In vivo effects of TGFbeta1 on the growth of gastric epithelium in suckling rats

As the content of Transforming Growth Factor-beta (TGFbeta) wanes in the milk of lactating rat, an increase in TGFbeta is observed in the gastric epithelia concomitant with differentiation of the glands upon weaning. Whereas TGFbeta has been shown to inhibit the proliferation of gastrointestinal cells in vitro, its functional significance and mechanisms of action have not been studied in vivo. Therefore, we administered TGFbeta1 (1 ng/g body wt.) to 14-day-old rats in which the gastric epithelium was induced to proliferate by fasting, and determined the involvement of signaling through Smads and the impact on epithelial cell proliferation and apoptosis. After the gavage, we observed the progressive increase of active TGFbeta1 while TbetaRII-receptor remained constant in the gastric mucosa. By immunohistochemistry, we showed Smad2/3 increase at 60 min (p<0.05) and Smad2 phosphorylation/activation and translocation to the nucleus most prominently between 0 and 30 min after treatment (p<0.05). Importantly, TGFbeta1 inhibited cell proliferation (p<0.05), which was estimated by BrDU pulse-labeling 12 h after gavage. Lower proliferation was reflected by increased p27(kip1) at 2 h (p<0.05). Also, TGFbeta1 increased apoptosis as measured by M30 labeling at 60 and 180 min (p<0.001), and by morphological features at 12 h (p<0.05). In addition, we observed higher levels of activated caspase 3 (17 kDa) from 0 to 30 min. Altogether, these data indicate a direct effect of TGFbeta1 signaling through Smads on both inhibiting proliferation, through alteration of cycle proteins, and inducing apoptosis of gastric epithelial cells in vivo. Further, the studies suggest a potential role for both milk and tissue-expressed TGFbeta1 in gastric growth during postnatal development.     

Transforming Growth Factor β1 Increases p27 Levels via Synthesis and Degradation Mechanisms in the Hyperproliferative Gastric Epithelium in Rats

Throughout postnatal development, the gastric epithelium expresses Transforming Growth Factor beta1 (TGFβ1), but it is also exposed to luminal peptides that are part of milk. During suckling period, fasting promotes the withdrawal of milk-born molecules while it stimulates gastric epithelial cell proliferation. Such response can be reversed by exogenous TGFβ1, as it directly affects cell cycle through the regulation of p27 levels. We used fasting condition to induce the hyperproliferation of gastric epithelial cells in 14-day-old Wistar rats, and evaluated the effects of TGFβ1 gavage on p27 expression, phosphorylation at threonine 187 (phospho-p27^Thr187) and degradation. p27 protein level was reduced during fasting when compared to suckling counterparts, while phospho-p27^Thr187/p27 ratio was increased. TGFβ1 gavage reversed this response, which was confirmed through immunostaining. By using a neutralizing antibody against TGFβ1, we found that it restored the p27 and phosphorylation levels detected during fasting, indicating the specific role of the growth factor. We noted that neither fasting nor TGFβ1 changed p27 expression, but after cycloheximide administration, we observed that protein synthesis was influenced by TGFβ1. Next, we evaluated the capacity of the gastric mucosa to degrade p27 and we recorded a higher concentration of the remaining protein in pups treated with TGFβ1, suggesting augmented stability under this condition. Thus, we showed for the first time that luminal TGFβ1 increased p27 levels in the rat gastric mucosa by up- regulating translation and reducing protein degradation. We concluded that such mechanisms might be used by rapidly proliferating cells to respond to milk-born TGFβ1 and food restriction.     

Expression and localization of Ski determine cell type-specific TGFbeta signaling effects on the cell cycle

Transforming growth factor beta (TGFbeta) promotes epithelial cell differentiation but induces Schwann cell proliferation. We show that the protooncogene Ski (Sloan-Kettering viral oncogene homologue) is an important regulator of these effects. TGFbeta down-regulates Ski in epithelial cells but not in Schwann cells. In Schwann cells but not in epithelial cells, retinoblastoma protein (Rb) is up-regulated by TGFbeta. Additionally, both Ski and Rb move to the cytoplasm, where they partially colocalize. In vivo, Ski and phospho-Rb (pRb) appear to interact in the Schwann cell cytoplasm of developing sciatic nerves. Ski overexpression induces Rb hyperphosphorylation, proliferation, and colocalization of both proteins in Schwann cell and epithelial cell cytoplasms independently of TGFbeta treatment. Conversely, Ski knockdown in Schwann cells blocks TGFbeta-induced proliferation and pRb cytoplasmic relocalization. Our findings reveal a critical function of fine-tuned Ski levels in the control of TGFbeta effects on the cell cycle and suggest that at least a part of Ski regulatory effects on TGFbeta-induced proliferation of Schwann cells is caused by its concerted action with Rb.     

