Three-dimensional tissue structure affects sensitivity of fibroblasts to TGF-beta 1

Transforming growth factor-(TGF-) is known to induce -smooth muscle actin (-SMA) infibroblasts and is supposed to play a role in myofibroblastdifferentiation and tumor desmoplasia. Our objective was to elucidatethe impact of TGF-1 on -SMA expression in fibroblasts in athree-dimensional (3-D) vs. two-dimensional (2-D) environment. Inmonolayer culture, all fibroblast cultures responded in a similarfashion to TGF-1 with regard to -SMA expression. In fibroblastspheroids, -SMA expression was reduced and induction by TGF-1 washighly variable. This difference correlated with a differentialregulation in the TGF- receptor (TGFR) expression, in particularwith a reduction in TGF-RII in part of the fibroblast types. Ourdata indicate that 1) sensitivity to TGF-1-induced -SMA expression in a 3-D environment is fibroblast-type specific, 2) fibroblast type-independent regulatory mechanisms, suchas a general reduction/loss in TGF-RIII, contribute to an altered TGFR expression profile in spheroid compared with monolayer culture, and 3) fibroblast type-specific alterations in TGFR typesI and II determine the sensitivity to TGF-1-induced -SMAexpression in the 3-D setting. We suggest that fibroblasts that can beinduced by TGF-1 to produce -SMA in spheroid culture reflect a"premyofibroblastic" phenotype.

     

TGF-beta effects on epithelial ion transport and barrier: reduced Cl- secretion blocked by a p38 MAPK inhibitor

Growthfactors affect a variety of epithelial functions. We examined theability of TGF- to modulate epithelial ion transport andpermeability. Filter-grown monolayers of human colonic epithelia, T84and HT-29 cells, were treated with TGF- (0.1-100 ng/ml,15 min-72 h) or infected with an adenoviral vector encodingTGF- (Ad-TGF) for 144 h. Ion transport (i.e., short-circuitcurrent, I_sc) and transepithelial resistance(TER) were assessed in Ussing chambers. Neither recombinant TGF- norAd-TGF infection affected baseline I_sc;however, exposure to 1 ng/ml TGF- led to a significant (30-50%) reduction in the I_sc responses toforskolin, vasoactive intestinal peptide, and cholera toxin (agentsthat evoke Cl secretion via cAMP mobilization) and to thecell-permeant dibutyryl cAMP. Pharmacological analysis of signalingpathways revealed that the inhibition of cAMP-driven epithelialCl secretion by TGF- was blocked by pretreatment withSB-203580, a specific inhibitor of p38 MAPK, but not by inhibitors ofJNK, ERK1/2 MAPK, or phosphatidylinositol 3'-kinase. TGF- enhanced the barrier function of the treated monolayers by up to threefold asassessed by TER; however, this event was temporally displaced from thealtered I_sc response, being statisticallysignificant only at 72 h posttreatment. Thus, in addition toTGF- promotion of epithelial barrier function, we show that thisgrowth factor also reduces responsiveness to cAMP-dependentsecretagogues in a chronic manner and speculate that this serves as abraking mechanism to limit secretory enteropathies.

     

Differential expression of TGF-beta isoforms by normal and inflammatory bowel disease intestinal myofibroblasts

First publishedSeptember 5, 2001; 10.1152/ajpcell. 00048.2001.Intestinalstrictures are frequent in Crohn's disease but not ulcerative colitis.We investigated the expression of transforming growth factor (TGF)-isoforms by isolated and cultured primary human intestinalmyofibroblasts and the responsiveness of these cells and intestinalepithelial cells to TGF- isoforms. Normal intestinal myofibroblastsreleased predominantly TGF-₃ and ulcerative colitismyofibroblasts expressed both TGF-₁ andTGF-₃, whereas in myofibroblast cultures from fibroticCrohn's disease tissue, there was significantly lower expression ofTGF-₃ but enhanced release of TGF-₂.These distinctive patterns of TGF- isoform release were sustainedthrough several myofibroblast passages. Proliferation of Crohn'sdisease myofibroblasts was significantly greater than that ofmyofibroblasts derived from normal and ulcerative colitis tissue. Incontrast to cells from normal and ulcerative colitis tissue,neutralization of the three TGF- isoforms did not affect theproliferation of Crohn's disease intestinal myofibroblasts. Studies onthe effect of recombinant TGF- isoforms on epithelial restitutionand proliferation suggest that TGF-₂ may be the least effective of the three isoforms in intestinal wound repair. In conclusion, the enhanced release of TGF-₂ but reducedexpression of TGF-₃ by Crohn's disease intestinalmyofibroblasts, together with their enhanced proliferative capacity,may lead to the development of intestinal strictures.

