Growth retardation, duodenal lesions, and aberrant ileum architecture in triple null mice lacking EGF, amphiregulin, and TGF-alpha.

BACKGROUND AND AIMS: Mice lacking epidermal growth factor (EGF), transforming growth factor alpha, and amphiregulin were used to identify roles for these EGF receptor (EGF-R) ligands in gastrointestinal development and mucosal integrity. METHODS: Gastrointestinal tract development was examined in knockout mice and correlated with expression of EGF-R protein and EGF family members throughout the gut. Crossfostering experiments addressed roles of maternal- and neonatal-derived ligands in pup growth and intestinal development. Cysteamine-induced ulceration in EGF(-/-) mice was used to examine its role in mucosal cytoprotection. RESULTS: Neonatal mice lacking all 3 ligands were growth retarded, even when reared by wild-type dams; conversely, lack of maternal ligands transiently impaired wild-type pup growth. Triple null neonates displayed spontaneous duodenal lesions, and ileal villi were truncated and fragile with reduced cellular proliferation in the crypts. However, maturation of digestive enzymes was unaffected. Adult EGF(-/-) mice displayed more severe lesions in response to cysteamine treatment compared with wild-type counterparts, although triple null mice were not more susceptible to dextran sulfate sodium-induced colitis, suggesting a differential role for these ligands in the injury response. CONCLUSIONS: EGF-R ligands are required for development and mucosal maintenance in mouse small intestine. Both maternal and neonatal sources of growth factors are required for optimal pup growth.

Targeted inactivation of the EGF and amphiregulin genes reveals distinct roles for EGF receptor ligands in mouse mammary gland development.

Targeted mice lacking functional EGF or amphiregulin (AR) were derived and bred to the TGFalpha-knockout to generate mice lacking various combinations of the three ligands. In contrast to EGF receptor (EGFR) knockout mice, triple null mice lacking half of the EGFR ligand family were healthy and fertile, indicative of overlapping or compensatory functions among EGF family members. Nevertheless, pups born to triple null dams frequently died or were runted, suggesting a mammary gland defect. Comparison of individual and combinatorial knockouts established that specific loss of AR severely stunted ductal outgrowth during puberty, consistent with dramatic expression of AR transcripts in normal developing ducts. Surprisingly, loss of all three ligands did not significantly affect cellular proliferation, apoptosis, or ERK activation within terminal end buds. Following pregnancy, most AR single null females, but few triple null females could nurse their young, revealing collaborative roles for EGF and TGFalpha in mammopoiesis and lactogenesis. In triple null glands, alveoli were poorly organized and differentiated, and milk protein gene expression was decreased. Additionally, Stat5a activation was frequently reduced in AR single and combinatorial nulls in association with impaired lactation. Collectively, our results provide genetic confirmation of a requirement for EGFR signaling throughout the development of the mouse mammary gland, and reveal stage-dependent activities for different EGFR ligands. Finally, the additional loss of growth factors from pups nursed by triple null dams further worsened their survival and growth, establishing functions for both maternal- and neonatal-derived growth factors.

Role of integrins in mouse eyelid development: studies in normal embryos and embryos in which there is a failure of eyelid fusion.

Eyelid fusion normally occurs between E15.5 and E16.5 of mouse embryonic development and results from the migration of a population of periderm-derived epithelial cells over the corneal surface. Cell migration is known to depend on extracellular matrix receptors of the integrin family and to be regulated by growth factors. We were therefore interested that a failure of eyelid fusion has been reported in mice that are homozygous null for the transforming growth factor alpha (TGF-alpha) gene and in mice (invalpha5beta1) in which a transgenic alpha5beta1 integrin under the control of the involucrin promoter is misexpressed in differentiating keratinocytes. We examined expression of the alpha2beta1, alpha3beta1, alpha5beta1 and alpha6beta4 integrins during eyelid fusion in wild-type embryos and found selective upregulation of the alpha5beta1 integrin and its ligand, fibronectin, in the migrating eyelid tip cells. In TGF-alpha null embryos, the failure of eyelid fusion was correlated with a failure to upregulate the alpha5beta1 integrin and fibronectin in the tip cells. Using beta-galactosidase as a reporter gene in transgenic mice, we observed specific activity of the involucrin promoter in the eyelid tip cells. In invalpha5beta1 mice the transgenic human integrin was overexpressed not only in the tip cells but throughout the eyelid epidermis. In contrast, the endogenous, murine, alpha5beta1 integrin was only weakly expressed in the tip cells. We speculate that selective and coordinated expression of the alpha5beta1 integrin and fibronectin in eyelid tip cells is required for eyelid fusion and may be under the control of growth factors that include TGF-alpha.

