| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Toxicology (M.C., W.R.), Institute for Cancer Research and Neuromuscular Research Department (M.C., W.R.), Institute of Anatomy, University of Vienna, 1090 Vienna, Austria; and Ludwig Boltzmann Institute for Pediatric Endocrinology and Immunology (K.H.), 1090 Vienna, Austria
Address all correspondence and requests for reprints to: Walter Rossmanith, Neuromuscular Research Department, Institute of Anatomy, University of Vienna, Währinger Strasse 13, 1090 Wien, Austria. E-mail: walter.rossmanith{at}univie.ac.at.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Based on the similarity to known activins, two further mammalian subunits, termed activin ßC and ßE, have recently been cloned from man, mouse, and rat (9, 10, 11, 12, 13, 14). So far, however, no biological role for activin ßC or ßE has been elucidated. Recombinant protein as far as available (ßCßC dimers from Chinese hamster ovary cells) has not revealed any biological effect, even in cellular systems typically responsive to activin A (15, 16). Moreover, mice deficient in activin ßC, activin ßE, or both develop normally and are both viable and fertile (17). And although activin ßC and ßE are predominantly and abundantly expressed in the liver of wild-type mice, knockout mice do not show any impairment of liver function or development. Thus, the function of activin C and E remains unclear.
Because expression of activin ßC and ßE is confined to the liver (10, 13, 18) and both activin A and TGF-ß are well-known key players in liver growth regulation, a hepatic role of activin C and E is still an attractive hypothesis. Whereas TGF-ß is expressed by nonparenchymal liver cells (13, 19), activin ßA, ßC, and ßE are primarily expressed by the parenchyma (13). Yet both, TGF-ß and activin A have been shown to inhibit hepatocyte DNA synthesis and induce cell death (for reviews, see Refs. 20 and 21). A partially redundant, apparently nonessential role of activin C and E in liver proliferation control thus remains a reasonable basis for further analyses.
Given the lack of suitable recombinant protein, targeted overexpression of activin ßC and ßE in the liver appears to be the approach of choice to study gain-of-function effects of these two genes in vivo. Recent advances in nonviral gene transfer to the liver now enabled the direct transfer of plasmids encoding the activin ßC, ßE, or ßA gene to adult mice. Predictably and reproducibly, the hydrodynamics-based method that uses high-volume tail-vein injection to selectively transfer naked DNA to the mouse liver (22, 23, 24), resulted in the transfection of hepatocytes and the transient overexpression of activin C, E, or A. Moreover, as this type of gene transfer is associated with a distinct damage of liver tissue (24), the model allowed us to study activin overexpression in the light of an inherent regenerative proliferation of hepatic cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). All three activin-expressing plasmids (paßC/lacZ, paßE/lacZ, paßA/lacZ) contain the complete open reading frame of the respective rat activin ß-subunit precursor (13, 25) followed by the encephalomyocarditis virus internal ribosomal entry site (IRES) and the E. coli lacZ gene (Fig. 1A). Mature peptide sequences of mouse and rat activins are either identical (activin A) or almost identical (activin C and E) (13). To achieve efficient translation of the activin ß-subunits, the nucleotide sequence around the start codon was changed in each case to the Kozak consensus sequence (GCCACCATGG) by PCR mutagenesis (26). All plasmid vectors were built on the backbone of pEGFP-N3 (BD Biosciences, CLONTECH Laboratories, Inc., Palo Alto, CA) and expressed the encoded genes from the human cytomegalovirus (CMV) immediate early promoter.
|
Animal procedures and gene transfer
Mice were acquired and used in compliance with the Austrian law, and all experiments were approved by the responsible institutional and ministerial committees. Animal maintenance was performed in accordance with the Guide for the Care and Use of Laboratory Animals (27).