     

Identification of mediators stimulating proliferation and matrix synthesis of rat pancreatic stellate cells

The aim of thisstudy was to identify fibrogenic mediators stimulatingactivation, proliferation, and/or matrix synthesis of rat pancreaticstellate cells (PSC). PSC were isolated from the pancreas of normalWistar rats and from rats with cerulein pancreatitis. Cell activationwas demonstrated by immunofluorescence microscopy of smooth muscle-actin (SMA) and real-time quantitative RT-PCR of SMA, fibronectin,and transforming growth factor (TGF)-₁. Proliferationwas measured by bromodeoxyuridine incorporation. Matrix synthesis wasdemonstrated on the protein and mRNA level. Within a few days inprimary culture, PSC changed their phenotype from fat-storing toSMA-positive myofibroblast-like cells expressing platelet-derivedgrowth factor (PDGF) - and PDGF -receptors. TGF-₁and tumor necrosis factor (TNF)- accelerated the change in thecells' phenotype. Addition of 50 ng/ml PDGF and 5 ng/ml basicfibroblast growth factor (bFGF) to cultured PSC significantly stimulated cell proliferation (4.37 ± 0.49- and 2.96 ± 0.39-fold of control). Fibronectin synthesis calculated on the basis of DNA was stimulated by 5 ng/ml bFGF (3.44 ± 1.13-fold), 5 ng/ml TGF-₁ (2.46 ± 0.89-fold), 20 ng/ml PDGF (2.27 ± 0.68-fold), and 50 ng/ml TGF- (1.87 ± 0.19-fold). As shownby RT-PCR, PSC express predominantly the splice variant EIII-A offibronectin. Immunofluorescence microscopy and Northern blot confirmedthat in particular bFGF and TGF-₁ stimulated thesynthesis of fibronectin and collagens type I and III. In conclusion,our data demonstrate that 1) TGF-₁ andTNF- accelerate the change in the cell phenotype, 2) PDGF represents the most effective mitogen, and 3) bFGF,TGF-₁, PDGF, and, to a lesser extent, TGF- stimulateextracellular matrix synthesis of cultured rat PSC.

     

Rapid and dynamic regulation of TGF-beta receptors on blood vessels and fibroblasts during ischemia-reperfusion injury

Thepathophysiological mechanisms involved in ischemia-reperfusioninjury are poorly understood. Although transforming growth factor(TGF)- has been shown to provide protection againstischemia-reperfusion injury in different organ systems, littleis known about the regulation of TGF- action during this process.Here we analyzed the effect of ischemia and reperfusion on theexpression of TGF- and its receptors in vivo with a pig skin flapmodel. Analysis of unoperated skin, nonischemic control flap,ischemic flap, and reperfused flap by immunohistochemistryindicates that ischemia and reperfusion result in rapid anddynamic regulation of type I, II, and III TGF- receptors andTGF-1 in a cell type-specific manner. Furthermore, hypoxiaupregulates type II TGF- receptor mRNA in skin fibroblasts inculture. Together, our results reveal that TGF- receptors andTGF-1 are markedly increased under acute ischemic conditions in the blood vessels and fibroblasts of the skin. We conclude thatTGF- action is enhanced under ischemic conditions and that it may represent an adaptive response to ischemic injury. The augmented TGF- responsiveness may be a critical determinant of theprotective effect of TGF- during ischemia-reperfusion injury.

