Origin of the dorsal surface of the neural tube by progressive delamination of epidermal ectoderm and neuroepithelium: implications for neurulation and neural tube defects

Knowledge of the morphogenetic events involved in the development of the dorsal portion of the neural tube is important for understanding neural tube closure, neural crest cell formation and emigration, and the origin of neural tube defects. Here, I characterize the progressive development of the tips of the neural folds during fold elevation in the trunk of mouse and chick embryos and the events leading to formation of the dorsal portion of the neural tube as the epidermal ectoderm (EE) and neuroepithelium (NE) separate from each other. The nature and timing of appearance of collagen IV, laminin and fibronectin were analysed by immunofluorescent and immunogold labelling, and ruthenium red and tannic acid were used to enhance staining for proteoglycans and glycosaminoglycans. As the neural folds elevate, the NE and EE delaminate progressively beginning at the basal surface of the lateral extremes of the neural plate. Nevertheless, the two epithelia remain connected across the zone of delamination by their previously existing basal laminae. In each fold, proteoglycan granules appear at the interface between the NE and EE before delamination begins, and then an (interepithelial) space begins to open and propagate dorsally. Other extracellular matrix (ECM) molecules appear within the space a short distance behind its tip and basal lamina deposition begins shortly thereafter. As fusion occurs, the interepithelial spaces of the two folds coalesce and the final separation of the EE from the NE is accomplished. These observations suggest that the previously recognized delay in deposition of ECM and basal lamina on the dorsal portion of the neural tube and on the overlying EE is a direct consequence of the delamination of the two epithelia and the establishment of two new basal surfaces. The observation that the surface of the dorsal third of the neural tube forms by delamination rather than by juxtaposition of previously existing basal surfaces of the two epithelial is discussed in terms of possible implications for models of neurulation and the origin of neural tube defects.     

Studies on the mechanisms of neurulation in the chick: interrelationship of contractile proteins, microfilaments, and the shape of neuroepithelial cells

Electron microscopy and indirect immunofluorescence were employed to correlate the distribution patterns of major contractile proteins (actin and myosin) with 1) the organizational state of microfilaments, 2) the apical cell surface topography, 3) the shape of the neuroepithelial cells, and 4) the degree of bending of the neuroepithelium during neurulation in chick embryos at Hamburger and Hamilton stages 5-10 of development. Both actin and myosin are present at these developmental stages and colocalize in the neural plate as well as in later phases of neurulation. During elevation of neural folds, actin- and myosin-specific fluorescence is always most intense in regions where the greatest degree of bending of the neuroepithelium takes place [e.g., the midline of the V-shaped neuroepithelium (early neural fold stage) and the midlateral walls of the "C"-shaped neuroepithelium (mid-neural-fold stage)]. This intense fluorescence coincides with 1) a particularly dense packing of microfilaments and 2) highly constricted cell apices. After neural folds make contact, there is an overall reduction in both the intensity of apical fluorescence and the thickness of apical microfilament bundles, especially in the roof and floor of the neural tube. The remaining fluorescence in the contact area is apparently related to cellular movements during fusion of neural folds.     

The effects of 5-bromodeoxyuridine on fusion of the cranial neural folds in the mouse embryo

The effects of 500 and 300 mg/kg bromodeoxyuridine (BUdR) on the process of fusion of the neural folds were tested after injection into pregnant mice on day 8 of gestation (192 hours postcoitum). Various doses of the natural nucleoside, thymidine (TdR), were also tested. Both doses of BUdR retarded growth to the same extent, but only the larger dose caused neural tube defects in 28.8% of embryos. Treatment with the larger dose also caused extensive cell necrosis to appear in the neuroepithelium of the neural folds between 12 and 15 hours after treatment. No changes were detectable with the light microscope up to this time. Measurement of the cell generation time in treated and control embryos indicated that the BUdR prolonged the cycle by about 2 hours and that the dying cells were in the second DNA synthetic phase following incorporation of the analog. Treatment with the smaller dose of BUdR caused minimal cell necrosis. This was taken as evidence for the importance of cell necrosis in the pathogenesis of BUdR-induced neural tube defects. Treatment with excess TdR did not cause either neural tube defects or cell necrosis, and a dose of TdR equimolar with the large dose of BUdR (400 mg/kg TdR) did not retard growth. Doses of 800 and 1,200 mg/kg TdR retarded growth to the same extent as BUdR. The administration of an equimolar amount of TdR, along with the teratogenic dose of BUdR, prevented the occurrence of cell necrosis and neural tube defects. When treatments were given on day 9 of gestation, 500 mg/kg BUdR caused cell necrosis in the neuroepithelium about 15 hours after treatment but no neural tube defects were produced by day 9 after treatment. It is suggested that in this case cell necrosis occurred too late to interfere with neural fold fusion. It was concluded that the ability of BUdR to cause exencephaly in mouse embryos was due to cell necrosis in the neuroepithelium.     

