Expression pattern of Galectin 4 in rat placentation.

Galectin 4 (Gal4) is abundantly expressed in the epithelium of the gastrointestinal tract, and functional analysis has concentrated on its roles associated with polarized membrane trafficking. This study aimed to investigate the expression of Gal4 in placentation. The expression level of Gal4 was revealed to be lower in differentiated Rcho-1 cells (a model system of rat trophoblast differentiation) than in proliferative cells. In the rat placenta, immunohistochemical analysis showed that Gal4 is preferentially located in the maternal-fetal junctional zone. These results suggest that down-regulation of Gal4 may be involved in the promotion of trophoblast cell differentiation.

Differential expression and the anti-apoptotic effect of human placental neurotrophins and their receptors.

Neurotrophin (NT) is important in the survival, maintenance and differentiation of neuronal tissue, and functions in follicle maturation, tumor growth, angiogenesis and immunomodulation; however, the expression of NT and its receptors (NTR) in human placenta and their influence on fetal growth are unclear. Here we investigated the correlation of NT and NTR in human placenta with uterine environment and fetal growth. TrkB, a NTR, mRNA was expressed on decidual and villous tissue and increased with gestational age, localizing in the trophoblast layer and endothelium by immunohistochemistry. Villous TrkB mRNA was significantly increased in preeclampsia (PE) than in controls and was higher in the normotensive small for gestational age (SGA) placenta, although it was not significant. It was also significantly increased in the small twin of discordant twin pregnancies. Brain-derived neurotrophic factor (BDNF), the main ligand of TrkB, was expressed in membranous chorion and villous tissue and was significantly higher in maternal plasma in normotensive SGA and PE than in controls. TrkB mRNA expression was up-regulated on cultured villous tissue explants and on JEG-3, a choriocarcinoma cell line, by H(2)O(2) treatment. BDNF decreased apoptotic cells in H(2)O(2)-treated JEG-3, indicating that BDNF/TrkB signaling had anti-apoptotic effects against oxidative stress in JEG-3, suggesting a protective role of BDNF/TrkB in human villous tissue under unfavorable conditions in utero.

Branched-chain amino acid supplemented diet during maternal food restriction prevents developmental hypertension in adult rat offspring.

Maternal food restriction is known to cause developmental hypertension in offspring. We have previously shown that maternal high-protein diet can reverse fetal programming of hypertension and that branched-chain amino acid (BCAA) concentrations in maternal and fetal plasma were increased by maternal high-protein intake. Then, we hypothesized that isocaloric supplementation with BCAA to a maternal food restriction can reverse the adverse outcome. Pregnant rats were divided into four groups at 7.5 days postcoitum: normally nourished (NN) and 70% undernourished (UN) groups with and without BCAA supplementation (NN-standard diet (SD), NN-BCAA, UN-SD and UN-BCAA groups). Compared with pups in the NN groups, those in the UN-SD group had significantly increased systolic blood pressure (SBP) at 8 and 16 weeks of age (P < 0.05). However, the elevation of SBP was not observed in offspring in the UN-BCAA group. Offspring glomeruli number of the UN groups was significantly lower (P < 0.05) than that of the NN groups, independent of BCAA supplementation. Angiotensin II receptor type 2 (ATR2) mRNA and protein expression in the kidney was significantly augmented in the UN-BCAA group at 30 weeks of age. In conclusion, BCAA supplementation during maternal food restriction prevents developmental hypertension together with increased ATR2 expression in adult offspring kidney.

Activated protein C attenuates coagulation-associated over-expression of fibrinolytic activity by suppressing the thrombin-dependent inactivation of PAI-1.

Several activated coagulation factors have been reported to enhance fibrinolysis by neutralizing plasminogen activator inhibitor type 1 (PAI-1) activity. We evaluated the physiological relevance of this mechanism using the euglobulin clot lysis time (ECLT) assay in the presence and absence of Ca2+, which is controlled by PAI-1 and mimics physiological thrombolysis. We found that the ECLT (18.5 +/- 0.6 h) was shortened by Ca2+ (5 mm) (6.6 +/- 0.1 h). A significant difference was observed in thrombin generation by the presence of Ca2+ in the euglobulin fraction. Prothrombin was almost fully converted to thrombin within 15 min in the presence of Ca2+, whereas essentially no conversion was observed without Ca2+. The presence of activated protein C (aPC) suppressed thrombin generation, and attenuated the shortening of ECLT in a dose-dependent manner, an effect enhanced by phospholipid and protein S. In the absence of Ca2+, aPC did not prolong the ECLT. After addition of biotin-labeled recombinant PAI-1 to the euglobulin fraction, PAI-1 was cleaved to lower molecular weight forms only in the presence of Ca2+. This cleavage did not occur in the presence of aPC, suggesting that thrombin was the catalyst for PAI-1 cleavage. The cleavage and inactivation of PAI-1 by generated thrombin is proposed to be responsible for the shortening of ECLT by Ca2+ and for coagulation-associated over-expression of fibrinolysis. Under such conditions, aPC appeared to suppress thrombin generation and to normalize highly activated fibrinolysis.