Male BALB/cOlaHsd mice (Harlan Winkelmann, Borchen, Germany) were housed individually in macrolon cages on standard softwood bedding under standardized conditions (controlled temperature and humidity, 12-h light, 12-h dark rhythm, food pellets and tap water ad libitum). Daily recordings of body weight and food consumption were made 2 h after the beginning of the light period. At the time of gene transfer mice were 9–10 wk old.
Gene transfer was performed as recently described (24). Briefly, plasmid DNA was diluted in sterile Ringer’s solution to a final concentration of 10 µg/ml; 100 ml/kg body weight of the plasmid DNA solution were injected into the tail vein at a rate of 0.4 ml/sec using a 27-gauge needle. Injections were administered between 1 and 2.5 h after dawn (light on).
Ninety minutes before they were killed, animals received 100 mg/kg body weight of 5-bromo-2'-deoxyuridine (BrdU) by ip injection of a 10 mg/ml solution.
Blood and tissue sampling
Animals were anesthetized with diethylether vapor and blood collected by venipuncture of the orbital plexus using a heparinized cannula. Anesthetized animals were euthanized by cervical dislocation. Liver, kidneys, heart, spleen, brain, and testes were collected in toto, weighed, eventually dissected, snap frozen in liquid nitrogen cooled isopentane, and stored at -80 C for further processing. Pieces of the lung, intestine, and muscle were processed similarly. Livers were dissected into their five individual lobes as previously described (24). Some lobes were further divided and pieces subjected to formalin or Carnoy fixation with subsequent paraffin embedding instead of cryopreservation. To exclude lobe to lobe variabilities all analyses were carried out on defined liver lobes.
Serum analysis
Serum was prepared after 2 h of coagulation at room temperature by collection of the supernatant after centrifugation at 2000 x g for 6 min. Serum samples were frozen and stored at -80 C until analysis. The following parameters were determined using standard serum diagnostic methods: Na, K, glucose, bilirubin, total protein, albumin, triglycerides, cholesterol, cholinesterase, aspartate aminotransferase, and alanine aminotransferase.
Plasma was prepared from citrate-blood by collection of the supernatant after centrifugation at 3000 x g for 6 min. Thrombin time was determined using standard laboratory diagnostic methods.
Histological analysis
To exclude a possible bias in the histological analysis, liver sections were encoded and examined without knowledge of which treatment the animals had received.
Five-micrometer paraffin sections of the liver were stained with hematoxylin and eosin and subjected to histopathological examination. Any abnormalities were recorded. Mitotic figures and necrotic and apoptotic cells were quantitatively recorded, applying strict adherence to established criteria for recognition and counting (28, 29, 30, 31). Between 4000 and 8000 hepatocytes/animal were inspected. The mitotic index is defined as the number of mitotic figures per 100 hepatocytes. The cell death index is the cumulative index of both types of cell death (apoptosis as well as necrosis) per 100 unaltered hepatocytes. The apoptotic index indicates the number of apoptotic cells per 100 hepatocytes.
BrdU incorporation into liver cell DNA was analyzed by immunohistochemistry. Briefly, 5-µm sections were deparaffinated, rehydrated, and subjected to immunohistochemistry as previously described (24). A segment of duodenum from each animal was stained in parallel to verify the successful systemic administration of BrdU. Between 10,000 and 20,000 hepatocyte nuclei per animal were inspected microscopically and labeled hepatocyte nuclei registered. Any labeled nonparenchymal cell nuclei within the same area were also recorded. The labeling index is defined as the number of labeled hepatocyte nuclei or labeled nonparenchymal cell nuclei per 100 hepatocyte nuclei, respectively.
DNA analysis
Liver DNA content was determined essentially as described by Labarca and Paigen (32). A 40- to 80-mg piece of liver was thawed, accurately weighed, and homogenized with an ultraturrax in 9 ml phosphate-saline buffer [50 mM NaPO4 (pH 7.4), 2 M NaCl, and 2 mM EDTA] per gram of liver. Aliquots of the homogenate were further diluted 33.3-fold with phosphate-saline buffer and SDS was added to 0.01% final concentration. After sonication (10 sec) the homogenate was diluted with 1 volume of phosphate-saline buffer containing 2 µg/ml Hoechst 33258 (Riedel-de Haën, Seelze, Germany), and the fluorescence was measured with a fluorometer (356 nm for excitation, 458 nm bandpass filter for emitted light). Measurements were calibrated using calf thymus DNA.
