Changes in TGF-β receptors of rat hepatocytes during primary culture and liver regeneration: Increased expression of TGF-β receptors associated with increased sensitivity to TGF-β-mediated growth inhibition

To clarify the role of transforming growth factor-β (TGF-β) and its receptors in hepatocyte growth, we studied the expression of TGF-β1 and its receptors and the sensitivity to growth inhibition by TGF-β1 protein in rat hepatocytes derived from resting and regenerating livers. In hepatocytes derived from resting livers, mRNAs for TGF-β type II receptor (TβR-II), insulin-like growth factor-II/mannose 6-phosphate receptor (IGF-II/M-6-PR), and TGF-β1 increased with time in primary culture. The cell surface TGF-β receptor proteins (TβR-I, II, and III), examined by the receptor affinity-labeling assay using ¹²⁵I-TGF-β1, also increased, especially after 48 hr of culture. Hepatocytes were more sensitive to inhibition of DNA synthesis, when the TGF-β1 protein was added at later times in culture, corresponding to the presence of increased TGF-β receptors. In hepatocytes from regenerating livers after a partial hepatectomy (PH), an increase of TβR-I, TβR-II, TβR-III, IGF-II/M-6-PR, and TGF-β1 mRNAs was found, compared with hepatocytes from resting livers. Similarly, using TGF-β receptor affinity-labeling assay, hepatocytes from PH livers were found to have an increase in TβR-I, II, and III proteins, with a peak at 4 days post-PH, compared with hepatocytes from resting livers. When TGF-β1 protein was added for a short period (6 or 24 hr) after cell attachment to hepatocyte cultures, it inhibited DNA synthesis more effectively in hepatocytes from regenerating compared with resting livers. Our results show that hepatocyte TGF-β receptors and sensitivity to growth inhibition by TGF-β1 protein change together and are modulated during liver regeneration, as well as during the conditions of primary culture. J. Cell. Physiol. 176:612–623, 1998. © 1998 Wiley-Liss, Inc.     

Clonal heterogeneity of the sensitivity of human colon carcinoma cell lines to TGFβ isoforms

Spontaneously arising, TGFβ₁-resistant colonies were isolated directly from the soft agarose plates of MOSER human colon carcinoma cells grown in the presence of TGFβ₁ but in the absence of serum. The colonies were cloned by limiting dilution and screened in a monolayer proliferation assay for sensitivity to TGFβ₁ and TGFβ₂ isoforms. Cell clones selectively sensitive or resistant to these isoforms in the growth inhibition assay displayed similar differential sensitivities to TGFβ isoforms for production of the extracellular matrix proteins laminin and fibronectin, as well as for the expression of the colon cell differentiation marker carcinoembryonic antigen. Differential receptor binding profiles for TGFβ₁ and TGFβ₂ were observed among the clones. The isolation of cell clones selectively resistant or sensitive to TGFβ isoforms as well as the identification of differential receptor binding profiles among the clones indicate the heterogeneity of TGFβ responsiveness that exists naturally in human colon tumor cells and stress the importance of defining mechanisms underlying differential responsiveness to TGFβ isoforms. © 1995 Wiley-Liss Inc.     

Requirement for receptor-bound urokinase in plasmin-dependent cellular conversion of latent TGF-β to TGF-β

