Search

Supplementary Figure Legend from Angiotensinogen Delays Angiogenesis and Tumor Growth of Hepatocarcinoma in Transgenic Mice

Angiotensinogen, a member of the serpin family, is involved in the suppression of tumor growth and metastasis. To investigate whether human angiotensinogen protects against tumor progression in vivo, we established an original bitransgenic model in which transgenic mice expressing human angiotensinogen (Hu-AGT-TG mice) were crossed with a transgenic mouse model of hepatocellular carcinoma (HCC-TG mice). Bitransgenic mice overexpressing human angiotensinogen (HCC/Hu-AGT-TG) had a significantly longer survival time than the HCC-TG mice and a reduction of both tumor growth and blood flow velocities in the liver. This antitumor effect of angiotensinogen is related to a reduced angiogenesis, impaired expression of endothelial arterial markers (active Notch4, Delta-like 4 ligand, and ephrin B2) with a decrease of arterial vessel density in HCC/Hu-AGT-TG mice liver. Overexpression of human angiotensinogen decreases angiogenesis, and prevents tumor sinusoids from remodeling and arterialization, thus delaying tumor progression in vivo. [Cancer Res 2009;69(7):2853–60]

Supplementary Figure 1 from Angiotensinogen Delays Angiogenesis and Tumor Growth of Hepatocarcinoma in Transgenic Mice

Background: Hereditary Hemorrhagic Telangiectasia type 2 (HHT2) is an inherited genetic disorder characterized by vascular malformations and hemorrhage. HHT2 results from ACVRL1 haploinsufficiency, the remaining wild-type allele being unable to contribute sufficient protein to sustain endothelial cell function. Blood vessels function normally but are prone to respond to angiogenic stimuli, leading to the development of telangiectasic lesions that can bleed. How ACVRL1 haploinsufficiency leads to pathological angiogenesis is unknown. Methods: We took advantage of Acvrl1 +/− mutant mice that exhibit HHT2 vascular lesions and focused on the neonatal retina and the airway system after Mycoplasma pulmonis infection, as physiological and pathological models of angiogenesis, respectively. We elucidated underlying disease mechanisms in vitro by generating Acvrl1 +/− mouse embryonic stem cell lines that underwent sprouting angiogenesis and performed genetic complementation experiments. Finally, HHT2 plasma samples and skin biopsies were analyzed to determine whether the mechanisms evident in mice are conserved in humans. Results: Acvrl1 +/− retinas at postnatal day 7 showed excessive angiogenesis and numerous endothelial “tip cells” at the vascular front that displayed migratory defects. Vascular endothelial growth factor receptor 1 (VEGFR1; Flt-1) levels were reduced in Acvrl1 +/− mice and HHT2 patients, suggesting similar mechanisms in humans. In sprouting angiogenesis, VEGFR1 is expressed in stalk cells to inhibit VEGFR2 (Flk-1, KDR) signaling and thus limit tip cell formation. Soluble VEGFR1 (sVEGFR1) is also secreted, creating a VEGF gradient that promotes orientated sprout migration. Acvrl1 +/− embryonic stem cell lines recapitulated the vascular anomalies in Acvrl1 +/− (HHT2) mice. Genetic insertion of either the membrane or soluble form of VEGFR1 into the ROSA26 locus of Acvrl1 +/− embryonic stem cell lines prevented the vascular anomalies, suggesting that high VEGFR2 activity in Acvrl1 +/− endothelial cells induces HHT2 vascular anomalies. To confirm our hypothesis, Acvrl1 +/− mice were infected by Mycoplasma pulmonis to induce sustained airway inflammation. Infected Acvrl1 +/− tracheas showed excessive angiogenesis with the formation of multiple telangiectases, vascular defects that were prevented by VEGFR2 blocking antibodies. Conclusions: Our findings demonstrate a key role of VEGFR1 in HHT2 pathogenesis and provide mechanisms explaining why HHT2 blood vessels respond abnormally to angiogenic signals. This supports the case for using anti-VEGF therapy in HHT2.