The postnatal age of rat lung fibroblasts influences G1/S phase transition in vitro

Summary In the neonatal rat lung, alveolar development occurs from postnatal Days 4–13, during which time there is a fourfold increase in interstitial fibroblasts. Factors influencing emergence of new septa and cell proliferation associated with septal elongation have yet to be identified, in part because of difficulties inherent in studying this process in vivo. Using flow cytometric analysis of the DNA content of freshly isolated lung fibroblasts, we found that proliferation, as indicated by the percentage of cells in S plus G₂/M phases, peaked on postnatal Day 4 (P<0.04). By Days 9–10 the proliferation rate was lower than on Days 3, 4, 5, or 6 (P<0.005). We then evaluated rates of in vitro proliferation as a function of postnatal age in first passage fibroblasts and found that the proliferative phenotype expressed in vivo persists in vitro. Fibroblasts from 4–5-d-old pups increased in number and incorporated ³H-thymidine at a faster rate than did fibroblasts obtained from pups at other postnatal ages (P<0.0001). Age-dependent differences in cell cycle transit time were compared in fibroblasts synchronized by serum starvation and analyzed by flow cytometry at 2-h intervals from 13–21 h after release from serum starvation. A greater percentage of cells from 5-d-old pups entered S phase during this period than was seen for cells obtained from 2-, 9-, 13-, or 23-d-old rat pups (P=0.0001). Cells from 5-, 9-, and 13-d-old pups reentered G₀/G₁ by 21 h after release from serum starvation, in contrast to fibroblasts from 2- and 23-d-old rats which did not. Throughout the 15-h period after release from serum starvation, levels of cyclin E, which peaks at the G₁/S border, were highest in the 5-d-old cells (P<0.025). Synchronization with 2.5 mM hydroxyurea which inhibits DNA synthesis completely abolished age-related differences in cell cycle transit time, implying that age-dependent differences in lung fibroblast proliferation rates are the result of events occurring before S-phase entry.     

Differential growth factor induction and modulation of human gastric epithelial regeneration

While several autocrine/paracrine growth factors (GFs) can all stimulate epithelial regeneration in experimentally wounded primary gastric cultures, clinical relevance for their non-redundant cooperative actions in human gastric ulcer healing is suggested by the sequential pattern of GF gene induction in vivo. Using new HGE cell lines able to form a coherent monolayer with tight junctions as well as using primary human gastric epithelial cultures, we show that EGF, TGFalpha, HGF and IGFs accelerate epithelial restitution upon wounding, independently of the TGFbeta pathway (as opposed to intestinal cells). However, they differently modulate cell behavior: TGFalpha exerts strong effects (even more than EGF) on cytoplasmic spreading and non-oriented protruding activity of bordering cells whereas HGF preferentially coordinates single lamella formation, cell elongation and migration into the wound. IGF-I and IGF-II rather induce the alignment of bordering cells and maintain a compact monolayer front. The number of mitotic cells maximally increases with EGF, followed by TGFalpha and IGF-I,-II. The current study demonstrates that GFs differentially regulate the regeneration of human gastric epithelial cells through specific modulation of cell shape adaptation, migration and proliferation, further stressing that a coordination of GF activities would be necessary for the normal progression of post-wounding epithelial repair.     