TGF alpha is dispensable for skin tumorigenesis in Tg.AC mice.

Alterations in growth factor signaling pathways frequently accompany the development and maintenance of epithelial neoplasia. Transforming growth factor alpha (TGF alpha) and its epidermal growth factor receptor have been thought to play an especially important role in epithelial neoplasia. In this study, mice were derived genetically deficient (null) in functional TGF alpha expression and carrying the Tg.AC/v-Ha-ras transgene. The goals were to determine if (a) papillomagenesis was dependent on TGF alpha and (b) progression to malignancy was dependent on TGF alpha expression. Groups of male and female mice heterozygous or homozygous for the TGF alpha null allele and hemizygous for the Tg.AC transgene were treated twice weekly for 10 or 15 wk with doses of 12-O-tetradecanoylphorbol-13-acetate (TPA) known to produce papillomas in Tg.AC mice. Papillomas were readily induced in both male and female TGF alpha null mice. Malignant progression of papillomas was observed in all TGF alpha null treatment groups. Additionally, we examined the response of TGF alpha null mice to full thickness dorsal wounds, a stimulus known to promote papillomagenesis in Tg.AC mice. As in the TPA study, papillomas were induced in both male and female TGF alpha null mice. These studies indicate that TGF alpha is not required for the induction and maintenance of papillomas nor is it essential for the malignant conversion of papillomas in Tg.AC mice.

Characterization of the mouse transforming growth factor alpha gene: its expression during eyelid development and in waved 1 tissues.

The spontaneous mouse waved 1 (wa1) mutation is allelic with the transforming growth factor alpha (TGF-alpha) gene and produces phenotypes similar to those of TGF-alpha knockout mice. Here, we show that TGF-alpha mRNA and protein levels are measurable in wa1 tissues but reduced 5- to 30-fold relative to wild type. Because the wa1-coding sequence is identical to that of the normal mRNA, wa1 is not a null mutation. Nuclear run-on analyses revealed decreased transcription of the TGF-alpha gene in wa1 tissues, but the sequence of a 3.2-kb 5' flanking fragment containing the promoter was unaltered. Moreover, pulsed field gel electrophoresis analysis did not reveal alterations within 750 kb upstream or 350 kb downstream of the gene, and chromosome 6 was karyotypically normal. Hence, we speculate that the wa1 mutation may be subtle and/or reside at a greater distance from the TGF-alpha gene. TGF-alpha deficiency elicits a spectrum of variably penetrant eye anomalies in wa1 and knockout mice that are associated with open eyes at birth. We found that late-gestation wa1 and TGF-alpha-null embryos display a significant delay in eyelid closure, although the eyes of most embryos fuse prior to birth. In situ hybridization localized TGF-alpha expression to the advancing margins of the eyelid epithelium and epidermal growth factor receptor expression throughout the eyelid and corneal epithelia. These results suggest that eye problems observed in TGF-alpha-deficient adult mice arise from premature exposure and trauma to open eyes during or following parturition.

Liver regeneration and hepatocarcinogenesis in transforming growth factor-alpha-targeted mice.