     

Expression of the TGF-beta receptor gene and sensitivity to growth inhibition following polyamine depletion

Our previous studieshave shown that inhibition of polyamine biosynthesis increases thesensitivity of intestinal epithelial cells to growth inhibition inducedby exogenous transforming growth factor- (TGF-). This study wentfurther to determine whether expression of the TGF- receptor genesis involved in this process. Studies were conducted in the IEC-6 cellline, derived from rat small intestinal crypt cells. Administration of-difluoromethylornithine (DFMO), a specific inhibitor of ornithinedecarboxylase (the rate-limiting enzyme for polyamine synthesis), for 4 and 6 days depleted cellular polyamines putrescine, spermidine, andspermine in IEC-6 cells. Polyamine depletion by DFMO increased levelsof the TGF- type I receptor (TGF-RI) mRNA and protein but had noeffect on the TGF- type II receptor expression. The inducedTGF-RI expression after polyamine depletion was associated with anincreased sensitivity to growth inhibition induced by exogenous TGF-but not by somatostatin. Extracellular matrix laminin inhibited IEC-6cell growth without affecting the TGF- receptor expression. Lamininconsistently failed to induce the sensitivity of TGF--mediatedgrowth inhibition. In addition, decreasing TGF-RI expression bytreatment with retinoic acid not only decreased TGF--mediated growthinhibition in normal cells but also prevented the increased sensitivityto exogenous TGF- in polyamine-deficient cells. These resultsindicate that 1) depletion of cellular polyamines by DFMOincreases expression of the TGF-RI gene and 2) increasedTGF-RI expression plays an important role in the process throughwhich polyamine depletion sensitizes intestinal epithelial cells togrowth inhibition induced by TGF-.

     

Ouabain and substrate affinities of human Na(+)-K(+)-ATPase alpha(1)beta(1), alpha(2)beta(1), and alpha(3)beta(1) when expressed separately in yeast cells

HumanNa⁺-K⁺-ATPase₁₁,₂₁, and ₃₁heterodimers were expressed individually in yeast, and ouabainbinding and ATP hydrolysis were measured in membrane fractions. Theouabain equilibrium dissociation constant was 13-17 nM for₁₁ and ₃₁at 37°C and 32 nM for ₂₁, indicatingthat the human -subunit isoforms have a similar high affinity forcardiac glycosides. K_0.5 values for antagonism of ouabain binding by K⁺ were ranked in order as follows:₂ (6.3 ± 2.4 mM) > ₃(1.6 ± 0.5 mM)  ₁ (0.9 ± 0.6 mM),and K_0.5 values for Na⁺ antagonismof ouabain binding to all heterodimers were 9.5-13.8 mM. Themolecular turnover for ATP hydrolysis by₁₁ (6,652 min¹) was abouttwice as high as that by ₃₁ (3,145 min¹). These properties of the human heterodimersexpressed in yeast are in good agreement with properties of the humanNa⁺-K⁺-ATPase expressed in Xenopusoocytes (G Crambert, U Hasler, AT Beggah, C Yu, NN Modyanov, J-DHorisberger, L Lelievie, and K Geering. J Biol Chem275: 1976-1986, 2000). In contrast to Na⁺ pumpsexpressed in Xenopus oocytes, the₂₁ complex in yeast membranes wassignificantly less stable than ₁₁ or₃₁, resulting in a lower functionalexpression level. The ₂₁ complex was also more easily denatured by SDS than was the₁₁ or the₃₁ complex.

     

The gamma-subunit of ENaC is more important for channel surface expression than the beta-subunit

The amiloride-sensitiveepithelial sodium channel (ENaC) plays a critical role in fluid andelectrolyte homeostasis and is composed of three homologous subunits:, , and . Only heteromultimeric channels made of ENaCare efficiently expressed at the cell surface, resulting in maximallyamiloride-sensitive currents. To study the relative importance ofvarious regions of the - and -subunits for the expression offunctional ENaC channels at the cell surface, we constructedhemagglutinin (HA)-tagged --chimeric subunits composed of -and -subunit regions and coexpressed them with HA-tagged - and-subunits in Xenopus laevis oocytes. The whole cellamiloride-sensitive sodium current (I_ami) andsurface expression of channels were assessed in parallel using thetwo-electrode voltage-clamp technique and a chemiluminescence assay.Because coexpression of ENaC resulted in largerI_ami and surface expression compared withcoexpression of ENaC, we hypothesized that the -subunit ismore important for ENaC trafficking than the -subunit. Usingchimeras, we demonstrated that channel activity is largely preservedwhen the highly conserved second cysteine rich domains (CRD2) of the- and -subunits are exchanged. In contrast, exchanging the wholeextracellular loops of the - and the -subunits largely reducedENaC currents and ENaC expression in the membrane. This indicates thatthere is limited interchangeability between molecular regions of thetwo subunits. Interestingly, our chimera studies demonstrated that theintracellular termini and the two transmembrane domains of ENaC aremore important for the expression of functional channels at the cellsurface than the corresponding regions of ENaC.