Gene expression analysis
Methods for the biochemical and histochemical analysis of lacZ gene expression in the liver were recently described (24).
Total RNA was isolated and 45 µg were analyzed by ribonuclease protection as previously described (13, 33, 34). Probes complementary to the following regions of the respective cDNAs were used (nucleotide positions generally refer to the A of the ATG initiation codon as +1): rat activin ßA (268 bp; corresponding to nucleotides (-)33 to 235) (25); rat activin ßC (196 bp; 861 to 1056) (13); rat activin ßE (169 bp; 470 to 638) (13); mouse glyceraldehyde 3-phosphate dehydrogenase (101 bp; 866 to 966) (35). Gels were analyzed with a PhosphorImager and the ImageMaster 1D Elite image analysis software (Amersham Biosciences, Uppsala, Sweden) (34).
Serum and liver activin A concentrations were measured with an activin A-specific ELISA (Serotec, Oxford, UK). The homogenates used for DNA quantitation were also employed in the determination of the activin A content of the liver. All dilutions of samples and standards were carried out in 0.1% BSA to minimize adsorptive losses.
Statistical analysis
Except where otherwise noted, mean and SD of five animals per group are shown. Activin expressing and control (lacZ) groups were compared by the two-tailed unpaired t test; P 0.05 or less was considered significant (*, P 
0.05; **, P 0.01; ***, P 0.001).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
) were transferred into mice by rapid high-volume tail-vein injection of naked plasmid DNA. This technique results in a rapid high-level expression of the transgene in the liver as the predominant organ of DNA uptake by selective transfection of a fraction of hepatocytes (24). A plasmid dose yielding maximal transgene expression was chosen. Based on the previously established kinetics of reporter gene expression with an early peak at 8 h after gene transfer but a rapid decline thereafter (24), 24 h, 2 d, and 6 d after gene transfer were selected as the time points for analysis of the effects of the transient activin overexpression. Animals transfected with an E. coli ß-galactosidase (lacZ)-encoding plasmid (Fig. 1B) served as control for side effects inherent to the gene transfer procedure, and activin A overexpression was analyzed for comparison at 2 d after its transfection.
Because lacZ was also linked to each activin ß-subunit cDNA via an IRES (Fig. 1A), the analysis of transgene expression in tissue sections was straightforward. As exemplified in Fig. 2, distribution and number of transfected hepatocytes were similar in activin and control groups as well as to previously reported results (24). One day after gene transfer about 6% of the hepatocytes stained positive for ß-galactosidase in all groups without any obvious zonal distribution. Reflecting the different mode of translation initiation, IRES linked lacZ expression in activin transfected mice was almost 10-fold lower than expression directly driven by the CMV promoter in control animals (data not shown). ß-Galactosidase levels declined to roughly 25% and 10% of their 24-h level after 2 and 6 d, respectively (data not shown).
|
). Surprisingly, high levels of rat activinßC and ßE mRNA at 24 h past transfection were followed by an approximately 100-fold decline at 2 d yet only a 2-fold further drop by 6 d. Activin ß-subunit mRNA (Fig. 3) and ß-galactosidase levels (data not shown) both confirmed the high reproducibility of the gene transfer procedure; maximal variability did not exceed a factor of 2 by 24 h post gene transfer.
|
). Two days after gene transfer, activin A levels were elevated 2-fold in serum and 3-fold in the liver.
