Effects of resistance training on tendon mechanical properties and rapid force production in prepubertal children.

Children develop lower levels of muscle force, and at slower rates, than adults. Although strength training in children is expected to reduce this differential, a synchronous adaptation in the tendon must be achieved to ensure forces continue to be transmitted to the skeleton with efficiency while minimizing the risk of strain-related tendon injury. We hypothesized that resistance training (RT) would alter tendon mechanical properties in children concomitantly with changes in force production characteristics. Twenty prepubertal children (age 8.9 ± 0.3 yr) were equally divided into control (nontraining) and experimental (training) groups. The training group completed a 10-wk RT intervention consisting of 2-3 sets of 8-15 plantar flexion contractions performed twice weekly on a recumbent calf-raise machine. Achilles tendon properties (cross-sectional area, elongation, stress, strain, stiffness, and Young's modulus), electromechanical delay (EMD; time between the onset of muscle activity and force), rate of force development (RFD; slope of the force-time curve), and rate of electromyographic (EMG) increase (REI; slope of the EMG time curve) were measured before and after RT. Tendon stiffness and Young's modulus increased significantly after RT in the experimental group only (∼29% and ∼25%, respectively); all other tendon properties were not significantly altered, although there were mean decreases in both peak tendon strain and strain at a given force level (14% and 24%, respectively; not significant) which may have implications for tendon injury risk and muscle fiber mechanics. A decrease of ∼13% in EMD was found after RT for the experimental group, which paralleled the increase in tendon stiffness (r = -0.59); however, RFD and REI were unchanged. The present data show that the Achilles tendon adapts to RT in prepubertal children and is paralleled by a change in EMD, although the magnitude of this change did not appear to be sufficient to influence RFD. These findings are of importance within the context of the efficiency and execution of movement.

EphrinB-mediated reverse signalling controls junctional integrity and pro-inflammatory differentiation of endothelial cells.

The EphB/ephrinB receptor-ligand system is pivotal for the development of the embryonic vasculature and for angiogenesis in the adult organism. We observed that (i) the expression of ephrinB2 and ephrinB1 is up-regulated in capillaries during inflammation, that (ii) these ligands are localised on the luminal endothelial surface, and that (iii) they interact with the ephrinB-receptor EphB2 on monocyte/macrophages. This study delineates the impact of ephrinB-mediated reverse signalling on the integrity and proinflammatory differentiation of the endothelium. To this end, in vitro analyses with human cultured endothelial cells reveal that knockdown of ephrinB2 or ephrinB1 impairs monocyte transmigration through the endothelium. While ephrinB2 but not ephrinB1 interacts with PECAM-1 (CD31) in this context, reverse signalling by ephrinB1 but not ephrinB2 elicits a c-Jun N-terminal kinase (JNK)-dependent up-regulation of E-selectin expression. Furthermore, treatment of endothelial cells with soluble EphB2 receptor bodies or EphB2-overexpressing mouse myeloma cells links ephrinB2 to PECAM-1 and induces its Src-dependent phosphorylation while diminishing Src homology phosphotyrosyl phosphatase-2 (SHP-2) activity and increasing endothelial cell permeability. We conclude that extravasation of EphB2 positive leukocyte populations is facilitated by lowering the integrity of endothelial cell junctions and enhancing the pro-inflammatory phenotype of the endothelium through activation of ephrinB ligands.

The tetraspanin network modulates MT1-MMP cell surface trafficking.

The membrane-type 1 matrix metalloproteinase (MT1-MMP) drives fundamental physiological and pathophysiological processes. Among other substrates, MT1-MMP cleaves components of the extracellular matrix and activates other matrix-cleaving proteases such as MMP-2. Trafficking is a highly effective means to modulate MT1-MMP cell surface expression, and hence regulate its function. Here, we describe the complex interaction of MT1-MMP with tetraspanins, their effects on MT1-MMP intracellular trafficking and proteolytic function. Tetraspanins are credited as membrane organizers that form a network within the membrane to regulate the trafficking of associated proteins. In short, we found MT1-MMP to interact with the tetraspanin-associated EWI-2 protein by a yeast two-hybrid screen. Immunoprecipitation analysis confirmed this interaction and further revealed that MT1-MMP also stably interacts with distinct tetraspanins (CD9, CD37, CD53, CD63, CD81, and CD82) and the tetraspanin-like MAL protein. By using different MT1-MMP truncation constructs and mutants, we observed that all tetraspanins and MAL associated with the hemopexin domain of MT1-MMP. Moreover, this interaction was independent of O-glycosylation of MT1-MMP and exclusively occurred in the endoplasmic reticulum. Here, the respective subcellular compartment was identified by fitting the MT1-MMP interaction pattern to a model for post-translational processing of MT1-MMP. In addition, tetraspanins differentially affected the cell surface localization of MT1-MMP, its capacity to activate pro-MMP-2, and the collagen invasion capacity. Interestingly, the degree of tetraspanin-MT1-MMP association did not correlate with its impact on MT1-MMP function. Tetraspanins thus distinctly affect MT1-MMP subcellular localization and function, and may constitute an effective mechanism to control MT1-MMP-dependent proteolysis at the cell surface.