Objective— To determine the role of Gja5 that encodes for the gap junction protein connexin40 in the generation of arteriovenous malformations in the hereditary hemorrhagic telangiectasia type 2 (HHT2) mouse model. Approach and Results— We identified GJA5 as a target gene of the bone morphogenetic protein-9/activin receptor-like kinase 1 signaling pathway in human aortic endothelial cells and importantly found that connexin40 levels were particularly low in a small group of patients with HHT2. We next took advantage of the Acvrl1 +/− mutant mice that develop lesions similar to those in patients with HHT2 and generated Acvrl1 +/− ; Gja5 EGFP/+ mice. Gja5 haploinsufficiency led to vasodilation of the arteries and rarefaction of the capillary bed in Acvrl1 +/− mice. At the molecular level, we found that reduced Gja5 in Acvrl1 +/− mice stimulated the production of reactive oxygen species, an important mediator of vessel remodeling. To normalize the altered hemodynamic forces in Acvrl1 +/− ; Gja5 EGFP/+ mice, capillaries formed transient arteriovenous shunts that could develop into large malformations when exposed to environmental insults. Conclusions— We identified GJA5 as a potential modifier gene for HHT2. Our findings demonstrate that Acvrl1 haploinsufficiency combined with the effects of modifier genes that regulate vessel caliber is responsible for the heterogeneity and severity of the disease. The mouse models of HHT have led to the proposal that 3 events—heterozygosity, loss of heterozygosity, and angiogenic stimulation—are necessary for arteriovenous malformation formation. Here, we present a novel 3-step model in which pathological vessel caliber and consequent altered blood flow are necessary events for arteriovenous malformation development.

Angiotensinogen, a member of the serpin family, is involved in the suppression of tumor growth and metastasis. To investigate whether human angiotensinogen protects against tumor progression in vivo, we established an original bitransgenic model in which transgenic mice expressing human angiotensinogen (Hu-AGT-TG mice) were crossed with a transgenic mouse model of hepatocellular carcinoma (HCC-TG mice). Bitransgenic mice overexpressing human angiotensinogen (HCC/Hu-AGT-TG) had a significantly longer survival time than the HCC-TG mice and a reduction of both tumor growth and blood flow velocities in the liver. This antitumor effect of angiotensinogen is related to a reduced angiogenesis, impaired expression of endothelial arterial markers (active Notch4, Delta-like 4 ligand, and ephrin B2) with a decrease of arterial vessel density in HCC/Hu-AGT-TG mice liver. Overexpression of human angiotensinogen decreases angiogenesis, and prevents tumor sinusoids from remodeling and arterialization, thus delaying tumor progression in vivo. [Cancer Res 2009;69(7):2853–60]

Angiogenesis is a complex process, requiring a finely tuned balance between numerous stimulatory and inhibitory signals. ALK1 (activin receptor like-kinase 1) is an endothelial-specific type 1 receptor of the transforming growth factor-β receptor family. Heterozygotes with mutations in the ALK1 gene develop hereditary hemorrhagic telangiectasia type 2 (HHT2). Recently, we reported that bone morphogenetic protein (BMP)9 and BMP10 are specific ligands for ALK1 that potently inhibit microvascular endothelial cell migration and growth. These data lead us to suggest that these factors may play a role in the control of vascular quiescence. To test this hypothesis, we checked their presence in human serum. We found that human serum induced Smad1/5 phosphorylation. To identify the active factor, we tested neutralizing antibodies against BMP members and found that only the anti-BMP9 inhibited serum-induced Smad1/5 phosphorylation. The concentration of circulating BMP9 was found to vary between 2 and 12 ng/mL in sera and plasma from healthy humans, a value well above its EC 50 (50 pg/mL). These data indicated that BMP9 is circulating at a biologically active concentration. We then tested the effects of BMP9 in 2 in vivo angiogenic assays. We found that BMP9 strongly inhibited sprouting angiogenesis in the mouse sponge angiogenesis assay and that BMP9 could inhibit blood circulation in the chicken chorioallantoic membrane assay. Taken together, our results demonstrate that BMP9, circulating under a biologically active form, is a potent antiangiogenic factor that is likely to play a physiological role in the control of adult blood vessel quiescence.

Angiotensin (Ang) II is produced locally in various tissues, but its role in the regulation of tissue metabolism is still unclear. Recent studies have revealed the role of type 2 Ang II receptor (AT2R) in the control of energy homeostasis and lipid metabolism. The contribution of the AT2R to adaptation to starvation was tested using AT2R-deficient (AT2R (y)(/-)) mice. Fasted AT2R (y)(/-) mice exhibited a lower loss of adipose tissue weight associated to a decreased free fatty acid (FFA) release from stored lipids than the controls. In vitro studies show that Ang II causes an AT1R-mediated antilipolytic effect in isolated adipocytes. AT1R expression is up-regulated by fasting in both genotypes, but the increase is more pronounced in AT2R (y/-) mice. In addition, the increased muscle beta-oxidation displayed in AT2R (y/-) mice on a fed state, persists after fasting compared with wild-type mice. In liver from fed mice, AT2R deficiency did not modify the expression of genes involved in fatty acid oxidation. However, in response to fasting, the large increase of the expression of this subset of genes exhibited by wild-type mice, was impaired in AT2R (y/-) mice. Taken together, decreased lipolytic capacity and increased muscle fatty acid oxidation participate in the decreased plasma FFA observed in fasted AT2R (y/-) mice and could account for the lower FFA metabolism in the liver. These data reveal an important physiological role of AT2R in metabolic adaptations to fasting.

