Cell elongation is key to in silico replication of in vitro vasculogenesis and subsequent remodeling

Vasculogenesis, the de novo growth of the primary vascular network from initially dispersed endothelial cells, is the first step in the development of the circulatory system in vertebrates. In the first stages of vasculogenesis, endothelial cells elongate and form a network-like structure, called the primary capillary plexus, which subsequently remodels, with the size of the vacancies between ribbons of endothelial cells coarsening over time. To isolate such intrinsic morphogenetic ability of endothelial cells from its regulation by long-range guidance cues and additional cell types, we use an in vitro model of human umbilical vein endothelial cells (HUVEC) in Matrigel. This quasi-two-dimensional endothelial cell culture model would most closely correspond to vasculogenesis in flat areas of the embryo like the yolk sac. Several studies have used continuum mathematical models to explore in vitro vasculogenesis: such models describe cell ensembles but ignore the endothelial cells' shapes and active surface fluctuations. While these models initially reproduce vascular-like morphologies, they eventually stabilize into a disconnected pattern of vascular "islands." Also, they fail to reproduce temporally correct network coarsening. Using a cell-centered computational model, we show that the endothelial cells' elongated shape is key to correct spatiotemporal in silico replication of stable vascular network growth. We validate our simulation results against HUVEC cultures using time-resolved image analysis and find that our simulations quantitatively reproduce in vitro vasculogenesis and subsequent in vitro remodeling.     

Effects of endothelial growth media on proepicardial cell gene expression and morphogenesis in 3D collagen matrices

Proepicardial cells (PE) contribute to embryonic coronary vessel and epicardial development. Cells from the PE region can differentiate into coronary vascular smooth muscle cells and fibroblasts in vitro, but the endothelial specification capability of these cells is controversial. We sought to examine the effects of endothelial cell growth media on gene expression and the morphogenic properties of proepicardial cells in three-dimensional (3D) matrices. A primary culture of avian PE cells was subjected to molecular characterization with selected endothelial specific markers. Morphogenic properties of PE cells were assessed by in vitro assays of coronary vasculogenesis and invasion, which utilized highly defined, serum free, three-dimensional matrix conditions. PE cells maintained mixed cell population properties in the culture based on morphogenic features, immunohistochemistry, and the gene expression data. When suspended in a 3D vasculogenesis in vitro assay, PE cells formed intracellular vacuoles and assembled into multicellular tubes. Further, ultrastructural analysis revealed the presence of pinocytic vacuoles, intercellular junctions, and endothelial specific Weibel Palade bodies. In the invasion assay, PE cells spontaneously invaded control matrices. This invasion was markedly enhanced by lysophosphatidic acid (94 ± 9.6 vs. 285.6 ± 54.9, p < 0.05) and was completely blocked with synthetic broad-spectrum metalloproteinase inhibitor GM6001. Isolated PE cells grown in endothelial cell media represent mixed-cell population, characterized by both smooth muscle and endothelial gene expression. When placed in 3D in vitro assays, PE cells manifest morphogenic properties, including multicellular tube assembly and invasion.     

Biochemical and mechanical extracellular matrix properties dictate mammary epithelial cell motility and assembly

Biochemical and mechanical cues of the extracellular matrix have been shown to play important roles in cell-matrix and cell-cell interactions. We have experimentally tested the combined influence of these cues to better understand cell motility, force generation, cell-cell interaction, and assembly in an in vitro breast cancer model. MCF-10A non-tumorigenic mammary epithelial cells were observed on surfaces with varying fibronectin ligand concentration and polyacrylamide gel rigidity. Our data show that cell velocity is biphasic in both matrix rigidity and adhesiveness. The maximum cell migration velocity occurs only at specific combination of substrate stiffness and ligand density. We found cell-cell interactions reduce migration velocity. However, the traction forces cells exert onto the substrate increase linearly with both cues, with cells in pairs exerting higher maximum tractions observed over single cells. A relationship between force and motility shows a maximum in single cell velocity not observed in cell pairs. Cell-cell adhesion becomes strongly favored on softer gels with elasticity ≤ 1250 Pascals (Pa), implying the existence of a compliance threshold that promotes cell-cell over cell-matrix adhesion. Finally on gels with stiffness similar to pre-malignant breast tissue, 400 Pa, cells undergo multicellular assembly and division into 3D spherical aggregates on a 2D surface.     