The role of receptor-bound urokinase-type plasminogen activator (uPA) in cellular activation of latent transforming growth factor-beta (LTGF-β) was investigated in a model system of mouse LB6 cells transfected with either a human uPA receptor cDNA (LhuPAR⁺). a human prouPA cDNA (LhuPA), or a control neomycinresistance cDNA (Lneo). When LhuPAR⁺ cells were co-cultured with LhuPA cells, the plasmin-dependent fibrinolytic activity generated was more than that observed in either homotypic cultures with fivefold greater number of LhuPA cells or co-cultures containing LhuPA and Lneo cells instead of the LhuPAR⁺ cells. The preferential activation of TGF-β by co-cultures with the greatest plasmin-generation potential, LhuPAR⁺ and LhuPA cells, was confirmed by three independent bioassays. In the first assay, a 48% decrease in PA activity, a measure of active TGF-β production, was observed with BAE cells treated with conditioned medium (CM) from co-cultures of LhuPA and LhuPAR⁺ cells. Inclusion of neutralizing antibodies to TGF-β abrogated the inhibitory effect of CM on PA activity demonstrating that the inhibitory molecule was TGF-β. Addition of the amino terminal fragment of uPA (ATF) or omission of plasminogen from co-cultures blocked both the fibrinolytic activity and the generation of TGF-β activity in the CM. In the second assay, CM from co-cultures of LhuPA and LhuPAR⁺ cells inhibited the migration of BAE cells in a wound assay. Controls with anti-TGF-β IgG indicated that the inhibition was due to TGF-β. In the third assay, proliferation of mink lung epithelial cells was inhibited by CM generated by co-cultures of LhuPA and LhuPAR⁺ cells as compared to CM from the same cells cultured in the absence of plasminogen or to CM from a co-culture of LhuPA with LhuPAR^? cells. Excess mannose-6-phosphate (M6P) blocked the generation of TGF-β as assayed by both the BAE migration and PA assays, presumably because it interfered with cellsurface localization of LTGF-β. Additionally, small numbers of LhuPA and LhuPAR⁺ cells co-cultured with BAE cells inhibited the BAE cell PA activity via the paracrine action of TGF-β. These results support the conclusion that plasmindependent activation LTGF-β by LB6 cells is promoted by the surface localization of uPA by its receptor. © 1994 Wiley-Liss, Inc.     

Cell-associated activation of latent transforming growth factor-β by calpain

Transforming growth factor-β (TGF-β) is normally secreted in a latent form, and plasmin-mediated proteolytic cleavage of latency-associated peptide (LAP), a component of latent TGF-β complex that makes the complex inactive, activates latent TGF-β. In the present study, we investigated the possible involvement of calpain, one of the cysteine proteases, in the activation of latent TGF-β. When recombinant latent TGF-β was incubated with calpain (1–10 u/ml) in a test tube, calpain cleaved LAP and released mature TGF-β from the latent complex. When calpain was applied to cultured bovine capillary endothelial (BCE) cells, a low concentration of calpain (0.05–0.1 u/ml) inhibited the migration and proliferation of the cells, and these inhibitory effects were abrogated by anti-TGF-β antibody as well as by calpain inhibitor peptide, but not by α₂-antiplasmin, a specific inhibitor of plasmin. Active TGF-β was detected in the conditioned medium of BCE cells collected in the presence of calpain. Chemical cross-linking of ¹²⁵I-calpain to BCE cells followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis indicated that calpain bound to the cell surface through chondroitinase ABC-sensitive proteoglycan. In addition, treatment of the BCE cells with chondroitinase ABC abrogated the inhibitory effect of calpain on the migration of these cells. Our data thus suggest that calpain is able to activate latent TGF-β through a mechanism independent of plasmin. This activation is efficient in the presence of cells, and calpain binds to the cell surface via proteoglycan and activates latent TGF-β, which is targeted to the same surface. J. Cell. Physiol. 174:186–193, 1998. © 1998 Wiley-Liss, Inc.     

Effects of dexamethasone on the expression of transforming growth factor-β in mouse embryonic palatal mesenchymal cells

The central role of TGF-β in the development of the embryonic palate has been well characterized. TGF-β inhibits mesenchymal cell proliferation, induces medial edge epithelial cell differentiation, and modulates the expression of extracellular matrix proteins as well as the proteases that act upon them. Mechanisms by which TGF-β expression itself is regulated are less well understood. Glucocorticoids are recognized in several cellular systems as able to regulate the expression of TGF-β. This study was therefore designed to examine whether glucocorticoids affect the expression of TGF-β isoforms in embryonic palatal cells. Based on flow cytometric analysis and viability determination, confluent primary cultures of mouse embryonic palate mesenchymal (MEPM) cells exposed to up to 10⁻⁶ M dexamethasone (dex) exhibited no signs of cytotoxicity after 24 hours of exposure. Northern blot analyses revealed that dexamethasone reduced steady-state mRNA levels of TGF-β3 in a dose-dependent manner as early as 4 hours after treatment but had little effect on TGF-β1 and TGF-β2 expression up to 24 hours of dex exposure. Dex also reduced the synthesis of both latent and mature forms of TGF-β protein by approximately four-fold as determined by the mink lung epithelial cell growth inhibition bioassay. Assessment of the ratio of mature to latent protein found in conditioned medium of control compared to dex-treated cultures indicated that dexamethasone may reduce the activation of latent TGF-β to mature biologically active TGF-β. Dexamethasone inhibited the proliferation of MEPM cells despite the down-regulation of TGF-β suggesting that dex-induced growth inhibition of MEPM cells is not mediated by TGF-β. These data suggest that dex modulates TGF-β signaling pathways directly by down-regulating TGF-β expression and possibly indirectly by altering the availability of mature TGF-β necessary to exert its biological effects in the developing palate. © 1996 Wiley-Liss, Inc.     