We have shown previously that angiotensinogen, like other members of the serine protease inhibitor family, has antiangiogenic properties in vitro on cultured endothelial cells and in ovo in the chick chorioallantoic membrane assay. The aim of this study was to show the effects of angiotensinogen on vascular wall remodeling in vivo. We measured the vessel wall thickness (tunica media) stained with an antibody to α-actin. In the kidney, arterioles were 21.5% thinner in male transgenic mice overexpressing human angiotensinogen than in male control animals. In other vascular beds, the arterial wall thickness was not affected. By using in situ hybridization and Northern blot analysis, human angiotensinogen expression was detected at a high level in the male kidney and at a much lower level in other organs. There is a relationship between the effect of angiotensinogen on arterial wall thickness and the local expression level of angiotensinogen in this model of transgenic mice. Because human angiotensinogen is not cleaved to a significant extent by mouse renin, the reduction in kidney arterial wall thickness is because of angiotensinogen itself and not angiotensin II, and we show that the reduction was not because of hypoplasia or hypotropia. In contrast, a marked difference in the expression of platelet-derived growth factor receptor-β was observed in the kidney arterioles at day 5 when compared with controls. Altogether, these observations provide the first quantitative evidence that a high level of angiotensinogen expression can inhibit the growth of kidney artery walls in vivo.

The members of the serine protease inhibitor (serpin) family, which share a common tertiary structure and a role as serin protease inhibitors, are involved in a variety of newly discovered functions. For example, antithrombin III exerts a strong antiangiogenic activity. Angiotensinogen, the renin substrate, has a folded structure and is a member of the noninhibitory serpin subfamily. Two other noninhibitory serpins, maspin and pigment epithelium-derived factor, have antiangiogenic properties. We investigated the antiangiogenic effect of angiotensinogen and 2 related compounds: (1) des(angiotensin I)angiotensinogen, the product of angiotensinogen cleavage by renin, and (2) the reactive center loop-cleaved angiotensinogen, which is produced after selective and limited proteolysis by the protease V8. We used well-established in vitro (endothelial cell proliferation and migration, and capillary-like tube formation on Matrigel) and in vivo (the chick chorioallantoic membrane assay) models of angiogenesis to evaluate the antiangiogenic activities of these 3 related molecules. Our data demonstrated that these compounds exerted a clear and equipotent antiangiogenic effect, thus attributing a novel function to angiotensinogen and des(angiotensin I)angiotensinogen, for which no function was previously known.

Successful embryo development requires an extensive endometrial angiogenesis in proximity of implantation site. The glycoprotein hCG is produced even before implantation by trophoblast in normal pregnancy. In this manuscript, we demonstrate an angiogenic effect of hCG in several in vivo (chick chorioallantoïc membrane, matrigel plug assay, aortic ring assay) and in vitro experimental models. In contrast, human placental lactogen (hPL) did not display angiogenic properties. LH/hCG receptor was detected in endothelial cells by reverse-transcriptase polymerase chain reaction (RT-PCR) and by Western blotting. In mice aortic ring assay, angiostimulation by hCG was abrogated by deletion of LH/hCG receptor (LuRKO mice). Use of recombinant hCG and anti-hCG antibody (Ab) further confirmed the specificity of this angiogenic activity. By using dibutyryl cAMP, adenylate cyclase, or protein kinase A inhibitors, we demonstrate that hCG-mediated angiogenesis involves adenylyl-cyclase-protein kinase A activation. Addition of hCG to endometrial epithelial epithelial cells, but not to cultured endothelial cells, stimulated vascular endothelial growth factor (VEGF). VEGF and hCG also displayed additive activities. Altogether, these data demonstrate that peritrophoblastic angiostimulation may result from a paracrine dialogue between trophoblast, epithelial, and endothelial cells through hCG and VEGF.

The glycolytic enzyme enolase (EC 4.2.1.11) exists as dimers formed from three structurally related subunits α, β, and γ, encoded by separate genes. The gene encoding the β-subunit is expressed only in striated muscles. We have previously shown that the β-enolase gene belongs to a small subset of muscle-specific genes showing transcriptional activity in cultured myoblasts, prior to withdrawal from the cell cycle. An increase in the level of β-enolase mRNA occurs during terminal differentiation of myoblasts. To investigate the mechanisms underlying this increase, we have simultaneously estimated, under steady state conditions, the rate of synthesis and the stability of β-enolase mRNA in proliferating C2.7 myoblasts as well as in differentiating myotubes. The method used is based on the isolation of newly synthesized RNA from the total RNA pool, following pulse-labeling of intact cells in the presence of 4-thiouridine. The results described here demonstrate a coordinate increase in newly synthesized and total β-enolase mRNA, while the mRNA half-life, about 4 hr, remains unchanged in the course of terminal differentiation. The expression of the gene for insulin-like growth factor-II (IGF-II), a major positive regulator of myogenesis, was analyzed using the same approach. It is concluded that the up-regulation of β-enolase as well as IGF-II gene expression in differentiating muscle cells reflects an increased rate of entry of newly synthesized mRNAs into the general pool of transcripts without changes in their respective half-lives. © 1995 Wiley-Liss, Inc.

