothelial cells and fat-storing (Ito-) cells. Total RNA was extracted from fat-storing cells at different times after isolation and from cells treated with different amounts of transforming growth factor beta. The RNA was hybridized with HGF, fibronectin-, and alpha-actin-specific cDNA probes, consecutively. We found an abundant amount of HGF mRNA in freshly isolated fat-storing cells, but not in other liver cells. The amount of the HGF transcripts decreases significantly in FSC during the time of culture, while fibronectin gene expression increases and alpha-actin gene expression as well. TGF-beta dramatically inhibits HGF gene expression, but causes an enhanced fibronectin mRNA level. Northern blot hybridisation of total RNA from CCl4-chronically damaged liver with HGF cDNA shows a significant increase of HGF mRNA during development of liver fibrosis. We suggest that in damaged liver either non-parenchymal cells, others than FSC, became able to express the HGF in vivo, or other mediators overcome the inhibitory effect of TGF-beta.

Hepatocyte growth factor (HGF) mediates the sustained formation of 1,2-diacylglycerol via phosphatidylcholine-phospholipase C in cultured rat hepatocytes

The addition of hepatocyte growth factor (HGF) to rat hepatocytes in primary culture resulted in the formation of inositol 1,4,5-trisphosphate (Ins(1,4,5)P3) and 1,2-diacylglycerol (DG) by a phosphoinositide-specific phospholipase C (PI-PLC). DG showed a biphasic increase; the first phase, corresponding with the peak of Ins(1,4,5)P3 and a second larger and prolonged phase. The HGF stimulates the phosphatidylcholine (PC)-derived prolonged DG formation by a phospholipase C pathway (PC-PLC) but not by a phospholipase D pathway. HGF also was found to elicit [Ca2+] oscillations which may be associated with the prolonged DG production from PC via the PC-PLC phospholipase C pathway.