TGF-β regulates the survival of ciliary ganglionic neurons synergistically with ciliary neurotrophic factor and neurotrophins

We investigated putative roles of transforming growth factor (TGF)-β expressed in peripheral ganglia in the regulation of neuronal cell survival during the period of ontogenetic neuron death (OD). The chick ciliary ganglion (CG), where OD occurs between embryonic days (E) 6 and 10, was employed as a model system. We show that CG neurons (E8) are immunoreactive (ir) for TGF-β2 and -β3 as well as the TGF-β receptor TβR-II, but are not ir for TGF-β1. Ciliary neurotrophic factor (CNTF) and fibroblast growth factor (FGF)-2, established neurotrophic molecules for CG neurons, up-regulate TGF-β3 mRNA and TGF-β biological activity in cultures of E8 CG neurons. None of the TGF-β isoforms—β1, β2, or β3—has a trophic, survival-promoting effect on cultured CG neurons. However, all isoforms enhance CG neuron survival mediated by CNTF or FGF-2, significantly and over a wide range of concentrations. In combination with the neurotrophins (NT) nerve growth factor (NGF) and NT-3, which are not neurotrophic for CG neurons, TGF-β significantly promotes CG neuron survival. However, TGF-β does not act synergistically with the neuropoietic cytokines oncostatin M, leukemia inhibiting factor, or interleukin-6. Immunoneutralization of endogenous TGF-β released from CG neurons using an antibody to TGF-β1/-β2/-β3 significantly reduces the potency of CNTF or FGF-2 to promote CG neuron survival. The blocking effect of the anti–pan-TGF-β antibody could be rescued by adding exogenous TGF-β. Together, these data suggest that para-/autocrine TGF-β signaling has an important effect on the regulation of neuron survival in a model system of peripheral neurons. © 1998 John Wiley & Sons, Inc. J Neurobiol 37: 563–572, 1998     

Role of receptor complexes in resistance or sensitivity to growth inhibition by TGFβ in intestinal epithelial cell clones

Untransformed rat intestinal epithelial cells (IEC-18) were chemically mutagenized, selected in the presence of TGFβ₁, and cloned by limiting dilution. Two clones (4–5, 4–6) were resistant to growth inhibition by both TGFβ₁ and TGFβ₂. Another clone (4–1) was more sensitive to both TGFβ isoforms (relative to parental IEC-18 cells). IC₅₀ values for TGFβ_{1 and 2} in the 4–1 cells were at least 1/9 those of the parental cells; growth rates were reduced by 49% for TGFβ₁ and by 26% for TGFβ₂ in this clone. This increased sensitivity to TGFβ was explained by the 5- to 10-fold increase, relative to parental cells, in binding of TGFβ₁ and TGFβ₂ to both the type I and II receptors. In contrast, the resistance to growth inhibition by TGFβ in the 4–5 and 4–6 cells could not be explained by a decrease in either TGFβ binding affinities or in total number of receptors expressed, by the presence of serum binding components, or by occupation of receptor binding sites with autocrine TGF-β₁. However, in comparison to TGFβ-sensitive cells (IEC-18, 4–1), the resistant cells displayed a higher ratio of type II relative to type I receptor binding by TGF-β₁. Thus, a critical ratio of binding to receptor subtypes correlated with growth inhibition by TGF-β₁. Resistance to TGF-β₂ in the same clones did not appear to be receptor related. Thus different mechanisms for resistance to TGF-β₁ and TGF-β₂ were observed within a given clone. © 1993 Wiley-Liss, Inc.     