Enolase (EC 4.2.1.11) is a glycolytic enzyme active as a dimer. In adult brain extracts, three forms, alpha alpha, alpha gamma and gamma gamma, have been described, with the alpha gamma hybrid accounting for 30% of total enolase activity (Fletcher et al., Dev Biol 65:462-475, 1978; Lucas et al., Dev Neurosci 10:91-98, 1988). Previous biochemical studies strongly suggest that this hybrid is not generated artefactually during the extraction procedures (Keller et al., J Neurochem 36:1389-1397, 1981; Shimizu et al., BBA 748:278-284, 1983). Immunocytological observations have demonstrated the cell specific localization of the alpha subunit in astrocytes and of the gamma subunit in neurons at the adult stage, but failed to identify a cell type containing both the alpha and gamma subunits necessary for the formation of the alpha gamma hybrid isoform (Ghandour et al., Exp Brain Res 41:271-279, 1981; Vinores et al., J Histochem Cytochem 32:1295-1302, 1984; Iwanaga et al., Arch Histol Cytol [Suppl] 52:13-24, 1989). We sought to approach this question by performing in situ hybridization studies in order to visualize the alpha and gamma mRNAs. In agreement with the immunocytological reports, we observe a specific accumulation of the gamma enolase transcripts in neurons and a high accumulation of alpha enolase transcripts in some glial cells such as the ependymocytes lining the ventricles. Our observations, following hybridization with 35S labeled oligonucleotide specific probes on adjacent thin sections, demonstrate for the first time that transcription of both alpha and gamma enolase genes occurs in many neurons of different brain regions. These results render highly probable the formation of the alpha gamma hybrid in mature neurons. Furthermore, we observe a differential expression of the genes encoding the alpha and gamma enolase subunits in various neuronal populations of the brain. The implications of these observations are discussed.

Antineutrophil cytoplasmic antibody (ANCA)–associated vasculitis exhibits endothelial damage, but the capacity for vessel repair in this disorder is not well understood. Here, we observed a marked increase in serum levels of soluble Flt1 (sFlt1), a potent inhibitor of vascular endothelial growth factor, in patients with active ANCA-associated vasculitis compared with patients during remission and other controls. Serum levels of sFlt1 correlated with C5a, an anaphylatoxin released after complement activation. Serum from patients with acute ANCA-associated vasculitis disrupted blood flow in the chicken chorioallantoic membrane assay, suggesting an antiangiogenic effect. Preincubation with excess human vascular endothelial growth factor prevented this effect. Anti–proteinase-3 (PR3) mAb and serum containing PR3-ANCA from patients with active vasculitis both induced a significant and sustained release of sFlt1 from monocytes, whereas anti-myeloperoxidase (MPO) mAb or polyclonal antibodies did not. However, the serum containing polyclonal PR3-ANCA did not induce release of sFlt1 from cultured human umbilical vein endothelial cells. In summary, these data suggest that anti-PR3 antibodies, and to a much lesser extent anti-MPO antibodies, increase sFlt1 during acute ANCA-associated vasculitis, leading to an antiangiogenic state that hinders endothelial repair.

Serum levels of soluble Flt1 (sFlt1), a potent inhibitor of VEGF, are significantly increased during the acute phase of PR3-AAV (2592 ± 11347 pg/ml) and MPO-AAV (476 ± 4258 pg/ml) compared to controls (118 ± 269 pg/ml). sFlt1 levels decreased during disease remission but remained increased compared to controls in patients with PR3-AAV. sFlt1 serum levels correlated with serum levels of C5a, an anaphylatoxin released following complement activation. Serum from patients with acute PR3-AAV induced the release of sFlt1 by human monocytes in vitro, but failed to induce a similar effect on endothelial cells. Pre-treatment of monocytes with an anti-C5a receptor blocking antibody attenuated sFlt1 release. Serum from patients with acute AAV induced a disruption of blood flow in the chicken chorioallantoic membrane assay and this effect was prevented by incubating patients’ serum with an excess of human VEGF.