Active mitochondria surrounding the pancreatic acinar granule region prevent spreading of inositol trisphosphate-evoked local cytosolic Ca(2+) signals.

Agonist-evoked cytosolic Ca(2+) spikes in mouse pancreatic acinar cells are specifically initiated in the apical secretory pole and are mostly confined to this region. The role played by mitochondria in this process has been investigated. Using the mitochondria-specific fluorescent dyes MitoTracker Green and Rhodamine 123, these organelles appeared as a bright belt concentrated mainly around the secretory granule area. We tested the effects of two different types of mitochondrial inhibitor on the cytosolic Ca(2+) concentration using simultaneous imaging of Ca(2+)-sensitive fluorescence (Fura 2) and electrophysiology. When carbonyl cyanide m-chlorophenylhydrazone (CCCP) was applied in the presence of the Ca(2+)-releasing messenger inositol 1,4, 5-trisphosphate (IP(3)), the local repetitive Ca(2+) responses in the granule area were transformed into a global rise in the cellular Ca(2+) concentration. In the absence of IP(3), CCCP had no effect on the cytosolic Ca(2+) levels. Antimycin and antimycin + oligomycin had the same effect as CCCP. Active mitochondria, strategically placed around the secretory pole, block Ca(2+) diffusion from the primary Ca(2+) release sites in the granule-rich area in the apical pole to the basal part of the cell containing the nucleus. When mitochondrial function is inhibited, this barrier disappears and the Ca(2+) signals spread all over the cytosol.

Calcium binding capacity of the cytosol and endoplasmic reticulum of mouse pancreatic acinar cells.

1. The droplet technique was used in this study to measure total calcium loss from pancreatic acinar cells due to calcium extrusion. The calcium binding capacity of the cytosol (kc) was measured as the ratio of the decrease in the total calcium concentration of the cytosol of the cell (Delta[Ca]c) and the synchronously occurring decrease in the free calcium ion concentration in the cytosol (Delta[Ca2+]c). The calcium dependency of the calcium binding capacity was determined by plotting values of kc against the corresponding [Ca2+]c. The rise in the cytosolic Ca2+ concentration of pancreatic acinar cells was triggered by stimulation with a supramaximal dose of cholecystokinin (CCK). The recovery of [Ca2+]c during continued exposure to the agonist was due to calcium extrusion from the cell. 2. The calcium binding capacity was about 1500-2000 for the [Ca2+]c range 150-500 nM. The mechanism of buffering was not investigated in this study. The calcium binding capacity of the cytosol did not vary significantly with [Ca2+]c in this range. The CCK-evoked decrease in the total calcium concentration in the lumen of the endoplasmic reticulum (ER) can be estimated from our data, taking into account previously published values for the volume of the ER in pancreatic acinar cells. Comparing the decrease in the total ER calcium concentration with our recently reported values for agonist-induced reductions in the free Ca2+ concentration inside the ER, we estimate that the calcium binding capacity of the ER is approximately 20. In pancreatic acinar cells we have therefore found a difference of two orders of magnitude in the efficiency of calcium buffering in the cytosol and the ER lumen.

Two cases of intrapulmonary lymph node presenting as a peripheral nodular shadow: diagnostic differentiation from lung cancer.

We present two cases of intrapulmonary lymph node. The patients were a 44-year-old woman and a 71-year-old man each with a small peripheral nodule in the lung. On computed tomography (CT) scans, both nodules were spiculated. Since histological diagnosis could not be obtained by bronchoscopic examination or CT-guided needle biopsy, they underwent video-assisted thoracoscopic surgery. Histological examination of the resected material revealed that both nodules were composed of lymph node. Intrapulmonary lymph node has until recently been assigned no clinical significance; however, differential diagnosis of this lesion from lung cancers and other metastatic tumors is now clinically important.