Transforming growth factor-alpha (TGF alpha), a member of the epidermal growth factor receptor ligand family, has been implicated in the regeneration and transformation of liver. Our recent development of mice that are homozygous for a disrupted TGF alpha gene allowed us to assess the requirement for this growth factor in these complex processes. We report here that although a 70% hepatectomy produced a significant increase in hepatic TGF alpha protein levels in wild-type mice, liver regeneration nevertheless proceeded normally in the absence of the growth factor. The hepatocyte labeling indices determined for homozygous targeted and wild-type mice at 36 and 48 h after hepatectomy were comparable, and the total liver DNA to body weight ratios 8 d after hepatectomy were essentially identical for the two genotypes. These results indicate that TGF alpha, is not necessary for liver regeneration. To test its requirement in liver carcinogenesis, young mice were administered single doses of diethylnitrosamine (DEN) with or without subsequent chronic treatment with the promoting agent phenobarbital (PB). Both wild-type and homozygous mutant male mice treated with DEN or DEN plus PB developed multiple preneoplastic foci or tumors by 9 mo of age with relatively high incidence. However, while five of 88 tumors in wild-type mice attained a diameter greater than 5 mm and were classified as hepatocellular carcinomas, none of 132 tumors in livers of targeted mice reached this size. Furthermore, three of these large wild-type tumors expressed significantly elevated levels of TGF alpha protein compared with normal liver. These results indicate that TGF alpha is not required for early events in chemically induced hepatocarcinogenesis but suggest that it could be important in the progression from small preneoplastic foci to large tumors.

Mouse chromosomal location of three EGF receptor ligands: amphiregulin (Areg), betacellulin (Btc), and heparin-binding EGF (Hegfl).

The products of five distinct loci, Egf, Tgfa, Areg, Btc, and Hegfl act as ligands for the epidermal growth factor receptor. Egf and Tgfa have previously been mapped to mouse chromosomes 3 and 6, respectively, but the mouse chromosomal locations of Areg, Btc, and Hegfl have not been reported. Here, we show that Areg and Btc are tightly linked on mouse chromosome 5, while Hegfl maps to mouse chromosome 18. We also provide evidence that a putative sixth family member, Sdgf, is in fact a species variant of Areg. These results confirm and extend known relationships between mouse and human chromosomes, and they also provide new information regarding the evolution of this important gene family.

The mouse waved-2 phenotype results from a point mutation in the EGF receptor tyrosine kinase.

Mice harboring the waved-1 (wa-1) and waved-2 (wa-2) mutations exhibit skin and eye abnormalities that are strikingly similar to those of TGF-alpha-deficient mice, and wa-1 and TGF-alpha were recently shown to be allelic. Because the wa-2 mutation was mapped previously to the vicinity of the EGF/TGF-alpha receptor (EGFR) gene on mouse chromosome 11, we hypothesized that the wa-2 phenotype might result from a defect in either the expression or activity of EGFR, or both. In the present report, we show that EGFR mRNA and protein of normal size are expressed in wa-2 liver and skin at levels that are comparable to those in the corresponding normal tissues, and that the ability of wa-2 EGFR to bind ligand is unaltered. However, ligand-dependent autophosphorylation of wa-2 EGFR is diminished 5- to 10-fold in vitro, and the ability of wa-2 EGFR to phosphorylate an exogenous substrate is reduced by > 90% compared with that of the control receptor. EGF-induced tyrosine phosphorylation, including that of EGFR itself, is also diminished in skin, particularly at lower dose of exogenous EGF. To establish the nature of the wa-2 mutation, we determined the nucleotide sequence of the coding region of normal and wa-2 murine EGFR cDNAs. A comparison of these sequences revealed a single-nucleotide transversion resulting in the substitution of a glycine for a conserved valine residue near the amino terminus of the tyrosine kinase domain. The importance of this mutation was confirmed by showing that its introduction into an otherwise normal EGFR markedly reduced the receptor's tyrosine kinase activity in transfected Chinese hamster ovary cells. Finally, in situ hybridization analysis demonstrated expression of EGFR predominantly in the outer root sheath of active hair follicles in neonatal mice. As we previously localized TGF-alpha mRNA to the inner root sheath, this pattern of EGFR expression is consistent with the effect of the wa-2 mutation on hair structure, and together with our previous characterization of TGF-alpha-deficient mice, reveals a critical role for signaling by this ligand/receptor system in skin.