Increased angiotensinogen (AGT) production by white adipose tissue has been related to not only obesity but also hypertension. Several studies have highlighted the importance of the angiotensin II type 2 receptor (AT2) in the regulation of blood pressure and fat mass, but the relevance of this transporter in a physiopathological model of increased AGT production, as it occurs in obesity, has not yet been investigated. We used transgenic mice that display either a deletion of AT2 (AT2 KO), an overexpression of AGT (OVEX), or both compound mutants (KOVEX). Results demonstrated that adipocyte hypertrophy and increased lipogenic gene expression induced by adipose AGT overproduction was rescued by deletion of AT2. In line with AGT overexpression, KOVEX and OVEX mice have similar increased plasma AGT levels. However, KOVEX mice display a higher blood pressure than OVEX mice. In kidney, renin expression was clearly reduced in OVEX mice, and its expression was normalized in KOVEX mice. Taken together, we demonstrated that the loss of AT2 expression was sufficient to rescue obesity induced by adipose tissue AGT overexpression and confirmed the necessary role of AT2 for the onset of obesity in this model. Furthermore, despite a reduction of adipose mass in KOVEX, AT2 deficiency caused increased renin production, further worsening the hypertension caused by AGT overexpression. Angiotensin type 2 receptor shows antihypertensive function but promotes the angiotensin II-mediated fat mass enlargement.

International audience; Angiogenesis is essential for tumor growth and metastatic dissemination. We have previously shown that human angiotensinogen (AGT) can in vitro inhibit endothelial cell proliferation and migration, capillary-like tube formation, and neovascularization. To determine whether AGT can exert an antitumoral effect through its antiangiogenic properties, we constructed a recombinant adenovirus carrying the human angiotensinogen gene under the control of the cytomegalovirus promoter (AdAGT). In vitro studies showed that AdAGT selectively inhibited endothelial cell proliferation. In vivo, injections of AdAGT into preestablished human MDA-MB-231 mammary carcinomas in nude mice inhibited tumor growth by 70% compared to controls, with 21% total regression. This effect was associated with the suppression of intratumoral vascularization and marked necrosis. Furthermore, in vitroAdAGT infection of MDA-MB-231 and murine melanoma B16F10 cells strongly blocked their in vivo tumorigenicity. Then, in mice expressing high levels of AGT (i.e., either iv injected with AdAGT or HuAGT transgenic mice), the number of B16F10 pulmonary metastases was 85% lower than in control C57BL/6 mice. Our data demonstrate that AGT is a very potent antiangiogenic factor in vivo, independent of angiotensin II generation. Its delivery by gene transfer represents a promising new strategy to block primary tumor growth and to prevent metastasis.

The chick chorioallantoic membrane (CAM) provides a suitable in vivo model to study angiogenesis and evaluate several pro- and anti-angiogenic factors and compounds. In the present work, new developments in image analysis are used to quantify CAM angiogenic response from optical microscopic observations, covering all vascular components, from the large supplying and feeding vessels down to the capillary plexus. To validate our methodology angiogenesis is quantified during two phases of CAM development (day 7 and 13) and after treatment with an antiangiogenic modulator of the angiogenesis. Our morphometric analysis emphasizes that an accurate quantification of the CAM vasculature needs to be performed at various scales.

Ischemic and solid tumor tissues are less well perfused than normal tissue, leading to metabolic changes and chronic hypoxia, which in turn promotes angiogenesis. We identified human angiopoietin-like 4 (angptl4) as a gene with hypoxia-induced expression in endothelial cells. We showed that the levels of both mRNA and protein for ANGPTL4 increased in response to hypoxia. When tested in the chicken chorioallantoic membrane assay, ANGPTL4 induced a strong proangiogenic response, independently of vascular endothelial growth factor. In human pathology, ANGPTL4 mRNA is produced in ischemic tissues, in conditions such as critical leg ischemia. In tumors, ANGPTL4 is produced in the hypoxic areas surrounding necrotic regions. We observed particularly high levels of ANGPTL4 mRNA in tumor cells of conventional renal cell carcinoma. Other benign and malignant renal tumor cells do not produce ANGPTL4 mRNA. This molecule therefore seems to be a marker of conventional renal cell carcinoma. ANGPTL4, originally identified as a peroxisome proliferator-activated receptor alpha and gamma target gene, has potential for use as a new diagnostic tool and a potential therapeutic target, modulating angiogenesis both in tumors and in ischemic tissues. This study also suggests that ANGPTL4 may provide a link between metabolic disorders and hypoxia-induced angiogenesis.