|
Animal health status, organ and serum analysis
Body weight and food consumption of each mouse were recorded daily, starting from 2 wk before gene transfer and continuing throughout the study. One day after gene transfer, food consumption was generally reduced (Fig. 4). Reduction was more pronounced in activin-expressing animals and significantly exceeded that of the lacZ control in the activin E and A overexpressing groups. Body weight was also reduced in both latter groups (98*** ± 1% and 96*** ± 1% of the average body weight 3 d before gene transfer). Body weight as well as food consumption, however, were back to near normal 1 d later. Overall appearance and behavior of the animals did not indicate any other apparent abnormalities induced by gene transfer or activin overexpression. Viability was 100%.
|
Necropsy and tissue dissection revealed the liver as the only organ displaying gross alterations induced by gene transfer and/or activin overexpression. Relative liver weight was considerably increased 1 d after gene transfer (Fig. 5A), probably as a result of the rapidly injected high volume (22, 24). However, relative liver weight of activin A transfected animals, analyzed 2 d after gene transfer, was significantly reduced, compared with the lacZ control group. Livers of activin C- and E-transfected animals showed a similar yet statistically nonsignificant reduction with respect to lacZ mice at this time. After 6 d, relative liver weight of all groups was restored to normal. To analyze whether these changes in liver weight were associated with alterations in hepatic cell number, the relative and absolute DNA content of the liver were determined. One and two days post injection, the relative DNA content was reduced in all groups by a similar amount (Fig. 5B; reduction was slightly more pronounced on 1 d post injection, mean relative DNA content being approximately 3.25 mg/g). Calculation of absolute DNA content furthermore demonstrated that the increased liver weights of different groups at 1 and 2 d past gene transfer were not associated with hyperplasia; the total amount of DNA per liver was unaltered by the gene transfer procedure (Fig. 5C; data for 1 d past gene transfer not shown). However, the reduced liver weight of activin A overexpressing mice was accompanied by a loss of DNA and thus apparently was due to cell loss and not to hypotrophic tissue shrinkage. A similar, though less pronounced trend was also observed in activin E-transfected mice. Six days after gene transfer DNA content like liver weight was within normal range (data not shown).
|
) were registered and covered by a cumulative cell death index (Fig. 7A) in comparison with the index of morphologically typical apoptosis (Fig. 7B). Painstaking histological examination and counting based on established morphological criteria was found to be the most appropriate discriminative method for a quantitative analysis of these two types of cell death (29, 30, 31).
|
|
). Activin C- and E-transfected mice had similar cell death rates 1 d post transfection, but the decrease was less pronounced and apoptoses were still registered after 6 d (Fig. 7). About half of the cell death on d 1 displayed characteristic apoptotic morphology (Fig. 6). The relative frequency of apoptosis increased on d 2 (Fig. 7). Activin A overexpression induced almost triple the amount of hepatic cell death, primarily by necrosis, compared with control mice (Fig. 7). Quantitative histological analysis thus further substantiated the biochemical evidence for cell loss induced by activin A.
Regenerative proliferation of hepatic cells
Because the liver typically responds to damage by regeneration, our gene transfer model allowed us to study the effects of activin overexpression in the context of an inherent regenerative proliferation of hepatic cells. Administration of BrdU identified cells undergoing DNA synthesis at the time of the animal’s death; the DNA-labeling index of hepatocytes and of nonparenchymal hepatic cells was determined (Fig. 8, A and C). Moreover, mitotic figures of hepatocytes were recorded from histological sections (Fig. 8B). Open bars in Fig. 8A show the extent and kinetics of regenerative DNA synthesis of hepatocytes as a result of gene transfer associated tissue damage (lacZ control), which was previously described in detail (24). Typically, regenerative DNA synthesis of nonparenchymal cells displayed a time lag (Fig. 8C).