Novel random peptide libraries displayed on AAV serotype 9 for selection of endothelial cell-directed gene transfer vectors.

We have demonstrated the potential of random peptide libraries displayed on adeno-associated virus (AAV)2 to select for AAV2 vectors with improved efficiency for cell type-directed gene transfer. AAV9, however, may have advantages over AAV2 because of a lower prevalence of neutralizing antibodies in humans and more efficient gene transfer in vivo. Here we provide evidence that random peptide libraries can be displayed on AAV9 and can be utilized to select for AAV9 capsids redirected to the cell type of interest. We generated an AAV9 peptide display library, which ensures that the displayed peptides correspond to the packaged genomes and performed four consecutive selection rounds on human coronary artery endothelial cells in vitro. This screening yielded AAV9 library capsids with distinct peptide motifs enabling up to 40-fold improved transduction efficiencies compared with wild-type (wt) AAV9 vectors. Incorporating sequences selected from AAV9 libraries into AAV2 capsids could not increase transduction as efficiently as in the AAV9 context. To analyze the potential on endothelial cells in the intact natural vascular context, human umbilical veins were incubated with the selected AAV in situ and endothelial cells were isolated. Fluorescence-activated cell sorting analysis revealed a 200-fold improved transduction efficiency compared with wt AAV9 vectors. Furthermore, AAV9 vectors with targeting sequences selected from AAV9 libraries revealed an increased transduction efficiency in the presence of human intravenous immunoglobulins, suggesting a reduced immunogenicity. We conclude that our novel AAV9 peptide library is functional and can be used to select for vectors for future preclinical and clinical gene transfer applications.

Age-related changes in mechanical properties of the Achilles tendon.

The stiffness of a tendon, which influences muscular force transfer to the skeleton and increases during childhood, is dependent on its material properties and dimensions, both of which are influenced by chronic loading. The aims of this study were to: (i) determine the independent contributions of body mass, force production capabilities and tendon dimensions to tendon stiffness during childhood; and (ii) descriptively document age-related changes in tendon mechanical properties and dimensions. Achilles tendon mechanical and material properties were determined in 52 children (5-12 years) and 19 adults. Tendon stiffness and Young's modulus (YM) were calculated as the slopes of the force-elongation and stress-strain curves, respectively. Relationships between stiffness vs. age, mass and force, and between YM vs. age, mass and stress were determined by means of polynomial fits and multiple regression analyses. Mass was found to be the best predictor of stiffness, whilst stress was best related to YM (< 75 and 51% explained variance, respectively). Combined, mass and force accounted for up to 78% of stiffness variation. Up to 61% of YM variability could be explained using a combination of mass, stress and age. These results demonstrate that age-related increases in tendon stiffness are largely attributable to increased tendon loading from weight-bearing tasks and increased plantarflexor force production, as well as tendon growth. Moreover, our results suggest that chronic increases in tendon loading during childhood result in microstructural changes which increase the tendon's YM. Regarding the second aim, peak stress increased from childhood to adulthood due to greater increases in strength than tendon cross-sectional area. Peak strain remained constant as a result of parallel increases in tendon length and peak elongation. The differences in Achilles tendon properties found between adults and children are likely to influence force production, and ultimately movement characteristics, which should be explicitly examined in future research.

Can Achilles tendon moment arm be predicted from anthropometric measures in pre-pubescent children?