The cDNA for rabbit muscle-specific (betabeta) enolase was cloned, sequenced and expressed in Escherichia coli. This betabeta-enolase differs at eight positions from that sequenced by Chin (17). Site-directed mutagenesis was used to change residue 414 from glutamate to leucine, thereby abolishing a salt bridge involved in subunit contacts. Recombinant wild-type and mutant enolase were purified from E. coli and compared to enolase isolated from rabbit muscle. Molecular weights were determined by mass spectrometry. All three betabeta-enolases had similar kinetics, and UV and circular dichroism (CD) spectra. The mutant enolase was far more sensitive to inactivation by pressure, by KCl or EDTA, and by sodium perchlorate. We confirmed, by analytical ultracentrifugation, that the sodium perchlorate inactivation was due to dissociation. DeltaG(o) for dissociation of enolase was decreased from 49.7 kJ/mol for the wild-type enzyme to 42.3 kJ/mol for the mutant. In contrast to the wild-type enzyme, perchlorate inactivation of E414L was accompanied by a small loss of secondary structure.

During muscle development, an isozymic transition of the glycolytic enzyme enolase occurs from the embryonic and ubiquitous alphaalpha-isoform to the muscle-specific betabeta-isoform. Here, we demonstrate a stimulatory role of thyroid hormones on these two enolase genes during rat development in hindlimb muscles and an inhibitory effect on the muscle-specific enolase gene in cardiac muscle. In hindlimb muscles the ubiquitous alpha-transcript level is diminished by hypothyroidism, starting at birth. On the contrary, the more abundant muscle-specific beta-transcript is insensitive to hypothyroidism before establishment of the functional diversification of fibers and is greatly decreased thereafter. Our data support the hypothesis of a role of thyroid hormones in coordinating the expressions of contractile proteins and metabolic enzymes during muscle development. The subcellular localization of isoenolases, established here, is not modified by hypothyroidism. Our results underline the specificity of action of thyroid hormones, which modulate differentially two isozymes in the same muscle and regulate, in opposite directions, the expression of the same gene in two different muscles.

The glycolytic enzyme enolase (EC 4.2.1.11) is active as dimers formed from three subunits encoded by different genes. The embryonic alphaalpha isoform remains distributed in many adult cell types, whereas a transition towards betabeta and gammagamma isoforms occurs in striated muscle cells and neurons respectively. It is not understood why enolase exhibits tissue-specific isoforms with very close functional properties. We approached this problem by the purification of native betabeta-enolase from mouse hindlimb muscles and by raising specific antibodies of high titre against this protein. These reagents have been useful in revealing a heterogeneity of the beta-enolase subunit that changes with in vivo and in vitro maturation. A basic carboxypeptidase appears to be involved in generating an acidic beta-enolase variant, and may regulate plasminogen binding by this subunit. We show for the first time that pure betabeta-enolase binds with high affinity the adjacent enzymes in the glycolytic pathway (pyruvate kinase and phosphoglycerate mutase), favouring the hypothesis that these three enzymes form a functional glycolytic segment. betabeta-Enolase binds with high affinity sarcomeric troponin but not actin and tropomyosin. Some of these binding properties are shared by the alphaalpha-isoenolase, which is also expressed in striated muscle, but not by the neuron-specific gammagamma-enolase. These results support the idea that specific interactions with macromolecules will address muscle enolase isoforms at the subcellular site where ATP, produced through glycolysis, is most needed for contraction. Such a specific targeting could be modulated by post-translational modifications.

During striated muscle development, the glycolytic enzyme enolase (EC 4.2.1.11) undergoes an isozymic transition, from the embryonic αα form towards the muscle-specific forms αβ and ββ. The regulation of this transition was analyzed in mouse hindlimb muscles from embryonic day 15 (E15) to the adult stage. The quantitative modulations of the levels of the transcripts and subunits of α and β enolase genes were determined. The absolute amounts of α and β enolase mRNAs were estimated using in vitro synthesized transcripts as calibration standards, thus allowing an evaluation of their relative contribution at each stage examined. The muscle-specific β enolase mRNA is already present at E15. Its level then increases and, from E17, this transcript becomes predominant. This accumulation is biphasic: a steep prenatal rise, corresponding to a net increase per fiber, accompanies the formation of secondary myofibers and the development of innervation; a second rise, beginning at postnatal day 5, is temporally correlated with the definitive specialization of the myofibers. Most of the decrease in α mRNA level occurs postnatally. No temporal or quantitative correlation between the up-regulation of β mRNA and the down-regulation of α mRNA levels is observed throughout hindlimb muscle development. Quantitative immunoblotting analyses carried out in parallel show that the enolase isozymic transition is mainly controlled at the mRNA level. The developmental profile of α-skeletal actin mRNA matches the biphasic pattern of β enolase mRNA. Interestingly, the second phase of transcript increases coincides with that of the mRNA of an adult-specific myosin heavy chain, the accumulation of which depends on increasing thyroid hormone levels. It has been proposed that thyroid hormones could coordinate the expression of contractile proteins and enzymes of energy metabolism during the postnatal muscle development. Our results support this notion and suggest a role of thyroid hormones in the regulation of the β enolase gene.