Epithelial cell proliferation and migration in the developing rat gastric mucosa

To investigate cell proliferation in developing gastric antrum and fundus, proliferating gastric epithelial cells were labeled in fetal rats by intravenously injecting mothers with [³H]thymidine. In addition 14-day postnatal (dPN) rats were given [³H]thymidine intraperitoneally. Tissue was removed 1.5 hr later, and autoradiographs were prepared to identify proliferating cells. Total epithelial labeling indices (L.I.) reached a peak at 20 days gestation (dG), coincident with the appearance of pit/glands, then declined again by 22 dG (gestation end). At 18 dG, labeled cells were distributed throughout all levels of the stratified epithelia. Between 20 and 22 dG, as pit/gland development proceeded, labeled cells became concentrated in the gland bases and were only rarely seen on the surface (L.I. of surface cells <1% at 22 dG). By 14 dPN, proliferating cells were entirely absent from the epithelial surface. Approximately 15% of the endocrine cells were labeled at 22 dG, compared to only 2% at 14 dPN; zymogen cells were labeled (～6%) at 14 dPN; parietal cells did not label at any age studied. In addition, cell migration to the epithelial surface was studied in rats labeled at 22 dG, 14 dPN, or 28 dPN, and killed 1–20 days later. Migration times were slightly shorter in the 28 dPN group and were longer in fundus than antrum in all groups.     

Transient coexpression of cytokeratin and vimentin in differentiating rat Sertoli cells

The expression of cytokeratin and vimentin type intermediate filaments were studied in fetal, postnatal, and adult rat testes. Immunocytochemical observations were correlated with the light and electron microscopic analysis of the developing organs. The Sertoli cell precursors in 15-day-old fetal testes contained both cytokeratin and vimentin. A gradual reorganization of both filaments, accompanied by a decrease of cytokeratin-positivity, was observed toward the end of the fetal period. The simultaneous presence of cytokeratin and vimentin in the same cells was shown by double immunofluorescence of newborn testes and the primary culture of dissociated testicular cells. In postnatal Sertoli cells, cytokeratin-positivity continued to decrease and disappeared by the age of 14 days. The increase in vimentin content and the appearance of axially oriented vimentin filaments coincided with the acquisition of the columnar shape of the Sertoli cells. The presence of cytokeratin and vimentin in fetal and newborn testes, and only vimentin in the adult testes was confirmed by immunoblotting. The present results suggest that major qualitative changes in the expression of intermediate filament proteins can take place during the embryonic development. The expression of cytokeratin in developing Sertoli cells, although only transient, supports the epithelial origin of these cells and can be applied as a marker for embryonic and early postnatal Sertoli cells.     

Rapid purification of mammalian cardiac troponin T and its isoform switching in rat hearts during development

A rapid purification of troponin T from adult hearts of various species has been developed. The purification procedure included 60 degrees C treatment of the high salt extract, ammonium sulfate fractionation, and DEAE-cellulose column chromatography. The troponin T purified from the bovine left ventricle contained two isoforms, which differed in both apparent molecular mass and isoelectric point. Both isoforms were able to bind to F-actin filaments only in the presence of tropomyosin. Monoclonal antibody JLT12 against rabbit skeletal troponin T cross-reacted with both isoforms of bovine cardiac troponin T. There was no detectable difference in the relative amount of these two isoforms among different portions (atria, right and left ventricles) of the bovine heart. The purified protein was used as an antigen to immunize mice, and a mouse antiserum with high titer and specificity to both isoforms was subsequently obtained. This antiserum also cross-reacted with cardiac troponin T from chicken, rabbit, and rat. The antibodies were further used to probe cardiac development in rats by Western blotting and immunoprecipitation. The results clearly showed that there was a switch of troponin T isoforms between hearts from 20-day-old rat embryos and hearts from 14-day-old rats. Immunoprecipitation of the in vitro translation products of poly(A)+ RNA isolated from day 5 rat hearts revealed the presence of two isoforms of troponin T, suggesting that two mRNAs coding for these two isoforms existed in the heart cells. It is of interest to not that some profound changes in the morphology and function of cardiac muscle have also been detected at this time of development. Troponin T isoform switching thus may well represent an important marker for cardiac development and function.     

Fetal and adult human skin fibroblasts display intrinsic differences in contractile capacity.