Intracellular glucose switches between cyclic ADP-ribose and inositol trisphosphate triggering of cytosolic Ca2+ spiking.

Cyclic ADP-ribose (cADPR) is a potentially important intracellular Ca2+ releasing messenger [1-5]. In pancreatic acinar cells where intracellular infusion of both inositol trisphosphate (IP3) and cADPR evoke repetitive Ca2+ spiking [6], the cADPR antagonist 8-NH2-cADPR [7], which blocks cADPR-evoked but not IP3-evoked Ca2+ spiking, can abolish Ca2+ spiking induced by physiological levels of the peptide hormone cholecystokinin (CCK) [8]. We have tested the effect of intracellular glucose on the ability of IP3, cADPR and CCK to induce cytosolic Ca2+ spikes in pancreatic acinar cells. In order to gain access to the intracellular cytosol, we used the whole-cell configuration of the patch-clamp technique [9] and monitored cytosolic Ca2+ concentration changes by measuring the Ca(2+)-dependent ionic current [10-13]. Glucose (300 microM to 10 mM) in the patch pipette/intracellular solution prevented cADPR from evoking Ca2+ spiking. The same effect was observed with 2-deoxy-glucose, but not L-glucose. In contrast, glucose potentiated IP3-evoked Ca2+ spiking. CCK evoked Ca2+ spiking irrespective of the presence or absence of intracellular glucose, but the cADPR antagonist 8-NH2-cADPR blocked CCK-evoked Ca2+ spiking only in the absence of intracellular glucose. This suggests that the hormone can evoke Ca2+ spiking via either the IP3 or the cADPR pathway. The intracellular glucose level may control a switch between these two pathways.

The calcium store in the nuclear envelope.

The nuclear envelope has a relatively small volume, but is connected up to the vastly larger endoplasmic reticulum. The Ca2+ concentration in the lumen of the interconnected nuclear envelope and endoplasmic reticulum network is in the resting state maintained at a level of more than 100 microM. There are specific Ca2+ release channels present in the inner nuclear membrane that can be activated by inositol trisphosphate or cADP ribose. The system, therefore, allows selective release of Ca2+ into the nucleoplasm which could be important for the control of specific types of gene expression.

Termination of cytosolic Ca2+ signals: Ca2+ reuptake into intracellular stores is regulated by the free Ca2+ concentration in the store lumen.

The mechanism by which agonist-evoked cytosolic Ca2+ signals are terminated has been investigated. We measured the Ca2+ concentration inside the endoplasmic reticulum store of pancreatic acinar cells and monitored the cytoplasmic Ca2+ concentration by whole-cell patch-clamp recording of the Ca2+-sensitive currents. When the cytosolic Ca2+ concentration was clamped at the resting level by a high concentration of a selective Ca2+ buffer, acetylcholine evoked the usual depletion of intracellular Ca2+ stores, but without increasing the Ca2+-sensitive currents. Removal of acetylcholine allowed thapsigargin-sensitive Ca2+ reuptake into the stores, and this process stopped when the stores had been loaded to the pre-stimulation level. The apparent rate of Ca2+ reuptake decreased steeply with an increase in the Ca2+ concentration in the store lumen and it is this negative feedback on the Ca2+ pump that controls the Ca2+ store content. In the absence of a cytoplasmic Ca2+ clamp, acetylcholine removal resulted in a rapid return of the elevated cytoplasmic Ca2+ concentration to the pre-stimulation resting level, which was attained long before the endoplasmic reticulum Ca2+ store had been completely refilled. We conclude that control of Ca2+ reuptake by the Ca2+ concentration inside the intracellular store allows precise Ca2+ signal termination without interfering with store refilling.

Phospholipase C inhibitor, U73122, releases intracellular Ca2+, potentiates Ins(1,4,5)P3-mediated Ca2+ release and directly activates ion channels in mouse pancreatic acinar cells.