|
). While activin C overexpression did not affect DNA synthesis 1 d after gene transfer, notably less hepatocytes were undergoing DNA synthesis in activin E-transfected livers. Yet 2 d after gene transfer the effects of activin expression were most striking. Activin C, E, and A all inhibited significantly the regenerative DNA synthesis of hepatocytes as well as nonparenchymal cells. Impairment by activin E was most prominent. The frequency of observed mitoses further confirmed the antiproliferative effect. Six days after gene transfer, however, when transgene expression was almost absent, proliferative activity in the livers of activin C- and E-transfected animals was increased, probably as a result of the delayed hepatic regeneration. Thus, activin C, E, and A have the potential to temporarily inhibit DNA synthesis in an in vivo experimental system. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Surplus amounts of a growth factor to study gain-of-function effects in an animal may be achieved by either application of the recombinant protein or introduction of a transgene. In the case of activin C or E, the first is not yet an option; however, germ line gene transfer and gene delivery to the adult animal in mice are both possible. Although the model we chose did not yield a sustained expression of recombinant activins in the liver, reproducible temporary high levels of apparently functional activin C, E, and A were achieved using minimal means. The method is based on the rapid injection of a high volume of plasmid DNA solution into the tail-vein of mice leading to a retrograde perfusion of the liver (22, 23, 24). Transgenes are almost exclusively expressed in the liver, although the mechanism of DNA uptake by hepatocytes is unclear. Kinetics of transgene expression as well as side effects of the gene transfer procedure are nevertheless predictable (24). Gene transfer causes a single wave of transgene expression with a peak around 8–10 h post injection; thereafter expression rapidly declines. Among the by-products of the high-volume injection procedure, the distinct damage of liver tissue leading to regeneration was the most important to consider. Peak of regenerative DNA synthesis of hepatocytes occurs around 24 h, but other cells within the liver generally enter into DNA synthesis 24 h later (36). Based on this knowledge, 24 h (1 d), 2 d, and 6 d were chosen as time points of analysis.
Did the gene transfer result in the expected overexpression of the transgenes? Means to directly determine activin C or E levels in serum or tissues are currently not available; however, mRNA analysis and reporter gene coexpression both indicated a reproducible temporal and spatial expression pattern. Furthermore, because similar transcription and translation rates can reasonable be assumed for all three activin expression cassettes transfected, the quantitative analysis of activin A levels allows suggestion of an idea of the activin C and E levels possibly achieved in liver and serum. Two days after gene transfer activin A levels were 2.7 ng/ml in serum and 55 ng/g in the liver, still 2- and 3-fold higher than control. Given the short half-life of activin A in serum (37) together with the observed mRNA expression pattern of activin transgenes, at least 10 times the amount of activins may be assumed 1 d after gene transfer. Yang et al. (38), using a similar gene transfer approach, recently reported levels of hepatocyte growth factor as high as 50 µg/g liver at 8 h after injection. However, such estimations of activin C or E concentration are indirect and direct quantitative assays are required to determine their exact degree of overexpression. Our experimental observations of the effects on liver tissue homeostasis nevertheless indicate a successful overexpression of biologically active activin C and E.
Activin C, E, and A temporarily inhibited regenerative DNA synthesis in the liver. Probably because of the offset between the regeneration initiating stimulus and the onset and peak of transgene expression, this effect became significant after only 2 d, although hepatocyte DNA synthesis in the activin E group was already decreased by d 1. The inhibitory effect on the proliferation of nonparenchymal cells may be indirect because they generally lag behind in liver regeneration (36). Inhibition was most pronounced in activin E-transfected animals; however, because no quantitative information on activin C and E protein levels is available, subtle quantitative differences in the effects of their overexpression have to be interpreted cautiously. Moreover, the increased cell loss because of activin A overexpression may have led to a secondary regenerative stimulus. The decline in transgene expression as well as activin turnover probably account for the apparent later relief from activin inhibition, although regenerating hepatocytes might also have the potential to overcome the action of a growth inhibitor. Even continuous supply of TGF-ß, a strong inhibitor of hepatocyte DNA synthesis, failed to permanently inhibit regenerative proliferation (39). Probably as a result of delayed regeneration and tissue remodeling, signs of proliferation and cell death were still increased in activin transfected animals 6 d past gene transfer in our study.