One of the differences between fetal and adult skin healing is the unique ability of fetal wounds to heal without contracture and scar formation. Studies have shown that the ratio between the three isoforms of TGFbeta is different in adult and fetal wounds. Thus, we analyzed the capacity of adult and fetal human skin fibroblasts to contract collagen gels after stimulation with TGFbeta isoforms. In control medium, fetal fibroblasts had a contractile capacity similar to that of adult fibroblasts. However, the growth capacity of fetal fibroblasts was completely inhibited, in contrast to adult fibroblasts. When cells were treated with TGFbeta, fetal fibroblasts showed an inhibition of their contractile capacity whereas adult fibroblasts further contracted gels. The contractile response was similar for all isoforms of TGFbeta although TGFbeta3 always had the strongest effect. We considered that the regulation of cell contractile capacity by TGFbeta may be dependent on receptor expression for this cytokine, on myofibroblast differentiation of the cells, or in cell links with matrix. Since TGFbeta receptor analysis did not show differences in receptor affinity, we studied the expression of alpha-smooth muscle (SM) actin, a fibroblast contractile marker and of three integrins, the cell surface receptors specific of the attachment of the fibroblasts with collagen matrix. We observed that the expression of alpha-SM actin and alpha3 and beta1 integrin subunits was increased when TGFbeta was added to the medium of adult fibroblasts whereas the levels of the alpha1 and alpha2 subunits were unchanged. In contrast, fetal fibroblasts treated with TGFbeta showed a decrease of alpha1, alpha2, and beta1 integrin expression but no change in alpha3 integrin and in alpha-SM actin expression. These results indicate that intrinsic differences between fetal and adult fibroblasts might explain their opposite responses to TGFbeta stimuli. The variations in their alpha-SM actin and integrin expression patterns represent potentially important mechanisms used by fetal fibroblasts to regulate their response to cytokines, and likely contribute to the resultant differences in the quality of wound repair.     

Syndecan-1 expression in epithelial cells is induced by transforming growth factor beta through a PKA-dependent pathway

Syndecans comprise a major family of cell surface heparan sulfate proteoglycans (HSPGs). Syndecans bind and modulate a wide variety of biological molecules through their heparan sulfate (HS) moiety. Although all syndecans contain the ligand binding HS chains, they likely perform specific functions in vivo because their temporal and spatial expression patterns are different. However, how syndecan expression is regulated has yet to be clearly defined. In this study, we examined how syndecan-1 expression is regulated in epithelial cells. Our results showed that among several bioactive agents tested, only forskolin and three isoforms of TGFbeta (TGFbeta1-TGFbeta3) significantly induced syndecan-1, but not syndecan-4, expression on various epithelial cells. Steady-state syndecan-1 mRNA was not increased by TGFbeta treatment and cycloheximide did not inhibit syndecan-1 induction by TGFbeta, indicating that TGFbeta induces syndecan-1 in a post-translational manner. However, TGFbeta induction of syndecan-1 was inhibited by transient expression of a dominant-negative construct of protein kinase A (PKA) and by specific inhibitors of PKA. Further (i) syndecan-1 cytoplasmic domains were Ser-phosphorylated when cells were treated with TGFbeta and this was inhibited by a PKA inhibitor, (ii) PKA was co-immunoprecipitated from cell lysates by anti-syndecan-1 antibodies, (iii) PKA phosphorylated recombinant syndecan-1 cytoplasmic domains in vitro, and (iv) expression of a syndecan-1 construct with its invariant Ser(286) replaced with a Gly was not induced by TGFbeta. Together, these findings define a regulatory mechanism where TGFbeta signals through PKA to phosphorylate the syndecan-1 cytoplasmic domain and increases syndecan-1 expression on epithelial cells.     

Postnatal development of the glycine cleavage system in rat liver

The development of the glycine cleavage system was studied in liver mitochondrial extracts from neonatal and adult rats. The enzyme activity in 2-day-old pups was 29.3% of that measured in the adult and was found to increase in an age-dependent manner. Measurement of hepatic free amino acid concentrations at the neonatal ages showed that glycine levels were highest at 2 days and at 14 days were about 48% of those in the 2-day-old pups. Serine levels did not change between the age of 2 and 14 days. A developmental delay in the glycine cleavage system is responsible for the high levels of glycine in the neonatal rat liver.     

The pubertal-related decline in cellular proliferation and neurogenesis in the dentate gyrus of male rats is independent of the pubertal rise in gonadal hormones