Muscle-tendon moment arm magnitudes are essential variables for accurately calculating muscle forces from joint moments. Their measurement requires specialist knowledge and expensive resources. Research has shown that the patellar tendon moment arm length is related to leg anthropometry in children. Here, we asked whether the Achilles tendon moment arm (MA(AT)) can be accurately predicted in pre-pubescent children from surface anthropometry. Age, standing height, mass, foot length, inter-malleolar ankle width, antero-posterior ankle depth, tibial length, lower leg circumference, and distances from the calcaneus to the distal head of the 1st metatarsal and medial malleolus were determined in 49 pre-pubescent children. MA(AT) was calculated at three different ankle positions (neutral, 10° plantarflexion, and 10° dorsiflexion) by differentiating tendon excursion, measured via ultrasonography, with respect to ankle angle change using seven different differentiation techniques. Backwards stepwise regression analyses were performed to identify predictors of MA(AT.) When all variables were included, the regression analysis accounted for a maximum of 49% of MA(AT) variance at the neutral ankle angle when a third-order polynomial was used to differentiate tendon excursion with respect to ankle angle. For this condition, foot length and the distance between calcaneus and 1st metatarsal were the only significant predictors, accounting for 47% of the variance (p<0.05). The absolute error associated with this regression model was 3.8±4.4 mm, which would result in significant error (mean=14.5%) when estimating muscle forces from joint moments. We conclude that MA(AT) cannot be accurately predicted from anthropometric measures in children.

Slow wall motion rather than electrical conduction delay underlies mechanical dyssynchrony in postinfarction patients with narrow QRS complex.

INTRODUCTION: The mechanism of mechanical dyssynchrony in postinfarction patients with a narrow QRS complex is not defined but essential for cardiac resynchronization therapy (CRT). METHODS AND RESULTS: Left ventricular electrical activation and subsequent wall motion were recorded for 16 patients with ischemic cardiomyopathy during intrinsic rhythm using a modified NOGA electromechanical mapping system. Ten patients presented mechanical dyssynchrony on tissue Doppler imaging, while 6 patients served as control subjects. The local activation time (LAT) was set by the maximum downslope of the unipolar electrogram. Local wall motion time (LMT) was defined as the time needed for the catheter tip to traverse half of its maximum inward deflection during systole. LAT and LMT were measured relative to the onset of the QRS complex. Electrical activation showed a septal-to-lateral pattern in all patients with a mean endocardial activation time of 65 +/- 13 ms. Control subjects exhibited 97.5% of all LMTs <290 +/- 17 ms. Delayed motion areas (cut-off LMT > 300 ms) showed no slowing of conduction. Wall motion time corrected for differences in electrical activation (LMT-LAT) was significantly longer in delayed (289 +/- 34 ms) than in regular (204 +/- 24 ms) motion areas (P = 0.002). Delayed motion segments were hypokinetic on echocardiography and presented a lower maximum inward motion (9.9 +/- 1.1 mm) compared to regular segments (10.9 +/- 1.2 mm) on electromechanical maps (P = 0.004). Viability, however, was preserved with unipolar and bipolar voltage amplitude >7 mV and >1.5 mV for 79% of all delayed motion areas. CONCLUSION: Dyssynchronous segments of an ischemic myocardium show unimpaired local activation but slow wall motion, thereby limiting the benefit of ventricular preexcitation via CRT.

Hepatic stellate cells display a functional vascular smooth muscle cell phenotype in a three-dimensional co-culture model with endothelial cells.

Hepatic stellate cells (HSCs) are pericytes of liver sinusoidal endothelial cells (LSECs) and activation of HSC into a myofibroblast-like phenotype (called transdifferentiation) is involved in several hepatic disease processes including neovascularization during liver metastasis, chronic and acute liver injury. While early smooth muscle cell (SMC) differentiation markers including SM alpha-actin and SM22alpha are expressed in a variety of non-SMC, expression of late-stage markers is far more restricted. Here, we found that in addition to early SMC markers, activated rat HSC express a large panel of characteristic late vascular SMC markers including SM myosin heavy chain, h1-calponin and h-caldesmon. Furthermore, myocardin, which is present exclusively in SMCs and cardiomyocytes and controls the transcription of a subset of early and late SMC markers, is highly expressed in activated HSC. We further studied activated HSC in a functional three-dimensional spheroidal co-culture system together with endothelial cells (EC). Co-culture spheroids of EC and SMC differentiate spontaneously and organize into a core of SMC and a surface layer of EC representing an inside-outside model of the physiological assembly of blood vessels. Replacing SMC by in vitro activated HSC resulted in a similar organized spheroid with differentiated, von-Willebrand factor producing, surface lining quiescent human umbilical vein endothelial cell and a core of HSC. In an in vitro angiogenesis assay, activated HSC induced quiescence in vascular EC-the hallmark of vascular SMC function. Co-spheroids of LSEC and activated HSC formed capillary-like sprouts in gel angiogenesis assays expressing the vascular EC marker VE-cadherin. Our findings indicate that activated HSC are capable to adapt a functional SMC phenotype and to induce formation of tubular sprouts by LSEC and vascular endothelial cells. Since tumors and tumor metastasis induce HSC activation, HSC may take part in tumor-induced neoangiogenesis by adapting SMC-like functions.