It is recognized in many cellular systems that the receptor/G-protein activation of phospholipase C and Ins(1,4,5)P3 production is the transduction pathway regulating the release of Ca2+ from internal stores. Ca2+ signals can now be monitored at the level of single cells but the biochemical detection of Ins(1,4,5)P3 cannot match this resolution. It is often difficult or impossible to directly attribute responses evoked in single cells by putative phospholipase C-coupled agonists to changes in Ins(1,4,5)P3 levels. U73122 is an aminosteroid that is reported to act as a specific inhibitor of phospholipase C and it has become an important tool in establishing the link between phospholipase C activation and cellular Ca2+ signalling. In the present study we use both patch-clamp electrophysiology and the imaging of fluorescent Ca2+ indicators to investigate the effect of U73122 in mouse pancreatic acinar cells. The study reveals that U73122 has effects other than the inhibition of phospholipase C. U73122 can directly activate ion channels. It can itself promote the release of Ca2+ from intracellular stores in permeabilized cells and in intact cells it triggers a release of Ca2+ that is initiated specifically at the secretory pole of these morphologically and functionally polarized cells. We also present evidence that U73122 can potentiate the response to Ins(1,4,5)P3; this is seen both in permeabilized cells and in patch-clamp protocols in which cells are internally dialysed with submaximal concentrations of Ins(1,4,5)P3. The effects of U73122 are therefore multiple and not specific for the inhibition of phospholipase C. Importantly, all the effects described influence Ca2+ signalling yet in many experimental protocols some of these effects can go unnoticed and might in error be attributed simply to the inhibition of Ins(1,4,5)P3 production.

Ca2+ flow via tunnels in polarized cells: recharging of apical Ca2+ stores by focal Ca2+ entry through basal membrane patch.

Intracellular Ca2+ store depletion induces Ca2+ entry across the plasma membrane, allowing the store to recharge. In our experiments, Ca2+ stores in pancreatic acinar cells were depleted by acetylcholine (ACh) stimulation in Ca2+-free solution. Thereafter, Ca2+ entry was only allowed through a CaCl2-containing pipette attached to the basal membrane. Recharging intracellular Ca2+ stores via a patch pipette occurred without a rise in the cytosolic Ca2+ concentration and depended on the operation of a thapsigargin-sensitive Ca2+ pump. After a period of focal Ca2+ entry, ACh could again evoke a rise in the cytosolic Ca2+ concentration, and this rise always started in the apical secretory pole. Recharging the apical Ca2+ store therefore depends on Ca2+ flow through a tunnel from the basal to the secretory pole, and the endoplasmic reticulum Ca2+ pump is essential for this process.

Blockade of DNA synthesis induced by platelet-derived growth factor by tranilast, an inhibitor of calcium entry, in vascular smooth muscle cells.

The present study was conducted to establish a pharmacological method of controlling growth of vascular smooth muscle cells (VSMC) by blocking calcium entry. In cultured rat VSMC, 1 nM platelet-derived growth factor (PDGF) induced a biphasic elevation of cytoplasmic free calcium concentration, ([Ca2+]c). The second sustained phase of [Ca2+]c was dependent on extracellular calcium. At lower concentrations, PDGF induced oscillatory changes in [Ca2+]c, and reduction of extracellular calcium attenuated the oscillation. An antiallergic compound, tranilast, abolished the sustained phase of [Ca2+]c induced by 1 nM PDGF. Tranilast also inhibited the oscillatory changes in [Ca2+]c induced by 200 pM PDGF. In addition, PDGF-induced calcium influx in the late G1 phase, as assessed by measuring the initial uptake of 45Ca, was inhibited by tranilast in a concentration-dependent manner. Tranilast also inhibited PDGF-augmented DNA synthesis; the ID50 for the inhibition of DNA synthesis was nearly identical to that for calcium influx. Although tranilast blocked PDGF-induced calcium entry, it did not affect PDGF-mediated autophosphorylation of the PDGF receptor, activation of phosphatidylinositol 3-kinase, activation of Ras or mitogen-activated protein kinase. Similarly, PDGF-induced elevation of diacylglycerol was not affected by tranilast. These results suggest that the antiallergic drug tranilast inhibits PDGF-induced DNA synthesis by blocking PDGF-mediated calcium entry. Tranilast may be of use in controlling PDGF-induced DNA synthesis in VSMC.

Primary lung cancer associated with diffuse granulomatous lesions in the pulmonary parenchyma.