Pubertal development is marked by significant decreases in cellular proliferation and neurogenesis in the dentate gyrus of the hippocampal formation. Although it is unclear what mediates these developmental changes in the dentate gyrus, gonadal hormones have been implicated in modulating many neurobiological processes during puberty and various parameters of neurogenesis in adulthood. Thus, it is possible that the gradual and sustained increase in gonadal hormones experienced during puberty plays a role in these changes in neurogenesis. In this experiments, we first quantified cellular proliferation and neurogenesis using 5-bromo-2'-deoxyuridine (BrdU) and doublecortin (DCX) immunohistochemistry, respectively, in the dentate gyrus of prepubertal (30 d), midpubertal (45 d), and adult (90 d) male rats. We found the decline in BrdU and DCX cell numbers throughout these ages was coincident with increases in their plasma testosterone levels. We next tested whether exposure to the pubertal rise in gonadal hormones was necessary for this decrease in hippocampal neurogenesis to occur. Thus, we examined cellular proliferation and neurogenesis in intact 30 day (prepubertal) and 60-day-old (late-pubertal) rats, as well as 60-day-old rats that had previously been gonadectomized or sham-gonadectomized at 30 days of age. Although we again found the expected decline in BrdU and DCX cell numbers between 30 and 60 days of age in the intact groups, there were no differences among the 60-day-old animals, regardless of gonadal status. These data indicate that the pubertal-related decline in hippocampal cellular proliferation and neurogenesis is independent of the pubertal change in gonadal hormones.     

Adoptive transfer of splenocytes in SCID mice implicates CD4+ T cells in apoptosis and epithelial proliferation associated with Helicobacter pylori-induced gastritis

Helicobacter pylori causes severe, rapidly progressive gastritis in severe combined immunodeficient (SCID) mouse recipients of congenic splenocytes. The H. pylori-infected and uninfected C57BL/6J and recipient SCID mice were evaluated to detect CD4+ and CD8+ T cells, B cells, apoptotic epithelial cells, and epithelial cell proliferation at postinoculation weeks 5, 6, 8, and 12. Serum was evaluated for anti-H. pylori IgG and IgM. In all H. pylori-infected mice, gastric CD4+ cell scores were increased, compared with scores for uninfected controls. Recipient mice differed, however, according to the source of the transferred CD4+ cells. The CD4+ cell scores for recipients of splenocytes from H. pylori-infected (immune) donors were indistinguishable from those for wild-type donor mice at all time points. In contrast, gastric mucosal CD4+ cell scores did not become significantly high until two weeks after transfer (postinoculation week 6) in recipients of cells from uninfected (na?ve) donors. Gastric epithelial apoptosis and the gastric epithelial proliferation zone were significantly (P < 0.05) increased in infected recipient and donor, compared with non-recipient and uninfected mice at postinoculation week 12. Results indicated that CD4 cells are sensitized in vivo and migrate to the gastric mucosa where they induce gastritis in response to H. pylori antigens. Influx of CD4 cells and gastritis are correlated with epithelial proliferation and apoptosis, and suggest that CD4-dependent H. pylori gastritis leads to epithelial damage with attendant proliferative and metaplastic responses.     

Attenuation of haptoglobin gene expression by TGFbeta requires the MAP kinase pathway.

In addition to important roles in the regulation of cell growth and cell restitution, both pro- and anti-inflammatory effects have been ascribed to TGFbeta in intestinal epithelial cells. However, the mechanisms involved in TGFbeta-dependent anti-inflammatory activities remain to be determined. In the rat intestinal epithelial cell line IEC-6, TGFbeta attenuated the glucocorticoid-dependent increases in mRNA levels of the acute phase protein gene haptoglobin, and of C/EBP isoforms beta and delta. Supershift assays demonstrated a TGFbeta-mediated decrease in the binding of C/EBP isoforms beta and delta to the haptoA and haptoC C/EBP DNA-binding sites from the haptoglobin promoter. Mutations of both HaptoA and HaptoC sites abolished the glucocorticoid-dependent activation and the TGFbeta-mediated attenuation of the haptoglobin promoter, as assessed by transient transfection assays. TGFbeta induced p42/p44 MAP kinase activities. Treatment with the MEK 1/2 inhibitor PD 98059 abolished TGFbeta attenuation. These results suggest that C/EBP isoforms are involved both in the glucocorticoid-dependent induction and in the TGFbeta-mediated attenuation of haptoglobin expression. Furthermore, p42/p44 MAP kinases may function in a TGFbeta-dependent signaling pathway leading to attenuation of haptoglobin expression.     