Design and rationale for the Myocardial Stem Cell Administration After Acute Myocardial Infarction (MYSTAR) Study: a multicenter, prospective, randomized, single-blind trial comparing early and late intracoronary or combined (percutaneous intramyocardial and intracoronary) administration of nonselected autologous bone marrow cells to patients after acute myocardial infarction.

BACKGROUND: Previous data suggest that bone marrow-derived stem cells (BM-SCs) decrease the infarct size and beneficially affect the postinfarction remodeling. METHODS: The Myocardial Stem Cell Administration After Acute Myocardial Infarction Study is a multicenter, prospective, randomized, single-blind clinical trial designed to compare the early and late intracoronary or combined (percutaneous intramyocardial and intracoronary) administration of BM-SCs to patients after acute myocardial infarction (AMI) with reopened infarct-related artery. The primary end points are the changes in resting myocardial perfusion defect size and left ventricular ejection fraction (gated single photon emission computed tomography [SPECT] scintigraphy) 3 months after BM-SCs therapy. The secondary end points relate to evaluation of (1) the safety and feasibility of the application modes, (2) the changes in left ventricular wall motion score index (transthoracic echocardiography), (3) myocardial voltage and segmental wall motion (NOGA mapping), (4) left ventricular end-diastolic and end-systolic volumes (contrast ventriculography), and (5) the clinical symptoms (Canadian Cardiovascular Society [CCS] anina score and New York Heart Association [NYHA] functional class) at follow-up. Three hundred sixty patients are randomly assigned into 1 of 4 groups: group A, early treatment (21-42 days after AMI) with intracoronary injection; group B, early treatment with combined application; group C, late treatment (3 months after AMI) with intracoronary delivery; and group D, late treatment with combined administration of BM-SCs. Besides the BM-SCs therapy, the standardized treatment of AMI is applied in all patients. CONCLUSIONS: The Myocardial Stem Cell Administration After Acute Myocardial Infarction Trial is the first randomized trial to investigate the effects of the combined (intramyocardial and intracoronary) and the intracoronary mode of delivery of BM-SCs therapy in the early and late periods after AMI.

Percutaneous endocardial injection of erythropoietin: assessment of cardioprotection by electromechanical mapping.

BACKGROUND: Apart from its well-known stimulation of erythropoiesis, erythropoietin (EPO) exhibits angiogenic and anti-apoptotic effects. These cellular protective effects have also been described in experimental acute myocardial infarction models. We investigated the effects of EPO in a porcine model of chronic progressive myocardial ischaemia. METHODS: At weeks 2 and 6 after implantation of a circumflex ameroid constrictor, endocardial electromechanical NOGA system (Biosense Webster, Inc., California, USA) mapping of the left ventricle, coronary and ventricular angiography, as well as echocardiography were performed. Two weeks after ameroid placement, 13 pigs were randomized with 7 pigs receiving 10.000 U EPO and 6 pigs receiving placebo into the ischaemic region using a NOGA guided percutaneous transendocardial injection catheter, MYOSTAR. After 6 weeks, histology (Masson's Trichrome) was analyzed. RESULTS: Endocardial electromechanical mapping showed an increase of mean unipolar voltage (UV) amplitude in the ischaemic myocardial segments in the EPO-treated animals (8.5 mV pre and 10.6 mV post treatment) and a significantly reduced ischaemic surface area compared to the control group (19% vs. 41%) suggesting a decline in ischaemic injury. Echocardiography revealed 2,2 hypokinetic segments of the lateral wall in the EPO group vs. 3,3 in the control groups. The mean ejection fraction was 64% in the EPO group and 55% in the placebo group. Quantitative histological analysis of the ischaemic regions revealed a reduction of myocardial fibrosis (8% vs. 28%) in the EPO group. CONCLUSION: Endocardial EPO injection may induce cardioprotective effects in hibernating myocardium and may attenuate the progression of ischaemic tissue damage.

Expression pattern of fibroblast growth factors (FGFs), their receptors and antagonists in primary endothelial cells and vascular smooth muscle cells.