TGF alpha deficiency results in hair follicle and eye abnormalities in targeted and waved-1 mice.

To explore the physiological roles of transforming growth factor alpha (TGF alpha), we disrupted the mouse gene by homologous recombination in embryonic stem cells. Homozygous mutant mice were viable and fertile, but displayed pronounced waviness of the whiskers and fur, accompanied by abnormal curvature, disorientation, and misalignment of the hair follicles. Homozygous and, to a lesser extent, heterozygous mice displayed eye abnormalities of variable incidence and severity, including open eyelids at birth, reduced eyeball size, and superficial opacity. Histological examination revealed eyelid and anterior segment dysgenesis, corneal inflammation and scarring, and lens and retinal defects. Although TGF alpha deficiency affected skin and eyes, wound healing in these tissues was not impaired. Similar hair and eye defects have been previously associated with the recessive mutation waved-1 (wa-1), and Northern analysis revealed reduced expression of TGF alpha in wa-1 mice. Crosses between wa-1 homozygotes and TGF alpha-targeted mice confirmed that wa-1 and TGF alpha are allelic.

Regulation of fat and muscle development by transforming growth factor alpha in transgenic mice and in cultured cells.

Transgenic mice overexpressing transforming growth factor alpha (TGF-alpha) under control of the metallothionein promoter had, on average, 20% reductions in body and carcass weights compared to nontransgenic littermates. This loss resulted from significant decreases in the comparative weights of bone, muscle, and especially fat. Transgenic epididymal fat pads were reduced by 40-80%, and total body fat content by 50%, relative to control animals. Distal hindlimb muscle weights were 20% below normal, and other skeletal muscles were visibly smaller in size. Weight reductions were accompanied by decreases in the cellularity of transgenic fat pads and muscles and by decreases in the number and area of striated muscle fibers. These findings were not obviously attributable to differences in metabolic rates since transgenic and control mice displayed similar levels of energy expenditure per unit lean body mass. The effects of TGF-alpha on the development of these tissues could be mimicked in culture for fat but not muscle. Thus, TGF-alpha did not inhibit the differentiation of the mouse skeletal myoblast cell line C2C12 as evidenced by the expression of muscle-specific actin and fusion to form multinucleated myotubes. However, TGF-alpha repressed the differentiation of the preadipocyte cell line 3T3-F442A in a dose-dependent and reversible manner as judged by morphological conversion and diminished expression of mRNAs encoding the adipocyte-specific markers adipsin and glycerophosphate dehydrogenase. This repression, which occurred without marked stimulation of proliferation, was incomplete even in the presence of high concentrations of growth factor. Despite its effects on adipose development, introduction of the metallothionein-TGF-alpha transgene into the ob/ob genetic background did not suppress the marked obesity characteristic of this mutation. Finally, endogenous TGF-alpha epidermal growth factor receptor mRNAs were detected in normal adipose tissue, suggesting that regulation of adipogenesis by this growth factor may be physiological.

Transforming growth factor alpha dramatically enhances oncogene-induced carcinogenesis in transgenic mouse pancreas and liver.

To characterize the effect(s) of transforming growth factor alpha (TGF alpha) during multistage carcinogenesis, we examined tumor development in pancreas and liver of transgenic mice that coexpressed TGF alpha with either viral (simian virus 40 T antigens [TAg]) or cellular (c-myc) oncogenes. In pancreas, TGF alpha itself was not oncogenic, but it nevertheless dramatically accelerated growth of tumors induced by either oncogene alone, thereby reducing the host life span up to 60%. Coexpression of TGF alpha and TAg produced an early synergistic growth response in the entire pancreas together with the more rapid appearance of preneoplastic foci. Coexpression of TGF alpha and c-myc also accelerated tumor growth in situ and produced transplantable acinar cell carcinomas whose rate of growth was TGF alpha dependent. In liver, expression of TGF alpha alone increased the incidence of hepatic cancer in aged mice. However, coexpression of TGF alpha with c-myc or TAg markedly reduced tumor latency and accelerated tumor growth. Significantly, expression of the TGF alpha and myc transgenes in hepatic tumors was induced up to 20-fold relative to expression in surrounding nonneoplastic liver, suggesting that high-level overexpression of these proteins acts as a major stimulus for tumor development. Finally, in both pancreas and liver, combined expression of TGF alpha and c-myc produced tumors with a more malignant (less differentiated) appearance than did expression of c-myc alone, consistent with an influence of TGF alpha upon the morphological character of c-myc-induced tumor progression. These findings demonstrate the importance of TGF alpha expression during multistage carcinogenesis in vivo and point to a major role for this growth factor as a potent stimulator of tumor growth.