We define the spatial and temporal patterns of expression of the gene encoding the glycolytic enzyme, β-enolase, during mouse ontogenesis. Transcripts were detected by in situ hybridization using 35 S labelled cRNA probes. The β-enolase gene is expressed only in striated muscles. It is first detected in the embryo, in the cardiac tube and in newly formed myotomes. In the muscle masses of the limb, β gene expression occurs at a low level in primary fibers, and subsequently greatly increases at a time which corresponds to the onset of innervation and secondary fiber formation. Later in development, it becomes undetectable in slow-twitch fibers. Our results demonstrate the multistep regulation of the β-enolase gene. The regulation of this muscle-specific gene in somites is discussed in terms of the myogenic sequences of the MyoD family shown to be present when it is activated.

It is now well established that a solid tumor cannot grow beyond a volume of a few mm3 without the formation of new capillaries to supply tumor cells with nutrients and oxygen, and to allow removal of waste products. We have already shown that recombinant human angiotensinogen (AGT) can inhibit endothelial cell proliferation and migration, capillary-like tube formation, and neovascularisation of the chick chorioallantoic membrane1. In order to determine whether these anti-angiogenic activities could impact tumor growth, we have delivered AGT by gene transfer into mice bearing experimental tumors using a replication defective adenovirus (AdAGT).

The possible influence of noradrenaline (NA) upon cell differentiation has been studied by comparing NA-supplemented cultures of foetal pallial cells with control cultures grown in normal medium. Two days after plating, the cultures were processed for immunocytochemical detection of either an adhesion molecule and marker of early stages of neuronal differentiation (L1) or a marker expressed at relatively late stages (γ-enolase). In both cases, the NA supplement reduced the expression of the antigen. The effects were more clear-cut for the late than for the early marker. In conclusion, the NA supplement to the culture medium, in our model, seemed to have a ‘differentiation regulating’ rather than a ‘neurotrophic’ function sensu stricto. It remains to be clarified, however, to which extent this finding can be generalized to in vivo situations.

Using the rabbit reticulocyte cell-free translation system we have estimated during ontogenesis the proportions of in vitro translatable alpha and gamma brain enolase mRNAs, which are two minor mRNA species. No polypeptide precursor to these enzyme subunits appears to be synthesized during translation in vitro. During brain development, the changes in translatable alpha and gamma mRNA content seem to parallel those of the corresponding antigens. The proportion of each of the enolase mRNAs is highest in adult mouse brain. Mechanisms controlling alpha and gamma antigen expression are discussed. In order to prepare the specific cDNA probes, purification of alpha and gamma mRNAs was undertaken.

Resume Les polysomes prepares a partie de poudre de cerveau de rat adulte prealablement congeles ont ete separes sur gradient de sucrose. La traduction in vitro dirigee par les polysomes ainsi fractionnes a ete realisee dans le systeme acellulaire de lysat de reticulocyte. Parmi les produits synthetises in vitro l'enolase specifique du cerveau (proteine 14.3.2) et la proteine S100 ont ete identifiees par immunoprecipitation; la tubuline a ete analysee par electrophorese bidimensionnelle. La synthese relative de ces proteines par l'ensemble des polysomes libres du cerveau de rat represente respectivement 0,1 p. cent, 0,05 p. cent et 0,7 p. cent des produits formes in vitro. Les polysomes responsables de la synthese de la proteine S100 sont separes de ceux dirigeant la formation des deux autres marqueurs par centrifugation sur gradient de sucrose. Ce fractionnement permet d'enrichir trois fois l'activite RNA messager specifique de chacune des proteines etudiees.

An analysis of enolase expression during differentiation of neuroblastoma clones in homogeneous culture is presented. The enolases expressed in these neuroblast-like cells are identical to those of mouse brain with respect to the examined properties. Our biochemical investigation has premitted us to demonstrate formally that neuroblastoma cells undergo a transition from the embryonic αα form to the neuronal γγ form and contain both enolases as well as the αγ hybrid form during maturation. These results suggest that the same phenomenon must exist in vivo for neuroblasts. In neuroblastoma cells, an increase in both αγ and γγ neuron specific enolases is related to cell maturation and expression of the αγ form precedes that of the γγ form during differentiation. Modulation of neuronal enolase activities is similar in the various conditions of differentiation studied and appears not to be necessarily related with morphological differentiation, although concomitant with an arrest of cell division. The evolution of specific neuronal enolases in neuroblastoma cells parallels that observed in vivo , in brain from embryonic day 15 to post-natal day 7. Moreover, at least one treatment (dimethylsulfoxide) causes an important decrease in the high specific αα activity of these cells as occurs in vivo . This enolase can therefore also be considered as a biochemical marker for neuroblastoma maturation. As observed with other markers and other cell types, neuroblastoma cells in culture express an immature biochemical differentiation of the enolase isozymes.