Fibroblast growth factors (FGFs) are important angiogenic growth factors. While basic FGF (FGF2) is well established as a potent inducer of angiogenesis much less is known about other FGFs possibly expressed by EC. We investigated the expression of all known FGFs, their main tyrosine kinase receptors and antagonists by RT-PCR analysis in human umbilical vascular endothelial cells (HUVECs) to obtain a complete expression profile of this important growth factor system in model endothelial cells (EC). In addition to FGFR1IIIc, which is considered as the major FGF receptor in EC, HUVECs express similar levels of FGFR3IIIc, detectable amounts of FGFR2IIIc and a new FGF receptor without an intracellular kinase domain (FGFR5). HUVECs express several secreted FGFs, including FGF5, 7, 8, 16 and 18 and two members of the fibroblast growth factor homologous factors (FHFs), not yet reported to be expressed in EC. The expression panel was compared with that obtained from human vascular smooth muscle cells (VSMCs) and human aortic tissue. Human umbilical artery smooth muscle cells (HUASMCs) and HUVECs express the identical FGF receptor and ligand panel implicating that both cell types act, according the FGF signals more as an entity than as individual cell types. Expression of Fgf1, 2, 7, 16 and 18 and the antagonists Sprouty 2,3 and 4 was demonstrated for all analysed cDNAs. The IIIc isoforms of FGFR1 and 2 and the novel FGFR5 were expressed in the aorta, but expression of the FGF receptor 3 was not detected in cDNAs derived from aortic tissue. In the VSMC of rat aortic tissue and in HUASM cultured cells we could demonstrate FGF18 immunoreactivity in the nucleus of the cells. The expression of several secreted FGFs by EC may focus the view more on their paracrine effects on neighbouring cells during tissue regeneration or tumor formation.

Reduction of myocardial scar size after implantation of mesenchymal stem cells in rats: what is the mechanism?

The use of a cellular therapy offers a promising approach for the treatment of heart disease. Besides other precursor cells, bone marrow (BM)-derived stem cells were discovered that migrate into ischemic myocardium and participate in myogenesis as well as angiogenesis. A subpopulation of those are the mesenchymal stem cells (MSC), which may be potential candidates for repairing ischemic heart tissue. MSC are easy to prepare and can be used in an autologous strategy. Here we demonstrate the effect of transplanted MSC in our autologous rat model of myocardial injury. BM was isolated from tibiae and femurs of Wistar rats. After 24 h, the adhering MSC were separated, expanded, retrovirally transduced using green fluorescent protein (GFP), and cloned. A cryo-infarct was generated in the rat hearts, and immediately after this the cells were injected into the border zone of the lesion. After a 10-week follow up, the hearts were excised and the myocardial scar areas were measured using computer-guided morphometry. When comparing transplanted rats (n = 8) with control animals (n = 5) treated rats demonstrated a significant reduction in the width (p < 0.05) of the myocardial scar area. The depth of the scars of the cell therapy rats was less extended (p > 0.05) and the myocardium of these animals was thicker than in the controls (p > 0.05). Immunohistochemical analyses revealed neither evidence of MSC transdifferentiation into cardiomyocytes, nor could an increased neovascularization be found. In conclusion, MSC are responsible for a remarkable reduction of the myocardial scar size in the treated animals. But, whether this strategy is directly transferable to the patient suffering from heart disease has to be determined. In addition, the mechanism by which MSC act in the ischemic heart remains to be determined.

The translating platform paradigm: perturbation displacement waveform alters the postural response.

The translating platform paradigm is widely used to investigate the regulation of upright standing and locomotion. This study investigated how the displacement waveform characteristics underlying the translating platform perturbation are revealed in the resulting postural response. Eight participants experienced a series of backward-directed perturbations using a hydraulically driven forceplate. Two ranges of platform displacement (5 and 15 cm) in combination with two peak velocities (40 and 60 cm/s) were achieved using three distinct waveforms for platform displacement: (a) RAMP: ramp onset and ramp offset, (b) Ramp-to-Parabola (R-P): ramp onset with parabolic offset and (c) SINE: sine-wave onset with sine wave offset. Our findings indicated that the unique and distinctive acceleration and deceleration characteristics that result from the three different platform displacement waveforms significantly altered the postural response to the perturbation.

Blood vessel maturation in a 3-dimensional spheroidal coculture model: direct contact with smooth muscle cells regulates endothelial cell quiescence and abrogates VEGF responsiveness.