An ultrastructural examination of the role of cell membrane surface coat material during neurulation

Data from neural crest cultures indicate that cell surface coat material (CSM) is directly involved in cellular migration and events surrounding differentiation. To investigate whether the CSM also has a morphogenetic role, embryos of the amphibian Ambystoma maculatum were examined ultrastructurally throughout the stages of neurulation. Segments of the neural axis were fixed in glutaraldehyde-containing Alcian blue 8GX, which reportedly enhances preservation of CSM, and were postfixed in OsO4 containing 1 percent lanthanum nitrate, which stains the CSM. The medial groove formed by the appearance of the neural ridges contains a large amount of CSM and numerous vesicles coated with lanthanum-positive material. In contrast, the lateral ridge surfaces are covered by a small amount of uniformly distributed CSM and a paucity of vesicles. As the ridges begin to fold there is a progressive increase in the amount of CSM within the presumptive neural tube region. Further convergence of the neural folds is accompanied by an increase of CSM at their leading edges. As the folds approximate each other, lanthanum-positive material physically bridges the gap. However, as the apposing tissue actually abuts to form the neural tube, no CSM is observed in the remaining interspace. The specific distribution and sequential accumulation of cell CSM during the events of neurulation strongly suggest its direct participation in the morphogenetic process.     

Developmental anomalies induced by all-trans-retinoic acid in fetal mice: II. Induction of abnormal neuroepithelium

All-trans-retinoic acid (RA) in olive oil was given in doses of 0, 40, or 60 mg/kg of body weight to pregnant mice on day 8 of gestation, and 2-6 hr later embryos were fixed in solutions with or without cetylpyridinium chloride (CPC). The neuroepithelium of the presumptive midbrain was processed for light and electron microscopy. Distorted contours of the neuroepithelium were induced by both doses of RA and the incidence and the severity of the disorganized neuroepithelium showed dose-related results. Abnormal neuroepithelium showed wide intercellular spaces with degenerated cytoplasmic processes or cell debris, separation of the apical side from adjacent cells, retention of mitotic and/or postmitotic cells on the apical side, presence of mitotic cells on the basal side, and detachment of degenerated structures from the neuroepithelium. Ultrastructurally, the affected neuroepithelium showed (1) appearance of degenerating filamentous or tubular coagulating bundles in the cytoplasm and the cytoplasmic process of the neural crest cells, (2) dispersal of polysomes into monosomes especially in the degenerating neural crest cells, (3) and a collecting of microfilament-like structures at the contact area between the neural crest cell and the presumptive neuroblast. These morphological changes suggest that RA affects the nature of cytoskeletal elements and the protein synthesis of the neuroepithelial cells. The selective susceptibility of neural crest cells to RA causes more degenerating neural crest cells in the neuroepithelium, which causes nonapproximation of the neural folds and scantiness of the migrating neural crest cells; these results lead to neural tube defects and craniofacial anomalies, respectively.     

Shaping and bending of the avian neuroepithelium: morphometric analyses

Changes in the size and shape of the neuroepithelium were measured from serial transverse sections of 30 plastic-embedded chick embryos at stages 4-11. The neural plate folds into a neural tube during this period. Changes in volume, length, apical and basal widths, apical and basal surface areas, and thickness of the neuroepithelium were measured and correlated with the amount of folding that had occurred. These measurements were made to provide data for comparison with those available from other systems, to gain insight into the mechanisms of shaping and bending of the neuroepithelium, and to obtain normal parameters for eventual comparison with those obtained from embryos with induced neural tube defects. During stages 4-11, the volume, length, apical and basal surface areas, and lateral thickness of the neuroepithelium increase, whereas apical and basal widths and median thickness of the neuroepithelium decrease. Models are presented to demonstrate the effects of possible changes in neuroepithelial cell number, position, and size on the shaping of the neural plate.     