A 60-year-old man was admitted to our hospital for productive cough. Chest roentgenography and CT scan disclosed a left hilar tumor invading the mediastinum, with mediastinal lymphadenopathy and diffuse micronodular shadows in both lung fields. A biopsied sample of the tumor revealed squamous cell carcinoma, while noncaseating epithelioid cell granulomas were observed in the samples obtained by transbronchial lung biopsy. The granulomas in the pulmonary parenchyma were determined to be sarcoid reactions secondary to lung cancer, since there was no evidence of sarcoidosis. Combination chemotherapy was effective for the tumor, and the granulomas disappeared after completion of the chemotherapy. These findings suggest the presence of a relationship between sarcoid reactions and lung cancer in this case.

[Intrahepatic arterial infusion chemotherapy for the colon cancer patients with liver metastases--a comparison of arterial embolization chemotherapy versus continuous arterial infusion chemotherapy].

To compare the prognostic benefit between arterial embolization chemotherapy (AEC) and continuous arterial infusion chemotherapy (CAIC), we clinicopathologically examined 32 colorectal cancer patients with liver metastases. Seventy patients were treated with AEC, and 15 patients with CAIC. In the AEC regimen, either ADR or CDDP dissolved in Lipiodol was infused by the implanted reservoir every one to two months. Otherwise, for the CAIC regimen, 360 mg/m2/day of 5-FU was continuously infused by the reservoir for 14 days. Subsequently, on day 22 after the initial 5-FU infusion, 180 mg/m2/day of 5-FU was continuously infused for 7 days. After a 7-day interval without infusion, 180 mg/m2/day of 5-FU was infused for 7 days. This 7-day infusion/7-day no-infusion cycle was repeated. The efficacy of the AEC was 6% (1 PR + 7 NC + 9 PD), CAIC (6 PR + 4 NC + 5 PD), suggesting that CAIC provides the better response rate. With regard to prognosis, the 1- and 2-year survival rates of AEC were 63% and 19%, respectively. The median survival time was 415 days. Otherwise, the 1- and 2-year survival with CAIC was 54% and 40%, respectively, and the median survival time was 470 days. No significant difference between the AEC and the CAIC was observed in the colorectal cancer patients with liver metastases.

Modulation of adenosine triphosphate-sensitive potassium channel and voltage-dependent calcium channel by activin A in HIT-T15 cells.

The ATP-sensitive potassium channel (KATP channel) determines the membrane potential of pancreatic beta-cells and plays a critical role in the regulation of insulin secretion. The present study was conducted to investigate the effect of activin A, a member of the transforming growth factor-beta supergene family, on the KATP channel in HIT-T15 clonal hamster insulinoma cells. In an excised inside-out patch, ATP-sensitive currents with a single channel conductance of approximately 20 picosiemens were observed. In an outside-out patch, currents with identical unitary conductance were also observed. In either case, the currents were augmented by diazoxide and blocked by glibenclamide, verifying that they were KATP channel currents. When KATP channel currents were monitored in an outside-out patch, activin A added to the bath solution inhibited KATP channel currents. Upon removal of activin A, the KATP channel currents were restored, suggesting that the inhibition was not due simply to spontaneous disappearance of channel activity (run-down). The KATP channel activity was markedly reduced after the addition of activin A and was reversed by diazoxide. Besides the inhibition of KATP channel, activin A increased, in a perforated patch, the amplitude of the inward Ba2+ current in response to a depolarizing pulse from -40 to +10 mV. Under the current clamp condition, activin A induced gradual depolarization, followed by a burst of action potentials. Activin-mediated action potentials were accompanied by an elevation of the cytoplasmic free calcium concentration. These results indicate that activin A causes depolarization of the plasma membrane by inhibiting the activity of the KATP channel. In addition, activin A directly modulates the voltage-dependent calcium channel and augments calcium entry.

Expression of calcium-permeable cation channel CD20 accelerates progression through the G1 phase in Balb/c 3T3 cells.