Dissecting coronary angiogenesis: 3D co-culture of cardiomyocytes with endothelial or mesenchymal cells

In embryogenesis, coronary blood vessels are formed by vasculogenesis from epicardium-derived progenitors. Subsequently, growing or regenerating myocardium increases its vasculature by angiogenesis, forming new vessels from the pre-existing ones. Recently, cell therapies for myocardium ischemia that used different protocols have given promising results, using either extra-cardiac blood vessel cell progenitors or stimulating the cardiac angiogenesis. We have questioned whether cardiomyocytes could sustain both vasculogenesis and angiogenesis. We used a 3D culture model of tissue-like spheroids in co-cultures of cardiomyocytes supplemented either with endothelial cells or with bone marrow-derived mesenchymal stroma cells. Murine foetal cardiomyocytes introduced into non-adherent U-wells formed 3D contractile structures. They were coupled by gap junctions. Cardiomyocytes segregated inside the 3D structure into clumps separated by connective tissue septa, rich in fibronectin. Three vascular endothelial growth factor isoforms were produced (VEGF 120, 164 and 188). When co-cultured with human umbilical cord endothelial cells, vascular structures were produced in fibronectin-rich external layer and in radial septa, followed by angiogenic sprouting into the cardiomyocyte microtissue. Presence of vascular structures led to the maintenance of long-term survival and contractile capacity of cardiac microtissues. Conversely, bone marrow mesenchymal cells formed isolated cell aggregates, which progressively expressed the endothelial markers von Willebrand's antigen and CD31. They proceeded to typical vasculogenesis forming new blood vessels organised in radial pattern. Our results indicate that the in vitro 3D model of cardiomyocyte spheroids provides the two basic elements for formation of new blood vessels: fibronectin and VEGF. Within the myocardial environment, endothelial and mesenchymal cells can proceed to formation of new blood vessels either through angiogenesis or vasculogenesis, respectively.     

Normal and Tumoral Melanocytes Exhibit q-Gaussian Random Search Patterns

In multicellular organisms, cell motility is central in all morphogenetic processes, tissue maintenance, wound healing and immune surveillance. Hence, failures in its regulation potentiates numerous diseases. Here, cell migration assays on plastic 2D surfaces were performed using normal (Melan A) and tumoral (B16F10) murine melanocytes in random motility conditions. The trajectories of the centroids of the cell perimeters were tracked through time-lapse microscopy. The statistics of these trajectories was analyzed by building velocity and turn angle distributions, as well as velocity autocorrelations and the scaling of mean-squared displacements. We find that these cells exhibit a crossover from a normal to a super-diffusive motion without angular persistence at long time scales. Moreover, these melanocytes move with non-Gaussian velocity distributions. This major finding indicates that amongst those animal cells supposedly migrating through Lévy walks, some of them can instead perform q-Gaussian walks. Furthermore, our results reveal that B16F10 cells infected by mycoplasmas exhibit essentially the same diffusivity than their healthy counterparts. Finally, a q-Gaussian random walk model was proposed to account for these melanocytic migratory traits. Simulations based on this model correctly describe the crossover to super-diffusivity in the cell migration tracks.     

Thymic accessory cell complexes in vitro and in vivo: morphological study

Summary Murine thymic macrophages and interdigitating cells, also called thymic accessory cells, were characterized by means of light- and electron microscopy. The cells were studied in suspension, during isolation by enzymatic digestion and in vivo. They were observed as isolated cells or as components of multicellular complexes, some of which were rosettes and were composed of lymphoid cells centered on each type of accessory cell. We also noted other cell complexes including macrophages that resembled classical epithelial nurse cells. We consider that multicellular complexes represent lymphostromal associations already existing in vivo, because we observed them at the periphery of thymic pieces undergoing enzymatic treatment. The heterogeneity of macrophages that we observed in vitro was also noted in vivo. In vivo macrophages were of three types: classical phagocytic cells distributed throughout the gland, cortical elongated cells in close contact with lymphoid blast cells, and atypical nurse cells containing mitotic cells and located in the inner cortex. The morphological aspects of the latter two cell types suggest that cortical macrophages in vivo have other roles: they can be interpreted as images of positive or negative cell selection. We also believe that rosettes are formed by elongated cortical macrophages when they are enzymatically isolated from the thymus.Part of this work was presented at the Second Thymus Workshop, Rolduc, The Netherlands, April 1989     

Mechanical Cell-Matrix Feedback Explains Pairwise and Collective Endothelial Cell Behavior In Vitro