Paracrine interactions between endothelial cells (EC) and mural cells act as critical regulators of vessel wall assembly, vessel maturation and define a plasticity window for vascular remodeling. The present study was aimed at studying blood vessel maturation processes in a novel 3-dimensional spheroidal coculture system of EC and smooth muscle cells (SMC). Coculture spheroids differentiate spontaneously in a calcium-dependent manner to organize into a core of SMC and a surface layer of EC, thus mimicking the physiological assembly of blood vessels with surface lining EC and underlying mural cells. Coculture of EC with SMC induces a mature, quiescent EC phenotype as evidenced by 1) a significant increase in the number of junctional complexes of the EC surface layer, 2) a down-regulation of PDGF-B expression by cocultured EC, and 3) an increased resistance of EC to undergo apoptosis. Furthermore, EC cocultured with SMC become refractory to stimulation with VEGF (lack of CD34 expression on VEGF stimulation; inability to form capillary-like sprouts in a VEGF-dependent manner in a 3-dimensional in gel angiogenesis assay). In contrast, costimulation with VEGF and Ang-2 induced sprouting angiogenesis originating from coculture spheroids consistent with a model of Ang-2-mediated vessel destabilization resulting in VEGF responsiveness. Ang-2 on its own was able to stimulate endothelial cells in the absence of Ang-1 producing SMC, inducing lateral sheet migration as well as in gel sprouting angiogenesis. Taken together, the data establish the spheroidal EC/SMC system as a powerful cell culture model to study paracrine interactions in the vessel wall and provide functional evidence for smooth muscle cell-mediated quiescence effects on endothelial cells.

Tensional forces in fibrillar extracellular matrices control directional capillary sprouting.

During angiogenesis, anastomosing capillary sprouts align to form complex three-dimensional networks of new blood vessels. Using an endothelial cell spheroid model that was developed to study endothelial cell differentiation processes, we have devised a novel collagen gel-based three-dimensional in vitro angiogenesis assay. In this assay, cell number-defined, gel-embedded endothelial cell spheroids act as a cellular delivery device, which serves as a focal starting point for the sprouting of lumenized capillary-like structures that can be induced to form complex anastomosing networks. Formation of capillary anastomoses is associated with tensional remodeling of the collagen matrix and directional sprouting of outgrowing capillaries towards each other. To analyze whether directional sprouting is dependent on cytokine gradients or on endothelial cell-derived tractional forces transduced through the extracellular matrix, we designed a matrix tension generator that enables the application of defined tensional forces on the extracellular matrix. Using this matrix tension generator, causal evidence is presented that tensional forces on a fibrillar extracellular matrix such as type I collagen, but not fibrin, are sufficient to guide directional outgrowth of endothelial cells. RGD peptides but not control RAD peptides disrupted the integrity of sprouting capillary-like structures and induced detachment of outgrowing endothelial cells cultured on top of collagen gels, but did not inhibit primary outgrowth of endothelial cells. The data establish the endothelial cell spheroid-based three-dimensional angiogenesis technique as a standardized, highly reproducible quantitative assay for in vitro angiogenesis studies and demonstrate that integrin-dependent matrix tensional forces control directional capillary sprouting and network formation.

Integration of endothelial cells in multicellular spheroids prevents apoptosis and induces differentiation.

Single endothelial cells (EC) seeded in suspension culture rapidly undergo apoptosis. Addition of survival factors, such as VEGF and FGF-2, does not prevent apoptosis of suspended EC. However, when cells are allowed to establish cell-cell contacts, they become responsive to the activities of survival factors. These observations have led to the development of a three-dimensional spheroid model of EC differentiation. EC spheroids remodel over time to establish a differentiated surface layer of EC and a center of unorganized EC that subsequently undergo apoptosis. Surface EC become quiescent, establish firm cell-cell contacts, and can be induced to express differentiation antigens (e.g., induction of CD34 expression by VEGF). In contrast, the unorganized center spheroid cells undergo apoptosis if they are not rescued by survival factors. The responsiveness to the survival factor activities of VEGF and FGF-2 was not dependent on cell shape changes since it was retained after cytochalasin D treatment. Taken together, these findings characterize survival factor requirements of unorganized EC and indicate that polarized surface EC differentiate to become independent of exogenous survival factors. Furthermore, they demonstrate that spheroid cell culture systems are useful not just for the study of tumor cells and embryonic stem cells but also for the analysis of differentiated functions of nontransformed cells.