CD20 is a transmembrane protein that functions as a Ca(2+)-permeable cation channel (Bubien, J. K., Zhou, L. J., Bell, P. D., Frizzel, R. A., and Tedder, T. F. (1993) J. Cell Biol. 121, 1121-1132) and is involved in growth regulation of B lymphocytes. In order to further investigate the role of calcium entry in cell cycle progression, we introduced the cDNA encoding a Ca(2+)-permeable cation channel, CD20, into Balb/c 3T3 cells. Balb/c 3T3 cells transfected with a vector containing cDNA encoding CD20 expressed the CD20 protein, which was detected by assaying the binding of a monoclonal antibody against CD20. Calcium-permeable cation channel activity was detected in CD20-expressing cells by whole cell patch clamp recording and microfluorometric determination of the cytoplasmic Ca2+ concentration using fura-2. The expression of CD20 induced significant alterations in the responses of the cells to insulin-like growth factor-I (IGF-I). IGF-I induced DNA synthesis by control cells only when they had been pretreated with both platelet-derived growth factor (PDGF) and epidermal growth factor (EGF). In contrast, DNA synthesis by 30% of the quiescent CD20-expressing cells was initiated in response to IGF-I in the absence of priming with PDGF and EGF. When control quiescent cells were primed with PDGF and EGF, the addition of IGF-I led to the initiation of DNA synthesis after 14 h or more, whereas it induced DNA synthesis by CD20-expressing cells primed with PDGF and EGF 4 h earlier. The IGF-induced DNA synthesis was dependent on extracellular Ca2+, and expression of CD20 reduced the concentration of extracellular Ca2+ required for it. Furthermore, DNA synthesis by approximately 25% of the CD20-expressing cells was initiated after priming with PDGF and EGF, even in the absence of the progression factor IGF-I. These results indicate that CD20 expressed in Balb/c 3T3 cells functions as a constitutively active Ca(2+)-permeable cation channel and that expression of CD20 accelerates G1 progression in a Ca(2+)-dependent manner.

Conversion of amylase-secreting rat pancreatic AR42J cells to neuronlike cells by activin A.

When AR42J cells, an amylase-secreting pancreatic exocrine cell line, were treated with activin A, cells extended neuritelike processes, and, concomitantly, amylase-containing vesicles disappeared. Immunofluorescence and immunoelectron microscopy revealed that these processes had neurite-specific cytoskeletal architectures: neurofilaments and microtubule bundles with cross-bridges of microtubule-associated protein 2. In addition to such morphological changes, activin-treated cells exhibited a marked increase in cytoplasmic free calcium concentration in response to depolarizing concentration of potassium. Moreover, activin-treated AR42J cells expressed mRNA for alpha 1 subunit of the neuroendocrine/beta cell-type voltage-dependent calcium channel. In naive AR42J cells, a sulfonylurea compound, tolbutamide, did not affect free calcium concentration, while it induced a marked elevation of free calcium in activin-treated cells. Single channel recording of the membrane patch revealed the existence of ATP-sensitive potassium channel in activin-treated cells. These results indicate that activin A converts amylase-secreting AR42J cells to neuronlike cells. Given that pancreatic endocrine cells possess neuronlike properties and express ATP-sensitive potassium channel as well as neuroendocrine/beta cell-type voltage-dependent calcium channel, activin treatment of AR42J cells may provide an in vitro model system to study the conversion of pancreatic exocrine cells to endocrine cells in islets.

Production of activin A and follistatin in cultured rat vascular smooth muscle cells.

Activin A, a member of the transforming growth factor beta supergene family, modulates DNA synthesis in cultured rat vascular smooth muscle cells (VSMC) (Kopma et al. (1993) Exp. Cell. Res. 206, 152-156). In the present study, we studied the production of activin A and follistatin in VSMC. When VSMCs cultured in a 24-well plate were cultured with 10% fetal calf serum (FCS) for 24 h, 0.94 +/- 0.20 pmol/well (mean +/- SE, n = 6) of bioactive activin was released into the culture media. Reverse-transcription polymerase chain-reaction revealed the expression of mRNA for the beta A subunit of inhibin but not for either the beta B or alpha subunit. Bioactivity of activin was increased in quiescent cells treated with FCS or platelet-derived growth factor (PDGF) but not with angiotensin II (Ang II) or insulin-like growth factor-I (IGF-I). Ang II or IGF-I did not stimulate DNA synthesis by itself but, when these two agents were combined, they increased nuclear labeling by 16.4% and release of bioactive activin by 170% of basal. The dose-response relationship and time course study indicated that PDGF-mediated release of activin correlated with initiation of DNA synthesis. Steady state expression of mRNA for the beta A subunit was markedly elevated 12 h after the addition of PDGF and was reduced thereafter. To assess the significance of autocrine activin, the effect of PDGF was determined in the presence and absence of excess of exogenous follistatin. The PDGF-mediated DNA synthesis was enhanced by the addition of excess follistatin.(ABSTRACT TRUNCATED AT 250 WORDS)