In vitro cultures of endothelial cells are a widely used model system of the collective behavior of endothelial cells during vasculogenesis and angiogenesis. When seeded in an extracellular matrix, endothelial cells can form blood vessel-like structures, including vascular networks and sprouts. Endothelial morphogenesis depends on a large number of chemical and mechanical factors, including the compliancy of the extracellular matrix, the available growth factors, the adhesion of cells to the extracellular matrix, cell-cell signaling, etc. Although various computational models have been proposed to explain the role of each of these biochemical and biomechanical effects, the understanding of the mechanisms underlying in vitro angiogenesis is still incomplete. Most explanations focus on predicting the whole vascular network or sprout from the underlying cell behavior, and do not check if the same model also correctly captures the intermediate scale: the pairwise cell-cell interactions or single cell responses to ECM mechanics. Here we show, using a hybrid cellular Potts and finite element computational model, that a single set of biologically plausible rules describing (a) the contractile forces that endothelial cells exert on the ECM, (b) the resulting strains in the extracellular matrix, and (c) the cellular response to the strains, suffices for reproducing the behavior of individual endothelial cells and the interactions of endothelial cell pairs in compliant matrices. With the same set of rules, the model also reproduces network formation from scattered cells, and sprouting from endothelial spheroids. Combining the present mechanical model with aspects of previously proposed mechanical and chemical models may lead to a more complete understanding of in vitro angiogenesis.     

Genetic circuitry controlling motility behaviors of Myxococcus xanthus

M. xanthus has a complex multicellular lifestyle including swarming, predation and development. These behaviors depend on the ability of the cells to achieve directed motility across solid surfaces. M. xanthus cells have evolved two motility systems including Type-IV pili that act as grappling hooks and a controversial engine involving mucus secretion and fixed focal adhesion sites. The necessity for cells to coordinate the motility systems and to respond rapidly to environmental cues is reflected by a complex genetic network involving at least three complete sets of chemosensory systems and eukaryotic-like signaling proteins. In this review, we discuss recent advances suggesting that motor synchronization results from spatial oscillations of motility proteins. We further propose that these dynamics are modulated by the action of multiple upstream complementary signaling systems. M. xanthus is thus an exciting emerging model system to study the intricate processes of directed cell migration.     

Colony Expansion of Socially Motile Myxococcus xanthus Cells Is Driven by Growth,Motility, and Exopolysaccharide Production

Myxococcus xanthus, a model organism for studies of multicellular behavior in bacteria, moves exclusively on solid surfaces using two distinct but coordinated motility mechanisms. One of these, social (S) motility is powered by the extension and retraction of type IV pili and requires the presence of exopolysaccharides (EPS) produced by neighboring cells. As a result, S motility requires close cell-to-cell proximity and isolated cells do not translocate. Previous studies measuring S motility by observing the colony expansion of cells deposited on agar have shown that the expansion rate increases with initial cell density, but the biophysical mechanisms involved remain largely unknown. To understand the dynamics of S motility-driven colony expansion, we developed a reaction-diffusion model describing the effects of cell density, EPS deposition and nutrient exposure on the expansion rate. Our results show that at steady state the population expands as a traveling wave with a speed determined by the interplay of cell motility and growth, a well-known characteristic of Fisher’s equation. The model explains the density-dependence of the colony expansion by demonstrating the presence of a lag phase–a transient period of very slow expansion with a duration dependent on the initial cell density. We propose that at a low initial density, more time is required for the cells to accumulate enough EPS to activate S-motility resulting in a longer lag period. Furthermore, our model makes the novel prediction that following the lag phase the population expands at a constant rate independent of the cell density. These predictions were confirmed by S motility experiments capturing long-term expansion dynamics.     

3D printing to construct in vitro multicellular models of melanoma

Currently, there is a lack of suitable models for in-vitro studies of malignant melanoma and traditional single cell culture models no longer reproduce tumor structure and physiological complexity well. The tumor microenvironment is closely related to carcinogenesis and it is particularly important to understand how tumor cells interact and communicate with surrounding nonmalignant cells. Three-dimensional (3D) in vitro multicellular culture models can better simulate the tumor microenvironment due to their excellent physicochemical properties. In this study, 3D composite hydrogel scaffolds were prepared from gelatin methacrylate and polyethylene glycol diacrylate hydrogels by 3D printing and light curing techniques, and 3D multicellular in vitro tumor culture models were established by inoculating human melanoma cells (A375) and human fibroblasts cells on them. The cell proliferation, migration, invasion, and drug resistance of the 3D multicellular in vitro model was evaluated. Compared with the single-cell model, the cells in the multicellular model had higher proliferation activity and migration ability, and were easy to form dense structures. Several tumor cell markers, such as matrix metalloproteinase-9 (MMP-9), MMP-2, and vascular endothelial growth factor, were highly expressed in the multicellular culture model, which were more favorable for tumor development. In addition, higher cell survival rate was observed after exposure to luteolin. The anticancer drug resistance result of the malignant melanoma cells in the 3D bioprinted construct demonstrated physiological properties, suggesting the promising potential of current 3D printed tumor model in the development of personalized therapy, especially for discovery of more conducive targeted drugs.     