SEM observations of the neural fold associated with neurulation in the rat

The topography of the ectoderm was examined by scanning electron microscopy during neurulation in rat embryos at stage 24 (somites 8-11). A zone of altered cell morphology was observed along the crest of neural folds. This zone was located between the presumptive neural tube and the surface ectoderm and exhibited numerous rounded cell blebs, immediately prior to fusion between the folds. It is suggested that the observed surface alterations may reflect a change in the properties of the altered cell which correlate with initial adhesion between the folds.     

Ultrastructural studies of amphibian neural fold fusion.

The fusion of neural folds to form the neural tube is a process in which presumptive contacting surfaces become adhering. An ultrastructural examination of regions of neural folds in the neurulae of three amphibian species (Hyla regilla, Rana pipiens, and Xenopus laevis), using both transmission and scanning electron microscopy, revealed that, prior to fusion, there is formation of vesicles within cells lining the neural groove, development of extracellular vesicles, changes in the surface morphology of the cells forming the fusion area, and extension of projections (filopodia) from cells lining the neural groove. The association of intra- and extracellular vesicles and filopodia with cells of the neural groove and folds suggests that these organelles may be involved in preparing the neural folds for initial contact, adhesion, and fusion. Ultrastructural differences in reaction of neural fold cell surfaces to staining by ruthenium red, colloidal iron, Alcian blue-lanthanum nitrate, and concanavalin A-hemocyanin indicate that the glycosaminoglycan compositions of these cell surfaces differ from those of presumptive epidermal cells.     

Observations on closure of the neuropores in the chick embryo.

Neuropore closure was studied in chick embryos by light and electron microscopy. Surface ectoderm reflects over the crests of the neural folds at all craniocaudal levels, merging with the neural ectoderm lining the neural groove. Apices of surface ectodermal cells have an essentially identical morphology prior to approximation of folds, both within the presumptive fusion sites and more laterally. Cells of these areas have slightly convex profiles exhibiting few cellular protrusions. Each neural fold contains a superficial half, composed of neural ectoderm covered by surface ectoderm, and a deep half consisting entirely of neural ectoderm. Initial contact between folds usually occurs near the junction between these halves in cranial regions, but is restricted primarily to surface ectoderm at caudal levels. Subsequent fusion of folds at all levels involves both ectodermal layers. Cellular protrusions and small, morphologically unspecialized intercellular junctions often interconnect cells of apposed folds in areas undergoing fusion. The anterior neuropore closes at stages 10-11, but fusion of folds in this region is not completed until stages 13-14. Fusion occurs dorsoventrally in this area and is more advanced internally than externally. Numerous pleomorphic inclusions and a few apparently necrotic cells are present in areas bordering the anterior neuropore. The posterior neuropore closes at stages 12-13 and fusion is completed in this region during stages 13-14. The caudal end of the posterior neuropore closes dorsal to the developing tail bud. Several morphological features of this closure may at least partially account for the high susceptibility to myeloschisis localized specifically at caudal spinal cord levels.     

Fusion of Neural Folds in the Rhombencephalic Region of Rat Embryos

The progression of fusion of neural folds in the rhombencephalic region of rat embryos was examined using actin-specific antibody and scanning electron microscopy. In the region of unfused neural folds, actin was distributed over the entire luminal surface of the neural plate. Subsequently, the actin became localized on the luminal surface of outwardly opened, dorsolateral portions of the neural plate. Reduction in the area over which actin was distributed corresponded to the decrease in the area of the luminal surface of outwardly opened, dorsolateral portions of the neural plate. Decrease in the area of the luminal surface synchronized with elevation of the neural folds and their approach to the midline and, finally, actin became localized in areas of the apices of the neural plate that formed the bridge between the two opposing neural folds. These results suggest that formation of the bridge is orchestrated and supported by microfilaments. The procession of fusion in the rostral direction may cause approach of the neural folds to the middle, and this approach may be guaranteed by reduction in the area of the luminal surface over which actin is distributed.     