Transforming growth factor-alpha is a potential mediator of estrogen action in the mouse uterus.

To better understand the role of peptide growth factors in sex steroid hormone-mediated growth of the female reproductive tract, the effect of estrogen on the expression of transforming growth factor-alpha (TGF alpha) in mouse uterus was investigated. Our results show that estrogen induces the expression of TGF alpha mRNA in the mouse uterus in a dose- and time-dependent manner. The up-regulation of TGF alpha transcripts occurs predominantly in uterine epithelial cells. RIA and Western blot analysis demonstrate that immunoreactive TGF alpha protein is secreted at high levels into mouse uterine luminal fluid after estrogen treatment. The induction of uterine TGF alpha mRNA is specific to estrogen; nonestrogenic steroids did not induce expression. Antibody specific to TGF alpha significantly reduces estrogen-mediated uterine growth, which supports the concept that TGF alpha is a mitogen for the reproductive tract. Analysis of TGF alpha/EGF receptors by binding, affinity labeling, and phosphorylation studies indicates that functional receptors are present in the mouse uterus after estrogen exposure. Thus, our data support a physiological role for TGF alpha and its receptor pathway in the female mouse reproductive tract.

Expression of the TGF alpha integral membrane precursor induces transformation of NRK cells.

TGF alpha is one member of a family of soluble growth factors that are derived from integral-membrane precursors. The mature form of TGF alpha is released from its transmembrane precursor (proTGF alpha) by a protease that, in many tumor cells, is inefficient or limiting. We have previously established that, in the absence of processing, membrane-anchored proTGF alpha is biologically active and can interact with the EGF receptor on adjacent cells, thereby inducing the receptor's intrinsic tyrosine kinase activity. We further showed that this interaction leads to immediate downstream signal transduction as evidenced by Ca2+ mobilization. To extend these observations, and to investigate its transforming potential, we infected normal rat kidney (NRK) cells with retroviral expression vectors that encode mutated forms of proTGF alpha containing amino acid substitutions at the proteolytic cleavage sites. NRK cells harboring these mutant constructs do not secrete mature growth factor, but do express biologically active proTGF alpha on the cell surface as shown by their ability to induce the autophosphorylation of EGF receptor on neighboring A431 cells in co-culture. Expression of the mutant proTGF alpha molecules promoted the anchorage-independent growth of NRK cells in soft agar, and caused them to be tumorigenic when injected into nude mice. These results demonstrate that an interaction between EGF receptor and the integral membrane precursor to TGF alpha can provide a mitogenic stimulus that leads to transformation. They further suggest that the accumulation of proTGF alpha on the surface of some transformed cells has physiological relevance.

Transforming growth factor alpha: expression, regulation and biological action of its integral membrane precursor.