Density-dependent inhibitory effect of transforming growth factor-β on human fibroblasts involves the down-regulation of platelet-derived growth factor α-receptors

We have previously found that transforming growth factor-β1 (TGF-β1) inhibits the mitogenic activity of platelet-derived growth factor (PDGF) in cultures of human neonatal fibroblasts in a density-dependent fashion. In the present investigation we determined the effect of TGF-β1 on the PDGF α-receptor, which binds all PDGF isoforms, as well as on the β-receptor, which binds only PDGF-BB with high affinity. We found that the inhibitory effect of TGF-β1 on PDGF-AA-induced mitogenesis was density-dependent; when dense cell cultures were preincubated with TGF-β1, there was an complete inhibition of ³H-thymidine incorporation, whereas the effect was less in sparse cultures. A similar density-dependent effect of TGF-β1 was seen in PDGF-BB treated cells, although less pronounced. The binding of ¹²⁵I-labeled PDGF-AA and PDGF-BB to the α-receptor was significantly reduced after treatment with TGF-β1 in dense cultures, whereas the sparse cultures were less affected. A decrease of α-receptor mRNA was also seen. The levels of β-receptor protein and mRNA were unaffected. We conclude that the growth inhibitory effect of TGF-β1 is cell density-dependent and involves down-regulation of PDGF α-receptors. © 1993 Wiley-Liss, Inc.     

Characterization of a 60-kDa cell surface-associated transforming growth factor-β binding protein that can interfere with transforming growth factor-β receptor binding

We have characterized a 60-kDa transforming growth factor-β (TGF-β) binding protein that was originally identified on LNCaP adenocarcinoma prostate cells by affinity cross-linking of cell surface proteins by using ¹²⁵I-TGF-β1. Binding of ¹²⁵I-TGF-β1 to the 60-kDa protein was competed by an excess of unlabeled TGF-β1 but not by TGF-β2, TGF-β3, activin, or osteogenic protein-1 (OP-1), also termed bone morphogenetic protein-7 (BMP-7). In addition, no binding of ¹²⁵I-TGF-β2 and ¹²⁵I-TGF-β3 to the 60-kDa binding protein on LNCaP cells could be demonstrated by using affinity labeling techniques. The 60-kDa TGF-β binding protein showed no immunoreactivity with antibodies against the known type I and type II receptors for members of the TGF-β superfamily. Treatment of LNCaP cells with 0.25 M NaCl, 1 μg/ml heparin, or 10% glycerol caused a release of the 60-kDa protein from the cell surface. In addition, we found that the previously described TGF-β type IV receptor on GH3 cells, which does not form a heteromeric complex with TGF-β receptors, could be released from the cell surface by these same treatments. This suggests that the 60-kDa protein and the similarly sized TGF-β type IV receptor are related proteins. The eluted 60-kDa LNCaP protein was shown to interfere with the binding of TGF-β to the TGF-β receptors. Thus, the cell surface-associated 60-kDa TGF-β binding protein may play a role in regulating TGF-β binding to TGF-β receptors. J. Cell. Physiol. 173:447–459, 1997. © 1997 Wiley-Liss, Inc.     

Expression,localization, and function of transforming growth factor-βs in embryonic chick spinal cord,hindbrain, and dorsal root ganglia

We have studied the localizations of transforming growth factor-beta (TGF-β) 2 and 3 immunohistochemically using isoform-specific antibodies and TGF-β3 mRNA by in situ hybridization in the nervous system of the 3- to 15-day-old chick embryo with special reference to spinal cord, hindbrain, and dorsal root ganglia (DRG). At embryonic day (E) 3, TGF-β3 mRNA as well as TGF-β2 and 3 immunoreactivities (IRs) were most prominent in the notochord, wall of the aorta, and dermomyotome. At E5 and E7, strong TGF-β2 and 3 IR were seen in or on radial glia of spinal cord and hindbrain. Radial glia in the floor plate region and ventral commissure gave the most intense signal. In the DRG, fiber strands of intense IRs representing extracellular matrix or satellite cells were seen. Neuronal perikarya did not become IR for TGF-β2 and 3 until E11, but even then the moderate signals for TGF-β3 mRNA could not be specifically localized to the neuronal cell bodies. In E11 and older embryos, spinal cord glial or glial progenitor cells, but not neuronal cell bodies were labeled for TGF-β3 mRNA. Immunocytochemistry and western blot analysis indicated that E8 DRG neurons have the TGF-β receptor type II, and treatment of these cells with NGF induces expression of TGF-β3 mRNA. The TGF-β isoforms 1, 2, and 3 did not promote survival of E8 DRG neurons in dissociated cell cultures. All three TGF-β isoforms, however, promoted neurite growth from E8 DRG explants, but were less potent than nerve growth factor. Our data suggest identical localizations of TGF-β2 and -β3 IR in the developing chick and mammalian nervous systems, underscoring the general importance of TGF-βs in fundamental events of neural development. © 1996 John Wiley & Sons, Inc.     