Helicobacter pylori vacuolating cytotoxin (VacA) alters cytoskeleton-associated proteins and interferes with re-epithelialization of wounded gastric epithelial monolayers

In previous studies, we demonstrated that Helicobacter pylori vacuolating cytotoxin (VacA) inhibits gastric epithelial cell proliferation and inhibits epidermal growth factor (EGF)-activated signal transduction. Cell proliferation and migration, both essential for mucosal healing are dependent on the cell cytoskeleton. Other investigators demonstrated that VacA induces vacuolation of eukaryotic cells. Since in some cells, control of actin cytoskeleton involves GTP-binding proteins of Rho family, in this study we examined whether VacA affects wound re-epithelialization, cell cytoskeleton-associated proteins Rho, Rac1 in a gastric epithelial (RGM1) cell monolayer wound model, and whether these changes correlate with vacuolation. VacA treatment significantly inhibited wound re-epithelialization, cell proliferation vs control. VacA-induced cell vacuolation strongly correlated with inhibition of wound re-epithelialization. Furthermore, VacA reduced Rac-1 protein expression and distribution, and C3-mediated ADP-ribosylation of Rho. These findings suggest that VacA may interfere with repair of gastric mucosal injury and ulcer re-epithelialization by altering cytoskeleton-dependent cell functions and signaling.     

Competition of glycerol with other oxidizable substrates in rat brain.

The rate of conversion of [1,3-14C]glycerol into 14CO2 was measured in the presence and absence of unlabelled alternative substrates in whole homogenates from the brains of young (4-6 and 18-20 days old) and adult rats. Unlabelled glucose decreased 14CO2 production from [1,3-14C]glycerol by about 40% at all ages studied. Unlabelled 3-hydroxybutyrate significantly decreased the 14CO2 production from both low (0.2 mM) and high (2.0 mM) concentrations of glycerol in 4-6- and 18-20-day-old rat pups. However, the addition of 3-hydroxybutyrate had no effect on the rate of 14CO2 production from 2.0 mM-glycerol in adult rats, suggesting that the interaction of 3-hydroxybutyrate with glycerol in adult rat brain is complex and may be related to the biphasic kinetics previously reported for glycerol oxidation. Unlabelled glutamine decreased the production of 14CO2 by brain homogenates from 18-20-day-old and adult rats, but not in 4-6-day-old rat pups. In the converse situation, the addition of unlabelled glycerol to whole brain homogenates had little effect on the rate of 14CO2 production from [6-14C]glucose, 3-hydroxy[3-14C]butyrate and [U-14C]glutamine, although some significant differences were noted. Collectively these results suggest that glycerol and these other substrates may be metabolized in separate subcellular compartments in brain such that the products of glucose, 3-hydroxybutyrate and glutamine metabolism can dilute the oxidation of glycerol, but the converse cannot occur. The data also demonstrate that there are complex age-related changes in the interaction of glycerol with 3-hydroxybutyrate and glutamine. The fact that glycerol oxidation was only partially suppressed by the addition of 1-5 mM-glucose, -3-hydroxybutyrate or -glutamine could also suggest that glycerol may be selectively utilized as an energy substrate in some discrete brain region.     

Induction of rat jejunal epithelial cell expression of sucrase-isomaltase by glucocorticoids in primary cell culture and in vivo

The cell cycle phase that mediates the induction of intestinal sucrase-isomaltase (SI) expression by glucocorticoids was investigated by measuring migration rates of 3H-DNA-labeled and of SI-containing epithelial cells by autoradiography and indirect immunofluorescent staining after simultaneous administration of [3H]thymidine and cortisone to 12-d-old rat pups. By 24 and 48 h, lead 3H-DNA-labeled cells had migrated 7.8 and 12.4 cell positions higher on the villus than lead cells expressing SI. Cell migration rates from 12 to 24 h and 24 to 48 h were 0.68 and 0.97 cell position/h. Thus, commitment to SI expression occurred in cells 11.5-12.8 h after the S phase, which is calculated to be in the G1 phase. To determine whether committed cells need to replicate to express SI, cell differentiation was examined in primary cultures of crypt cells originating from corticosterone-treated rats. About two-thirds of cultured cells were retarded in the S phase after plating, as judged by no increase of DNA labeling indices, no change in epithelial cell number, and the absence of mitosis (less than 0.01%). The proportion of cells expressing SI increased from 0 to 6-8% between 12 and 24 h, and reached 48% 48 h after plating on collagen-coated dishes. SI expression did not occur in cells plated on glass or plastic surfaces. Pulse labeling with [35S]methionine confirmed that de novo synthesis of SI occurred in cell cultures. Thus, additional cell cycling of committed cells occurring in vivo is not obligatory for the expression of SI.