A novel membrane protein influencing cell shape and multicellular swarming of Proteus mirabilis

Swarming in Proteus mirabilis is characterized by the coordinated surface migration of multicellular rafts of highly elongated, hyperflagellated swarm cells. We describe a transposon mutant, MNS185, that was unable to swarm even though vegetative cells retained normal motility and the ability to differentiate into swarm cells. However, these elongated cells were irregularly curved and had variable diameters, suggesting that the migration defect results from the inability of these deformed swarm cells to align into multicellular rafts. The transposon was inserted at codon 196 of a 228-codon gene that lacks recognizable homologs. Multiple copies of the wild-type gene, called ccmA, for curved cell morphology, restored swarming to the mutant. The 25-kDa CcmA protein is predicted to span the inner membrane twice, with its C-terminal major domain being present in the cytoplasm. Membrane localization was confirmed both by immunoblotting and by electron microscopy of immunogold-labelled sections. Two forms of CcmA were identified for wild-type P. mirabilis; they were full-length integral membrane CcmA1 and N-terminally truncated peripheral membrane CcmA2, both present at approximately 20-fold higher concentrations in swarm cells. Differentiated MNS185 mutant cells contained wild-type levels of the C-terminally truncated versions of both proteins. Elongated cells of a ccmA null mutant were less misshapen than those of MNS185 and were able to swarm, albeit more slowly than wild-type cells. The truncated CcmA proteins may therefore interfere with normal morphogenesis, while the wild-type proteins, which are not essential for swarming, may enhance migration by maintaining the linearity of highly elongated cells. Consistent with this view, overexpression of the ccmA gene caused cells of both Escherichia coli and P. mirabilis to become enlarged and ellipsoidal.     

Symmetry-breaking in mammalian cell cohort migration during tissue pattern formation: role of random-walk persistence

Coordinated, cohort cell migration plays an important role in the morphogenesis of tissue patterns in metazoa. However, individual cells intrinsically move in a random walk-like fashion when studied in vitro. Hence, in the absence of an external orchestrating influence or template, the emergence of cohort cell migration must involve a symmetry-breaking event. To study this process, we used a novel experimental system in which multiple capillary endothelial cells exhibit spontaneous and robust cohort migration in the absence of chemical gradients when cultured on micrometer-scale extracellular matrix islands fabricated using microcontact printing. A computational model suggested that directional persistence of random-walk and dynamic mechanical coupling of adjacent cells are the critical control parameters for this symmetry-breaking behavior that is induced in spatially-constrained cell ensembles. The model predicted our finding that fibroblasts, which exhibit a much shorter motility persistence time than endothelial cells, failed to undergo symmetry breaking or produce cohort migration on the matrix islands. These findings suggest that cells have intrinsic motility characteristics that are tuned to match their role in tissue patterning. Our results underscore the importance of studying cell motility in the context of cell populations, and the need to address emergent features in multicellular organisms that arise not only from cell-cell and cell-matrix interactions, but also from properties that are intrinsic to individual cells.     

Hela l-CaD is Implicated in the Migration of Endothelial Cells/Endothelial Progenitor Cells in Human Neoplasms

Caldesmon (CaD) is a major actin-binding protein distributed in a variety of cell types. No functional differences among the isoforms in in vitro studies were found so far. In a previous study we found that the low molecular caldesmon isoform (Hela l-CaD) is expressed in endothelial cells (ECs)/endothelial progenitor cells (EPCs) in tumor vasculature of various human tumors. Activation of cell motility is necessary for the navigation of the tip ECs during angiogenesis, and migration of EPCs from the bone marrow during vasculogenesis. In the present study we searched for features of motility and the intracellular expression sites of Hela l-CaD in ECs/EPCs of various human tumors under histologically preserved microenviroment. We discovered a variety of motility-related cell protrusions like filopodia, microspikes, lamellipodia, podosomes, membrane blebs and membrane ruffles in the activated ECs/EPCs. Hela l-CaD appeared to be invariably expressed in the subregions of these cell protrusions. The findings suggest that Hela l-CaD is implicated in the migration of ECs/EPC in human neoplasms where they contribute to tumor vasculogenesis and angiogenesis.Key words: Hela l-CaD, cell motility, angiogenesis, vasculogenesis, ECs/EPCs     