     

Transforming growth factor-beta 1 modulates the expression of vascular endothelial growth factor by osteoblasts

Angiogenesis is essential to both normal and pathological bonephysiology. Vascular endothelial growth factor (VEGF) has been implicated in angiogenesis, whereas transforming growth factor-1 (TGF-1) modulates bone differentiation, matrixformation, and cytokine expression. The purpose of this study was toinvestigate the relationship between TGF-1 and VEGF expression inosteoblasts and osteoblast-like cells. Northern blot analysis revealedan early peak of VEGF mRNA (6-fold at 3 h) in fetal rat calvarial cellsand MC3T3-E1 osteoblast-like cells after stimulation with TGF-1 (2.5 ng/ml). The stability of VEGF mRNA in MC3T3-E1 cells was not increasedafter TGF-1 treatment. Actinomycin D inhibited the TGF-1-inducedpeak in VEGF mRNA, whereas cycloheximide did not. Blockade of TGF-1signal transduction via a dominant-negative receptor II adenovirussignificantly decreased TGF-1 induction of VEGF mRNA. Additionally,TGF-1 induced a dose-dependent increase in VEGF protein expressionby MC3T3-E1 cells (P < 0.01).Dexamethasone similarly inhibited VEGF protein expression. BothTGF-1 mRNA and VEGF mRNA were concurrently present in rat membranousbone, and both followed similar patterns of expression during ratmandibular fracture healing (mRNA and protein). In summary,TGF-1-induced VEGF expression by osteoblasts and osteoblast-likecells is a dose-dependent event that may be intimately related to bonedevelopment and fracture healing.

     

Evidence for functional role of epsilonPKC isozyme in the regulation of cardiac Na(+) channels

Investigation of the role ofindividual protein kinase C (PKC) isozymes in the regulation ofNa⁺ channels has been largely limited by the lack ofisozyme-selective modulators. Here we used a novel peptide-specificactivator (V1-7) of PKC and other peptide isozyme-specificinhibitors in addition to the general PKC activator phorbol12-myristate 13-acetate (PMA) to dissect the role of individual PKCs inthe regulation of the human cardiac Na⁺ channel hH1,heterologously expressed in Xenopus oocytes. Peptides wereinjected individually or in combination into the oocyte. Whole cellNa⁺ current (I_Na) was recorded usingtwo-electrode voltage clamp. V1-7 (100 nM) and PMA (100 nM)inhibited I_Na by 31 ± 5% and 44 ± 8% (at 20 mV), respectively. These effects were not seen with thescrambled peptide for V1-7 (100 nM) or the PMA analog4-phorbol 12,13-didecanoate (100 nM). However, V1-7-and PMA-induced I_Na inhibition was abolished byV1-2, a peptide-specific antagonist of PKC. Furthermore,PMA-induced I_Na inhibition was not altered by100 nM peptide-specific inhibitors for -, -, -, or PKC. PMAand V1-7 induced translocation of PKC from soluble toparticulate fraction in Xenopus oocytes. This translocationwas antagonized by V1-2. In native rat ventricular myocytes,PMA and V1-7 also inhibited I_Na; thisinhibition was antagonized by V1-2. In conclusion, the resultsprovide evidence for selective regulation of cardiac Na⁺channels by PKC isozyme.

     

Upregulation of alpha(8)beta(1)-integrin in cardiac fibroblast by angiotensin II and transforming growth factor-beta1

Using a novel pharmacological tool with¹²⁵I-echistatin to detect integrins on the cell, we haveobserved that cardiac fibroblasts harbor five different RGD-bindingintegrins: ₈₁,₃₁, ₅₁, _v1, and _v3.Stimulation of cardiac fibroblasts by angiotensin II (ANG II) ortransforming growth factor-1 (TGF-1) resulted in an increase ofprotein and heightening by 50% of the receptor density of₈₁-integrin. The effect of ANG II wasblocked by an AT₁, but not an AT₂, receptorantagonist, or by an anti-TGF-1 antibody. ANG II and TGF-1increased fibronectin secretion, smooth muscle -actin synthesis, andformation of actin stress fibers and enhanced attachment of fibroblaststo a fibronectin matrix. The ₈- and₁-subunits were colocalized by immunocytochemistry with vinculin or ₃-integrin at focal adhesion sites.These results indicate that ₈₁-integrinis an abundant integrin on rat cardiac fibroblasts. Its positivemodulation by ANG II and TGF-1 in a myofibroblast-likephenotype suggests the involvement of₈₁-integrin in extracellularmatrix protein deposition and cardiac fibroblast adhesion.