The temporary block of hepatocyte proliferation might have also resulted in the recruitment of the oval stem cell compartment (40, 41). However, no histological evidence for oval cell proliferation was noted in our experiments. Apparently, neither the regeneration initiating stimulus nor the inhibitory action on hepatocyte proliferation were strong or long lasting enough to switch to a stem cell type of liver regeneration (40, 41).
Using the same gene transfer technology and vector system, we also succeeded in (conversely) increasing the regenerative response to tissue damage by expressing follistatin, a liver growth factor (Rossmanith, W., and M. Chabicovsky, unpublished observations): an elevated rate of DNA synthesis induced by follistatin overexpression, compared with control, corroborates that the inhibitory effect of activins is not unspecific; in fact, it verifies the specific biological activity of activins expressed in our experimental setting.
This is the first study to demonstrate the inhibitory effect of activins on hepatic DNA synthesis in vivo. Previously, TGF-ß has been shown to inhibit the early proliferative response in liver regeneration (39). Both, TGF-ß and activin A also inhibit hepatocyte DNA synthesis in various in vitro systems (for review see Refs. 20 and 21). Using the transfection-overexpression approach in vitro, we recently found that activin C and E also significantly impair the proliferation of human and rat hepatoma cell lines (Vejda, S., N. Erlach, B. Peter, W. Rossmanith, M. Grusch, and R. Schulte-Hermann, manuscript submitted) (16). Antiproliferative activity was lower than that of activin A, but comparisons were limited again by lack of quantitative information about activin levels.
TGF-ß and activin A, in addition to their inhibitory action, also induce hepatocyte cell death (for review see Refs. 20 and 21). Massive cell loss as a result of activin A overexpression was observed in our model (Fig. 5). Increases in cell death relative to control appeared to be due primarily to necrosis and not apoptosis (Fig. 7). However, part of the observed necroses may represent secondary necrosis because of massive apoptosis exceeding the phagocytotic capacity of neighboring cells (29, 42). Nevertheless, necroses have also been reported in an other model of activin overexpression: gonadal tumors overproducing activin A and activin B in inhibin
deficient mice (43).
Do activin C or E induce cell death? Several of our results indicate a similar capability for cell death induction as demonstrated for activin A, i.e. decreased relative liver weight and DNA content and increased cell death index but probably as a result of the background of tissue damage induced by the gene transfer procedure these parameters did not reach statistical significance for activin C or E. Above-mentioned in vitro studies, however, did indicate a cell death-inducing capacity for activin C and E: both liver activins induced apoptosis in the different hepatoma cell lines (Vejda, S., N. Erlach, B. Peter, W. Rossmanith, M. Grusch, and R. Schulte-Hermann, manuscript submitted) (16). Other more appropriate models are probably required to confirm this finding in vivo.
Further observations similarly suggest an involvement of activin C and E in the induction of cell death in the presented model. Loss of liver tissue by partial hepatectomy reduces the feeding activity of rodents (44). Food consumption was also reduced by gene transfer associated tissue damage. The reduction was significantly more pronounced in activin A transfected animals (Fig. 4), thus correlating with more extensive liver tissue damage. Activin C and E overexpressing animals displayed intermediate behavior. Unless activins directly affect feeding behavior, reduced feeding and concomitant reduced body weight may well reflect increased loss of liver mass.
This study reveals a biological activity for the two hitherto orphan growth factors activin C and E; furthermore, the proliferation-inhibiting activity of activin A is demonstrated in vivo, thereby confirming previous studies on isolated hepatocytes (45). The three activins, with their potential to reversibly inhibit hepatocyte DNA synthesis are probably involved in liver growth regulation. The liver, although in the adult animal normally mitotically quiescent, is an organ with a broad potential to adjust its size to changing physiological requirements by either proliferation or cell loss. Although the study of the proliferative response in regeneration has helped to understand the role of growth factors involved in proliferation initiation, little is known about the negative regulators that eventually stop proliferation or initiate tissue regression after hyperplasia (for review see Refs. 20, 36, 46). The candidate list, so far including TGF-ß and activin A, should now be supplemented by activin C and E.