Protection of transforming growth factor β activity by heparin and fucoidan

The transforming growth factor-β (TGF-β) family of proteins exert diverse and potent effects on proliferation, differentiation, and extracellular matrix synthesis. However, relatively little is known about the stability or processing of endogenous TGF-β activity in vitro or in vivo. Our previous work indicated that (1) TGF-β1 has strong heparin-binding properties that were not previously recognized because of neutralization by iodination, and (2) heparin, and certain other polyanions, could block the binding of TGF-β1 to α₂-macroglobulin (α₂-M). The present studies investigated the influence of heparin-like molecules on the stability of the TGF-β1 signal in the pericellular environment. The results indicate that heparin and fucoidan, a naturally occurring sulfated L-fucose polymer, suppress the formation of an initial non-covalent interaction between ₁₂₅I-TGF-β1 and activated α₂-M. Electrophoresis of ₁₂₅I-TGF-β1 showed that fucoidan protects TGF-β1 from proteolytic degradation by plasmin and trypsin. While plasmin caused little, if any, activation of latent TGF-β derived from vascular smooth muscle cells (SMC), plasmin degraded acid-activated TGF-β, and purified TGF-β1, and this degradation was inhibited by fucoidan. In vitro, heparin and fucoidan tripled the half-life of ₁₂₅I-TGF-β1 and doubled the amount of cell-associated ₁₂₅I-TGF-β1. Consistent with this protective effect, heparin- and fucoidan-treated SMC demonstrated elevated levels of active, but not latent, TGF-β activity. © 1994 wiley-Liss, Inc.     

Role of the transforming growth factor-β family in the expression of cranial neural crest-specific phenotypes

Cranial and trunk neural crest cells produce different derivatives in vitro. Cranial neural crest cultures produce large numbers of cells expressing fibronectin (FN) and procollagen I (PCol I) immunoreactivities, two markers expressed by mesenchymal derivatives in vivo. Trunk neural crest cultures produce relatively few FN or PCol I immunoreactive cells, but they produce greater numbers of melanocytes than do cranial cultures. Treatment of trunk neural crest cultures with transforming growth factor-β1 (TGF-β1) stimulates them to express both FN and PCol I immunoreactivities at levels comparable to those normally seen in cranial cultures and simultaneously decreases their expression of melanin. These observations raised the possibility that endogenous TGF-β is involved in specifying differences in the phenotypes expressed by cranial and trunk neural crest cells in vitro. Consistent with this idea, we found that treatment of cranial cultures with a function-blocking TGF-β antiserum inhibits the development of FN immunoreactive cells and stimulates the development of melanocytes. Cranial and trunk neural crest cells express approximately equal levels of TGF-β mRNA. However, trunk neural crest cells are significantly less sensitive to the FN-inducing effect of TGF-β1 than are cranial neural crest cells. These results suggest that: (1) endogenous TGF-β is required for the expression of mesenchymal phenotypes by cranial neural crest cells, and (2) differences in the phenotypes expressed by cranial and trunk neural crest cells in vitro result in part from differences in the sensitivities of these two cell populations to TGF-β. © 1995 John Wiley & Sons, Inc.     

Differential regulation of p34cdc2 and p33cdk2 by transforming growth factor-β1 in murine mammary epithelial cells

Cyclin-dependent kinases (cdks) are a family of proteins whose function plays a critical role in cell cycle traverse. Transforming growth factor-β₁ (TGF-β₁) is a potent growth inhibitor of epithelial cells. Since cdks have been suggested as possible biochemical markers for TGF-β growth inhibition, we investigated the effect of TGF-β₁ on cdc2 and cdk2 in a normal mouse mammary epithelial cell line (MME) and a TGF-β-resistant MME cell line (BG18.2). TGF-β₁ decreases newly synthesized cdc2 protein levels within 6 h after addition. Coincident with this decrease in newly synthesized cdc2 protein was a marked reduction in its ability to phosphorylate histone H1. This decrease in kinase activity is not due to a change in steady-state levels of cdc2 protein, since mRNA and total protein levels of cdc2 are not reduced until 12 h after TGF-β₁ addition. This suggests that the kinase activity of cdc2 is dependent on newly synthesized cdc2 protien. Moreover, the protein synthesis of another cyclin-dependent kinase, cdk2, is not effected by TGF-β₁ addition, but its kinase activity is substantially reduced. Thus, it appears that TGF-β decreases the kinase activity of both cdc2 and cdk2 by distinct mechanisms.