     

p38 MAPK mediates renal tubular cell TNF-{alpha} production and TNF-{alpha}-dependent apoptosis during simulated ischemia

Ischemia causes renal tubular cellloss through apoptosis; however, the mechanisms of this processremain unclear. Using the renal tubular epithelial cell lineLLC-PK₁, we developed a model of simulated ischemia(SI) to investigate the role of p38 MAPK (mitogen-activated proteinkinase) in renal cell tumor necrosis factor- (TNF-) mRNAproduction, protein bioactivity, and apoptosis. Resultsdemonstrate that 60 min of SI induced maximal TNF- mRNA productionand bioactivity. Furthermore, 60 min of ischemia induced renaltubular cell apoptosis at all substrate replacement time pointsexamined, with peak apoptotic cell death occurring after either 24 or 48 h. p38 MAPK inhibition abolished TNF- mRNA production andTNF- bioactivity, and both p38 MAPK inhibition and TNF- neutralization (anti-porcine TNF- antibody) preventedapoptosis after 60 min of SI. These results constitute theinitial demonstration that 1) renal tubular cells produceTNF- mRNA and biologically active TNF- and undergoapoptosis in response to SI, and 2) p38 MAPKmediates renal tubular cell TNF- production and TNF--dependent apoptosis after SI.

     

Effect of interleukin-1beta and tumor necrosis factor-alpha on gene expression in human endothelial cells

Interleukin-1(IL-1) and tumor necrosis factor- (TNF-) are two majorcytokines that rise to relatively high levels during systemicinflammation, and the endothelial cell (EC) response to these cytokinesmay explain some of the dysfunction that occurs. To better understandthe cytokine-induced responses of EC at the gene expression level,human umbilical vein EC were exposed to IL-1 or TNF- for varioustimes and subjected to cDNA microarray analyses to study alterations intheir mRNA expression. Of ~4,000 genes on the microarray, expressionlevels of 33 and 58 genes appeared to be affected by treatment withIL-1 and TNF-, respectively; 25 of these genes responded to bothtreatments. These results suggest that the effects of IL-1 andTNF- on EC are redundant and that it may be necessary to suppressboth cytokines simultaneously to ameliorate the systemic response.

     

IkappaBalpha-dependent regulation of low-shear flow-induced NF-kappa B activity: role of nitric oxide

We have investigated the role ofinhibitor B (IB) in the activation of nuclear factor B(NF-B) observed in human aortic endothelial cells (HAEC) undergoinga low shear stress of 2 dynes/cm². Low shear for 6 hresulted in a reduction of IB levels, an activation of NF-B,and an increase in B-dependent vascular cell adhesion molecule 1 (VCAM-1) mRNA expression and endothelial-monocyte adhesion.Overexpression of IB in HAEC attenuated all of these shear-induced responses. These results suggest that downregulation ofIB is the major factor in the low shear-induced activation ofNF-B in HAEC. We then investigated the role of nitric oxide (NO) inthe regulation of IB/NF-B. Overexpression of endothelial nitric oxide synthase (eNOS) inhibited NF-B activation in HAEC exposed to 6 h of low shear stress. Addition of the structurally unrelated NO donors S-nitrosoglutathione (300 µM) orsodium nitroprusside (1 mM) before low shear stress significantlyincreased cytoplasmic IB and concomitantly reduced NF-Bbinding activity and B-dependent VCAM-1 promoter activity. Together,these data suggest that NO may play a major role in the regulation ofIB levels in HAEC and that the application of low shear flowincreases NF-B activity by attenuating NO generation and thusIB levels.