Transforming growth factor alpha (TGF alpha) is a 6 kDa polypeptide mitogen that interacts with the epidermal growth factor receptor and activates its intrinsic tyrosine kinase. The mature 50 amino acid TGF alpha is released from a 159 or 160 amino acid integral membrane glycoprotein precursor, denoted proTGF alpha, via cleavage at both termini by an unknown protease with elastase-like specificity. Rat proTGF alpha is encoded by a 4.5 kb mRNA that is transcribed from a gene containing 6 exons and spanning 85 kb of DNA. Expression of TGF alpha is most prevalent and abundant in transformed cells and tumors, but also detectable at modest levels in a limited number of normal cells and tissues. In many neoplastic cells, proteolytic processing of proTGF alpha is incomplete and/or inefficient, resulting in the preponderance of soluble and/or membrane-bound forms larger than the mature TGF alpha. To characterize the biological activities of the transmembrane TGF alpha precursor in the absence of processing, amino acid substitutions were introduced at the cleavage sites by site-directed mutagenesis of the rat TGF alpha cDNA. Fibroblasts expressing the mutant proTGF alpha constructs did not secrete TGF alpha, but did accumulate proTGF alpha at the cell surface. Coincubation of these cells with A431 cells resulted in binding and autophosphorylation of EGF receptors, and mobilization of intracellular calcium in A431 cells, demonstrating that the transmembrane proTGF alpha can activate EGF receptors on adjacent cells, leading to signal transduction. In addition, rat fibroblasts constitutively expressing the wild-type or mutant proTGF alpha became morphologically transformed in culture, and induced tumors in nude mice. Thus, the interaction between membrane-anchored ligand and receptor triggers mitogenesis that can culminate in neoplastic transformation. To characterize the physiological and pathological effects of TGF alpha in vivo, particularly with respect to epithelial cells, transgenic mice were developed which overexpress the growth factor in multiple or specific tissues. Widespread overexpression of TGF alpha driven by the metallothionein promoter induced epithelial hyperplasia in several organs, including liver and intestine, without disrupting normal tissue architecture. In contrast, the pancreas displayed increased proliferation of both acinar cells and fibroblasts, and focal alteration of acinar cell differentiation. This pancreatic hyperplasia, fibroplasia, and metaplasia were reproduced when TGF alpha expression was placed under control of the elastase promoter, and thus locally restricted to acinar cells, suggesting autocrine and/or paracrine mode of action. Finally, overexpression of TGF alpha promoted neoplastic transformation of certain epithelia. In coagulation gland, there was dramatic hyperplasia and dysplasia with focal evidence of carcinoma in situ.(ABSTRACT TRUNCATED AT 400 WORDS)

Overexpression of TGF alpha in transgenic mice: induction of epithelial hyperplasia, pancreatic metaplasia, and carcinoma of the breast.

Metallothionein-directed expression of TGF alpha in transgenic mice induced a spectrum of changes in the growth and differentiation of certain adult tissues. First, TGF alpha promoted a uniform epithelial hyperplasia of several organs without otherwise causing major alterations in tissue architecture. Second, in pancreas it promoted proliferation of both acinar cells and fibroblasts and focally altered acinar cell differentiation. The magnitude of this response was proportional to the level of local, tissue-specific TGF alpha expression and was reproduced when expression of TGF alpha was placed under the control of the elastase promoter, implying an autocrine or paracrine mechanism. Third, TGF alpha was oncogenic in vivo. It caused dramatic hyperplasia and dysplasia of the coagulation gland epithelium, which displayed evidence of carcinoma in situ, and in postlactational mammary gland it induced secretory mammary adenocarcinomas. Thus, TGF alpha displays characteristics of both a potent epithelial cell mitogen and an oncogenic protein in vivo.

Characterization of high molecular weight transforming growth factor alpha produced by rat hepatocellular carcinoma cells.

In addition to the mature 50 amino acid transforming growth factor alpha (TGF alpha), some transformed cells appear to produce multiple higher molecular weight forms. The structure and derivation of most of these larger soluble TGF alpha species remain to be established. We previously reported that a chemically induced rat hepatocellular carcinoma cell line, JM1, secreted acid-stable proteins which bind to epidermal growth factor receptors and stimulate DNA synthesis in primary cultures of normal adult rat hepatocytes. Purification and characterization of these hepatoma-derived growth factors have indicated their relationship to TGF alpha. Two EGF-competing activities of apparent Mr 30K and 10K were separated by gel filtration of concentrated JM1-conditioned medium and further purified by ion-exchange chromatography and reverse-phase HPLC. Both growth factors were detected by a radioimmunoassay specific for TGF alpha. Western blotting with antibodies to the 50 amino acid TGF alpha revealed that the lower molecular weight factor comigrated with the synthetic 6-kDa rat TGF alpha. The higher molecular weight TGF alpha appeared on immunoblots as a diffuse band of 18-21 kDa, which converted to the mature 6-kDa form upon digestion with elastase, confirming a precursor-product relationship. However, the 18-21-kDa proteins did not react with antibodies directed against the carboxy-terminal cytoplasmic segment of the transmembrane TGF alpha precursor. Enzymatic deglycosylation of the 18-21-kDa TGF alpha species by sequential removal of sialic acids and O- and N-linked carbohydrate reduced the molecular weight to 11K. The size and soluble nature of this polypeptide suggest that it represents the extracellular domain of the transmembrane TGF alpha precursor.(ABSTRACT TRUNCATED AT 250 WORDS)