In vertebrates, the glycolytic enzyme enolase (EC 4.2.1.11) is present as homodimers and heterodimers formed from three distinct subunits of identical molecular weight, alpha, beta, and gamma. We report the cloning and sequencing of a cDNA encoding the beta subunit of murine muscle-specific enolase. The corresponding amino acid sequence shows greater than 80% homology with the beta subunit from chicken obtained by protein sequencing and with alpha and gamma subunits from rat and mouse deduced from cloned cDNAs. In contrast, there is no homology between the 3' untranslated regions of mouse alpha, beta, and gamma enolase mRNAs, which also differ greatly in length. The short 3' untranslated region of beta enolase mRNA accounts for its distinct length, 1600 bases. It is known that a progressive transition from alpha alpha to beta beta enolase occurs in developing skeletal muscle. We show that this transition mainly results from a differential regulation of alpha and beta mRNA levels. Analysis of myogenic cell lines shows that beta enolase gene is expressed at the myoblast stage. Moreover, transfection of premyogenic C3H10T1/2 cells with MyoD1 cDNA shows that the initial expression of beta transcripts occurs during the very first steps of the myogenic pathway, suggesting that it could be a marker event of myogenic lineage determination.

A specific event which accompanies the terminal differentiation of most neurons is the isozymic enolase transition from the αα form to the αγ and γγ adult neuronal forms. It is partly expressed during maturation of a mouse neuroblastoma clonal cell line (NIE-115). We demonstrate, in these cell cultures, that the significant increase in the concentration of γ gene product, expressed as αγ enolase during differentiation, is due to a parallel and similar increase in its synthesis. The capacities of poly (A)+ RNA from undifferentiated and differentiated cultures to direct the synthesis of γ gene product were compared in a reticulocyte cell-free protein-synthesizing system. This translating capacity is stimulated in the same proportion as the rate of synthesis of γ antigen in differentiating cell cultures. Therefore the increase in the level of γ protein (expressed mostly as αγ enolase) during differentiation results from the increased translating activity or relative amount of γmRNA.

Expression of the neuronal marker 14-3-2 or NSE (neuron-specific enolase) has been studied during in vitro differentiation of cells in culture. The 14-3-2 protein of neuroblastoma cells is immunologically identical with that found in mouse brain extract. The lack of detectable 14-3-2 in cultures of non-neuronal lines shows that this protein, as has been already shown in vivo, is also a specific marker of neurons in vitro. The presence of 14-3-2 in a differentiated hypothalamic clone—but not in its presumptive precursor—indicates selective initial derepression of 14-3-2. Moreover, modulation of the amount of 14-3-2 already present in dividing neuroblastoma cells is related to the confluent phase of growth or morphological differentiation of neuroblasts. Both mechanisms may be related to the mechanisms underlying initial differentiation and subsequent maturation of neurons in vivo. In dividing neuroblastoma cells modulation of the basal level of 14-3-2 is not necessarily associated with expression of the morphological differentiation, but seems generally concomitant with an arrest of cell division.

We have compared the rodent developmental pattern of the 14-3-2 antigen estimated by a microcomplement fixation technique with that of the cerebral enolases. Chromatographic separation of enolase isozymes on microcolumns demonstrates that the embryonic neuron specific enolase is firstly and mostly represented by the αγ isozyme. The most important increase in 14-3-2 antigen and γγ enolase occurs between post-natal days 7th and 15th. By post-natal day 30, adult levels have been reached. An interesting observation is—during embryonic development—the decrease in the specific activity of the cerebral enolase isozyme αα. This could be explained by the replacement—in neuroblasts—of αα enolase by neuron specific enolase. A comparison between 14-3-2 antigen and neuron specific enolase (γγ) purified by completely different methods is presented. The 14-3-2 antigen exhibits an enolase specific activity comparable to that of purified enzyme and has the same electrophoretic mobility. Antibodies raised against either antigen have an identical specificity. Pre and post-natal developmental pattern in rodent brains are similar for both proteins. Thus neuron specific 14-3-2 antigen is identical to neuron specific enolase. Thus we have precisely described the ontogenic transition between the three cerebral enolase isozymes at the tissue level. This study is completed by the analysis of these transitions at the neuronal cell level, using homogenous cell lines (Part II of this paper).