Activins appear to be further involved in maintaining hepatocytes in the liver in a quiescent state, although evidence for this hypothesis is indirect. Follistatin, a protein-sequestering activin from its signaling pathway (14, 47), is capable of initiating a proliferative response in the liver (48, 49). Similarly, transient overexpression of follistatin after hydrodynamics-based gene transfer, boosted regenerative DNA synthesis (Rossmanith W., and M. Chabicovsky, unpublished observations). Moreover, liver tumors may use follistatin overexpression to escape activin growth control (34).
With three activins and TGF-ß, liver growth regulation appears to involve an excess of negative regulators. However, preliminary gene expression studies under various physiological and unphysiological conditions suggest that even the closely related activins C and E are not fully redundant (16) (Rossmanith, W., unpublished observations), although this conception is at odds with the results of gene knockout experiments, apparently indicating that activin C and E are not essential for liver development or function (17). Thus, our picture of the biological role of activins in the liver remains vague. The elucidation of the biological activities of activin C and E in this study has nevertheless set an important first step toward a better understanding of this role.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Present address for M.C.: Igeneon Immunotherapy of Cancer AG, 1230 Vienna, Austria.
Abbreviations: BrdU, 5-Bromo-2'-deoxyuridine; CMV, cytomegalovirus; IRES, internal ribosomal entry site.
Received March 27, 2003.
Accepted for publication April 14, 2003.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
- and ß-subunit complementary deoxyribonucleic acids and expression in the ovary. Mol Endocrinol 1:561–568[Abstract]
2-Macroglobulin conformation determines binding affinity for activin A and plasma clearance of activin A/ 2-macroglobulin complex. Endocrinology 136:5343–5349[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  Y.-G. Chen, Q. Wang, S.-L. Lin, C. D. Chang, J. Chung, and S.-Y. Ying Activin Signaling and Its Role in Regulation of Cell Proliferation, Apoptosis, and Carcinogenesis. Experimental Biology and Medicine, May 1, 2006; 231(5): 534 - 544. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 U. Muenster, C. A. Harrison, C. Donaldson, W. Vale, and W. H. Fischer An Activin-A/C Chimera Exhibits Activin and Myostatin Antagonistic Properties J. Biol. Chem., November 4, 2005; 280(44): 36626 - 36632. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E J Gold, X Zhang, A M Wheatley, S L Mellor, M Cranfield, G P Risbridger, N P Groome, and J S Fleming {beta}A- and {beta}C-activin, follistatin, activin receptor mRNA and {beta}C-activin peptide expression during rat liver regeneration J. Mol. Endocrinol., April 1, 2005; 34(2): 505 - 515. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Gaedeke, T. Boehler, K. Budde, H.-H. Neumayer, and H. Peters Glomerular activin A overexpression is linked to fibrosis in anti-Thy1 glomerulonephritis Nephrol. Dial. Transplant., February 1, 2005; 20(2): 319 - 328. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Wada, A. Maeshima, Y.-Q. Zhang, Y. Hasegawa, H. Kuwano, and I. Kojima Assessment of the function of the {beta}C-subunit of activin in cultured hepatocytes Am J Physiol Endocrinol Metab, August 1, 2004; 287(2): E247 - E254. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Harrison, P. C. Gray, W. H. Fischer, C. Donaldson, S. Choe, and W. Vale An Activin Mutant with Disrupted ALK4 Binding Blocks Signaling via Type II Receptors J. Biol. Chem., July 2, 2004; 279(27): 28036 - 28044. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Vejda, N. Erlach, B. Peter, C. Drucker, W. Rossmanith, J. Pohl, R. Schulte-Hermann, and M. Grusch Expression of activins C and E induces apoptosis in human and rat hepatoma cells Carcinogenesis, November 1, 2003; 24(11): 1801 - 1809. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