The effects of continuous and discontinuous groove edges on cell shape and alignment

Nanofabricated model surfaces and digital image analysis of cell shape were used to address the importance of a continuous sharp edge in the alignment of cells to shallow surface grooves. The grooved model surfaces had either continuous or discontinuous edges of various depths (40-400 nm) but identical surface chemistry and groove/ridge dimensions (15 microm wide). Epithelial cells were cultured on the model surfaces for 10 and 24 h. Fluorescence microscopy combined with image analysis were used to quantify cell area and alignment and to make cell shape classifications of individual cells. The degrees of alignment of cells and the percentages of elongated cell classes increased with groove depth on samples with continuous grooves. Two main differences, with regard to cell response, were observed between the continuous and discontinuous grooved surfaces. First, significantly fewer cells aligned to surface grooves with discontinuous edges than to grooves with continuous edges. Second, there were lower percentages of the elongated cell classes on discontinuous grooves than on continuous ones. We concluded that grooved surfaces with continuous edges are more potent in aligning and inducing elongated cells. The results from the present study suggest that a mechanism of alignment involving orientation along a continuous edge is likely.     

Caryocyst-like host cell formation by Caryospora duszynskii (Apicomplexa: Eimeriidae) in human fetal lung cell cultures

Sporozoites of the coccidium, Caryospora duszynskii, penetrated human fetal lung cell cultures but did not undergo asexual or sexual multiplication during a 29-day observation period. Beginning three days postinoculation (PI), infected host cells lost their normal elongated fibroblast-like shape and became ellipsoidal in appearance and resembled caryocysts. These caryocyst-like infected cells were observed from 3 through 29 days PI. Sporozoites remained viable throughout the study as evidenced by motility of extracellular sporozoites in infected human fetal lung cell cultures. Results of this in vitro study suggest that some species of Caryospora may form caryocysts in secondary hosts without undergoing asexual or sexual multiplication in these hosts.     

N-cadherin deficiency impairs pericyte recruitment, and not endothelial differentiation or sprouting, in embryonic stem cell-derived angiogenesis

Endothelial cells express two classical cadherins, VE-cadherin and N-cadherin. VE-cadherin is absolutely required for vascular morphogenesis, but N-cadherin is thought to participate in vessel stabilization by interacting with periendothelial cells during vessel formation. However, recent data suggest a more critical role for N-cadherin in endothelium that would regulate angiogenesis, in part by controlling VE-cadherin expression. In this study, we have assessed N-cadherin function in vascular development using an in vitro model derived from embryonic stem (ES) cell differentiation. We show that pluripotent ES cells genetically null for N-cadherin can differentiate normally into endothelial cells. In addition, sprouting angiogenesis was unaltered, suggesting that N-cadherin is not essential for the early events of angiogenesis. However, the lack of N-cadherin led to an impairment in pericyte covering of endothelial outgrowths. We conclude that N-cadherin is necessary neither for vasculogenesis nor proliferation and migration of endothelial cells but is required for the subsequent maturation of endothelial sprouts by interacting with pericytes.     

Cell flexibility affects the alignment of model myxobacteria

Myxobacteria are social bacteria that exhibit a complex life cycle culminating in the development of multicellular fruiting bodies. The alignment of rod-shaped myxobacteria cells within populations is crucial for development to proceed. It has been suggested that myxobacteria align due to mechanical interactions between gliding cells and that cell flexibility facilitates reorientation of cells upon mechanical contact. However, these suggestions have not been based on experimental or theoretical evidence. Here we created a computational mass-spring model of a flexible rod-shaped cell that glides on a substratum periodically reversing direction. The model was formulated in terms of experimentally measurable mechanical parameters, such as engine force, bending stiffness, and drag coefficient. We investigated how cell flexibility and motility engine type affected the pattern of cell gliding and the alignment of a population of 500 mechanically interacting cells. It was found that a flexible cell powered by engine force at the rear of the cell, as suggested by the slime extrusion hypothesis for myxobacteria motility engine, would not be able to glide in the direction of its long axis. A population of rigid reversing cells could indeed align due to mechanical interactions between cells, but cell flexibility impaired the alignment.     