Differential effect of growth factors on growth stimulation and phenotypic stability of glutamine-synthetase-positive and -negative hepatocytes in primary culture.

In rat liver parenchyma, two subpopulations of hepatocytes can be distinguished by the absence or presence of the marker enzyme, glutamine synthetase (GS). Hepatocytes in the perivenous zone immediately adjacent to the hepatic venules in the liver acinus are positive for GS. Using autoradiography in combination with immunocytochemistry, the response of these two hepatocyte populations (GS positive and GS negative) to a variety of growth factors (defined compounds or complex stimuli) was investigated in vitro. Irrespective of the individual growth-promoting activity (which varied considerably), all stimuli led to much higher labeling indices in GS-negative cells as compared to GS-positive cells. In GS-negative cells, the strongest effect was exerted by serum obtained from partially hepatectomized rats (labeling index, 67%) and the conditioned media of JM1 and JM2 hepatoma cells (63%-82%), followed by a combination of insulin and either norepinephrine (46%) or epidermal growth factor (EGF; 42%). In contrast, serum had the weakest influence on GS-positive cells (0.3%), while the other potent stimuli enhanced the labeling index of these cells by between 6% and 15% within 48 h. The percentage of labeled nuclei was higher in mononucleated than in binucleated GS-positive hepatocytes. The time course of thymidine incorporation was also different for the two subpopulations. Under all growth-promoting conditions, the stimulation of GS-negative cells peaked between 72 and 96 h, while it increased continuously in GS-positive cells for at least 120 h, particularly in the case of serum. In proliferating cultures, both the absolute and the relative number of GS-positive hepatocytes decreased, while no such effect was found in various nonproliferating control cultures maintained at low and high cell density. Similar results were found for GS activity. In contrast, the hormonal induction of tyrosine aminotransferase (TAT) was not affected. It is suggested that these differences in the growth response of GS-positive and -negative cells contribute to the acinar gradient in hepatocyte proliferation that occurs during liver regeneration. Furthermore, the striking phenotypic instability of GS-positive cells that have undergone DNA synthesis and mitosis supports the hypothesis that cellular reprogramming depends on passage through the cell cycle.

Partial purification and characterization of a hepatocyte growth factor produced by rat hepatocellular carcinoma cells.

Serum-free medium conditioned by confluent cultures of JM1 of JM2 rat hepatocellular carcinoma cells stimulated DNA synthesis in primary cultures of adult rat hepatocytes in a dose-dependent, saturable manner and in the absence of epidermal growth factor. The hepatotrophic activity was nondialyzable in Mr 50,000 cutoff membranes, heat (60 degrees C) and acid stable, and sensitive to trypsin and dithiothreitol treatment. Gel filtration of concentrated JM1 or JM2 conditioned medium on Sephadex G-100 separated the activity into two regions, the major broad peak migrating with apparent Mr 25,000. Chromatography fractions active in the hepatocyte proliferation bioassay also inhibited specific binding of iodinated epidermal growth factor to cultures of A431 carcinoma cells and rat hepatocytes. These results suggest that neoplastic liver cells synthesize and secrete polypeptide growth factors which can bind to epidermal growth factor receptors and stimulate proliferation of normal adult rat hepatocytes in primary culture.