     

Protein kinase Calpha participates in activation of store-operated Ca2+ channels in human glomerular mesangial cells

Protein kinase C (PKC) plays animportant role in activating store-operated Ca²⁺ channels(SOC) in human mesangial cells (MC). The present study was performed todetermine the specific isoform(s) of conventional PKC involved inactivating SOC in MC. Fura 2 fluorescence ratiometry showed that thethapsigargin-induced Ca²⁺ entry (equivalent to SOC) wassignificantly inhibited by 1 µM Gö-6976 (a specific PKC andI inhibitor) and PKC antisense treatment (2.5 nM for 24-48h). However, LY-379196 (PKC inhibitor) and2,2',3,3',4,4'-hexahydroxy-1,1'-biphenyl-6,6'-dimethanoldimethyl ether(HBDDE; PKC and  inhibitor) failed to affect thapsigargin-evoked activation of SOC. Single-channel analysis in the cell-attached configuration revealed that Gö-6976 and PKC antisensesignificantly depressed thapsigargin-induced activation of SOC.However, LY-379196 and HBDDE did not affect the SOC responses. Ininside-out patches, application of purified PKC or I, but notII or , significantly rescued SOC from postexcision rundown.Western blot analysis revealed that thapsigargin evoked a decrease incytosolic expression with a corresponding increase in membraneexpression of PKC and . However, the translocation from cytosolto membranes was not detected for PKCI or II. These resultssuggest that PKC participates in the intracellular signaling pathwayfor activating SOC upon release of intracellular stores ofCa²⁺.

     

Laminins and TGF-beta maintain cell polarity and functionality of human gastric glandular epithelium

The human gastric glandularepithelium produces a gastric lipase enzyme (HGL) that plays animportant role in digestion of dietary triglycerides. To assess theinvolvement of extracellular matrix components and transforming growthfactor-1 (TGF-1) in the regulation of this enzymic function,normal gastric epithelial cells were cultured on collagen type I,Matrigel, and laminins (LN)-1 and -2 with or without TGF-1.Epithelial morphology and HGL expression were evaluated usingmicroscopy techniques, enzymic assays, Western blot, Northernhybridization, and RT-PCR. A correlation was observed between the cellpolarity status and the level of HGL expression. TGF-1 alone orindividual matrix components stimulated cell spreading and caused adownfall of HGL activity and mRNA. By contrast, Matrigel preserved themorphological features of differentiated epithelial cells andmaintained HGL expression. The combination of LNs with TGF-1 (twoconstituents of Matrigel) exerted similar beneficial effects onepithelial cell polarity and evoked a 10-fold increase of HGL levelsthat was blunted by a neutralizing antibody against the₂-integrin subunit and by mitogen-activated proteinkinase (MAPK) inhibitors PD-98059 (p42/p44) or SB-203580 (p38). Thisinvestigation demonstrates for the first time that a powerful synergismbetween a growth factor and basement membrane LNs positively influencescell polarity and functionality of the human gastric glandularepithelium through an activation of the₂₁-integrin and effectors of two MAPK pathways.

     

Functional overload increases beta-MHC promoter activity in rodent fast muscle via the proximal MCAT (betae3) site

Functional overload (OL)of the rat plantaris muscle by the removal of synergistic musclesinduces a shift in the myosin heavy chain (MHC) isoform expressionprofile from the fast isoforms toward the slow type I, or, -MHCisoform. Different length rat -MHC promoters were linked to afirefly luciferase reporter gene and injected in control and OLplantaris muscles. Reporter activities of 3,500, 914, 408, and215 bp promoters increased in response to 1 wk of OL. The smallest171 bp promoter was not responsive to OL. Mutation analyses ofputative regulatory elements within the 171 and 408 bp region wereperformed. The 408 bp promoters containing mutations of the e1,distal muscle CAT (MCAT; e2), CACC, or A/T-rich (GATA), were stillresponsive to OL. Only the proximal MCAT (e3) mutation abolished theOL response. Gel mobility shift assays revealed a significantly higherlevel of complex formation of the e3 probe with nuclear protein fromOL plantaris compared with control plantaris. These results suggestthat the e3 site functions as a putative OL-responsive element inthe rat -MHC gene promoter.