Hela l-CaD Is Implicated in the Migration of Endothelial Cells/Endothelial Progenitor Cells in Human Neoplasms

Caldesmon (CaD) is a major actin-binding protein distributed in a variety of cell types. No functional differences among the isoforms in in vitro studies were found so far. In a previous study we found that the low molecular caldesmon isoform (Hela l-CaD) is expressed in endothelial cells (ECs)/endothelial progenitor cells (EPCs) in tumor vasculature of various human tumors. Activation of cell motility is necessary for the navigation of the tip ECs during angiogenesis, and migration of EPCs from the bone marrow during vasculogenesis. In the present study we searched for features of motility and the intracellular expression sites of Hela l-CaD in ECs/EPCs of various human tumors under histologically preserved microenviroment. We discovered a variety of motility-related cell protrusions like filopodia, microspikes, lamellipodia, podosomes, membrane blebs and membrane ruffles in the activated ECs/EPCs. Hela l-CaD appeared to be invariably expressed in the subregions of these cell protrusions. The findings suggest that Hela l-CaD is implicated in the migration of ECs/EPC in human neoplasms where they contribute to tumor vasculogenesis and angiogenesis.     

Viscoelasticity of multicellular surfaces

Various modeling approaches have been applied to describe viscoelasticity of multicellular surfaces. The viscoelasticity is considered within three time regimes: (1) short time regime for milliseconds to seconds time scale which corresponds to sub-cellular level; (2) middle time regime for several tens of seconds to several minutes time scale which corresponds to cellular level; and (3) long time regime for several tens of minutes to several hours time scale which corresponds to supra-cellular level. Short and middle time regimes have been successfully elaborated in the literature, whereas long time viscoelasticity remains unclear. Long time regime accounts for collective cell migration. Collective cell migration could induce uncorrelated motility which has an impact to energy storage and dissipation during cell surface rearrangement. Uncorrelated motility influences: (1) volume fraction of migrating cells, (2) distribution of migrating cells, (3) shapes of migrating cell groups. These parameters influence mechanical coupling between migrating and resting subpopulations and consequently the constitutive model for long time regime.This modeling consideration indicates that additional experimental work is needed to confirm the feasibility of constitutive models which have been applied in literature for long time regime as: (1) relaxation of stress and strain, (2) storage and loss moduli as the function of time, (3) distribution of migrating cells.     

Isolation and differentation of cloned epithelial cell lines from normal rat mammary glands

Summary Single-cell-cloned cell lines have been established from primary cultures of neonatal rat mammary glands. A representative cuboidal cell line, Rama 704, shows the presence of intermediate filamental proteins keratin and vimentin, and occasional cells express milk fat globule membrane antigens on their apical surfaces. Rama 704 cells grow as a cuboidal pavement in culture and produce hemispherical blisters or domes when confluent. Noteworthy ultrastructural features are the presence of junctional complexes, desmosomes, and apical microvilli typical of epithelia. Cells seeded within floating collagen gels with form a variety of multicellular outgrowths, some of which are ductlike in morphology and are composed of polarized cells surrounding a central lumen. The cuboidal cells produce elongated cells under conditions of high cell density and also when cells float off collagen gels and reattach to the plastic substrate. The former elongated cells have been cloned and three cel lines established: Rama 710, 711, and 712; the latter uncloned elongated cells are termed Rama 704E. The cloned elongated cells show an increase in the amounts of basement membrane proteins deposited, a lack of junctional complexes and microvilli, and an increase in the amount of rough endoplasmic reticulum compared with their parental cells. Rama 704E cells show an enhanced deposition of basement membrane proteins and increased amounts of actin in the cytoplasm over the elongated cell lines and contain microfilaments and pincocytotic vesicles similar to those seen in myoepithelial cells. All the elongated cells and lines fail to form ductlike structures within collagen gels. None of the cell lines form tumors in syngeneic rats although they all produce some tumors in nude mice, which are composed of cords of epithelioid cells and spindle cells in varying proportions. In addition, some of the Rama 704 tumors contain rhabdomyoblastic elements that penetrate the host fat pad. This is the first report of the isolation and characterization of a stable cuboidal cell line from a neonatal rat mammary gland. The Rama 704 cell line shows morphological and biochemical features of mammary epithelial cells and converts at high cell density to elongated cells that have also been cloned.