     

GSK-3beta negatively regulates skeletal myotube hypertrophy

Todetermine whether changes in glycogen synthase kinase-3 (GSK-3)phosphorylation contribute to muscle hypertrophy, we delineated theeffects of GSK-3 activity on C₂C₁₂ myotubesize. We also examined possible insulin-like growth factor I (IGF-I) signaling of NFAT (nuclear factors of activated T cells)-inducible geneactivity and possible modulation of NFAT activation by GSK-3. Application of IGF-I (250 ng/ml) or LiCl (10 mM) alone (i.e., bothinhibit GSK-3 activity) increased the area ofC₂C₁₂ myotubes by 80 and 85%, respectively.The application of IGF-I (250 ng/ml) elevated GSK-3 phosphorylationand reduced GSK-3 kinase activity by ~800% and ~25%,respectively. LY-294002 (100 µM) and wortmannin (150 µM), specificinhibitors of phosphatidylinositol 3'-kinase, attenuated IGF-I-inducedGSK-3 phosphorylation by 67 and 92%, respectively. IGF-I suppressedthe kinase activity of GSK-3. IGF-I (250 ng/ml), but not LiCl (10 mM), induced an increase in NFAT-activated luciferase reporteractivity. Cotransfection of a constitutively active GSK-3(cGSK-3) inhibited the induction by IGF-I of NFAT-inducible reporteractivity. LiCl, which inhibits GSK-3, removed the block by cGSK-3on IGF-I-inducible NFAT-responsive reporter gene activity. These datasuggest that the IGF-I-induced increase in skeletal myotube size issignaled, in part, through the inhibition of GSK-3.

     

TIMAP,a novel CAAX box protein regulated by TGF-beta1 and expressed in endothelial cells

Representational difference analysis ofthe glomerular endothelial cell response to transforming growthfactor-1 (TGF-1) revealed a novel gene, TIMAP (TGF--inhibitedmembrane-associated protein), which contains 10 exons and maps to humanchromosome 20.q11.22. By Northern blot, TIMAP mRNA is highly expressedin all cultured endothelial and hematopoietic cells. The frequency ofthe TIMAP SAGE tag is much greater in endothelial cell SAGE databasesthan in nonendothelial cells. Immunofluorescence studies of rat tissuesshow that anti-TIMAP antibodies localize to vascular endothelium.TGF-1 represses TIMAP through a protein synthesis- and histonedeacetylase-dependent process. The TIMAP protein contains five ankyrinrepeats, a protein phosphatase-1 (PP1)-interacting domain, aCOOH-terminal CAAX box, a domain arrangement similar to that of MYPT3,and a PP1 inhibitor. A green fluorescent protein-TIMAP fusion proteinlocalized to the plasma membrane in a CAAX box-dependent fashion.Hence, TIMAP is a novel gene highly expressed in endothelial andhematopoietic cells and regulated by TGF-1. On the basis of itsdomain structure, TIMAP may serve a signaling function, potentiallythrough interaction with PP1.

     

Differential distribution of ERalpha and ERbeta mRNA in intrauterine tissues of the pregnant rhesus monkey

Twoestrogen receptor (ER) isoforms, ER and ER, have been described.However, no information is available in any species regarding thecomparison of ER and ER levels in pregnant intrauterine tissues.We investigated 1) distribution of ER and ER mRNA in myometrium, amnion, choriodecidua, and placenta; 2) theirabundance in intrauterine tissues at term not in labor (NIL) and inspontaneous term labor (STL); and 3) immunolocalization ofER and ER in pregnant rhesus monkey myometrium. Myometrium,amnion, choriodecidua, and placenta were obtained at cesarean sectionfrom monkeys in STL at 156-166 days gestational age(GA) (n = 4) and from control monkeys NIL at140-152 days GA (n = 4). RT-PCR was conducted to determineER and ER and glyceraldehyde-3-phosphate dehydrogenase mRNAabundance in four intrauterine tissues of the pregnant rhesus monkey.The cloned ER PCR fragment was subjected to sequence analysis. ERand ER were localized in the myometrium by immunohistochemistry. Wedemonstrated that 1) rhesus monkey ER shares >97%identity with human ER in the region sequenced; 2) both ERswere expressed in myometrium, amnion, and choriodecidua but not inplacenta in the current study; 3) ER and ER weredifferentially distributed in myometrium and amnion; 4) ERand ER were immunolocalized in myometrial smooth cells and smoothmuscle and endothelial cells of the myometrial blood vessels. Thebiological significance of these quantitative differences in ERsubtypes merits further study.