Arsenic-induced exencephaly in the mouse and associated lesions occurring during neurulation

Early tissue damage following a teratogenic dose of arsenic to the dam was studied in mice with the objective of detecting the primary lesion associated with the development of exencephaly. Animals were killed 6 to 21 h after a single 45 mg/kg intraperitoneal injection of sodium arsenate on day 8 of pregnancy and neurulation-stage embryos were fixed for histological and ultrastructural examination. In the prospective hindbrain, the most consistent feature associated with arsenate treatment was the widely separated neural folds which were not positioned for closure. Intracytoplasmic inclusions, interpreted as necrotic debris, were most numerous in the apical portion of the neural folds, sometimes extending into the mesenchyme, but they were not extensive in most embryos. In the prospective forebrain, necrotic debris was found throughout the neuroepithelium, in contrast to the posterior portions of the developing brain. It is not clear that necrosis of the neuroepithelium or mesenchyme would in itself be the primary lesion associated with exencephaly, although death of specific cells such as those participating in the fusion process could be involved. The potential effect of arsenate on physiological and biochemical processes which could affect neural tube closure is discussed.     

Cell migration into neural tube lumen provides evidence for the "fixed cortex" theory of cell motility

We present a model of cell motility based on emigration of neural crest cells into the neural tube lumen under in vitro conditions (10% fetal calf serum or YIGSR) that inhibit their normal emigration from the base of the neuroepithelium into surrounding extracellular matrix (ECM). Ultrastructural observations reveal that cells lining the lumen are joined by zonulae adherentes (ZA), which are points of strong intercellular attachment, and thereby serve as markers for fixed regions of plasmalemma and cortical actin. Three major observations of the relationship of cells to the ZA support the "fixed cortex" model of mesenchymal cell migration. First, cells extend apical cell processes past the ZA into the lumen. To do this, they must make new apical plasmalemma and actin cortex that the endoplasm slides into. Second, elongated cells are observed in the lumen that are still attached via ZA to the neuroepithelium. This indicates that all of the endoplasm finally slides past the ZA. Third, numerous cytoplasmic pieces, often attached to each other and to the neuroepithelium via ZA, are found at the site where cells appear to have detached from the epithelium after entering the lumen. Since the ZA is fixed in location, the endoplasm must have slid past it into newly manufactured anterior cortex and plasmalemma, with the trailing end of the cell finally snapping off. The "fixed cortex" theory of cell migration agrees with existing data in that it predicts the polarized insertion of new plasmalemma and actin at the leading end of the cell, but it differs significantly from existing theories of mesenchymal cell migration in that it states that the cell surface remains firmly attached to the substratum while the myosin-rich endoplasm slides past it.     

Xenopus ADAM 13 is a metalloprotease required for cranial neural crest-cell migration.

BACKGROUND: Cranial neural-crest (CNC) cells originate from the lateral edge of the anterior neuroepithelium and migrate to form parts of the peripheral nervous system, muscles, cartilage, and bones of the face. Neural crest-cell migration involves the loss of adhesion from the surrounding neuroepithelium and a corresponding increase in cell adhesion to the extracellular matrix (ECM) present in migratory pathways. While proteolytic activity is likely to contribute to the regulation of neural crest-cell adhesion and migration, the role of a neural crest-specific protease in these processes has yet to be demonstrated. We previously showed that CNC cells express ADAM 13, a cell surface metalloprotease/disintegrin. Proteins of this family are known to act in cell-cell adhesion and as sheddases. ADAMs have also been proposed to degrade the ECM, but this has not yet been shown in a physiological context. RESULTS: Using a tissue transplantation technique, we show that Xenopus CNC cells overexpressing wild-type ADAM 13 migrate along the same hyoid, branchial, and mandibular pathways used by normal CNC cells. In contrast, CNC cell grafts that express protease-defective ADAM 13 fail to migrate along the hyoid and branchial pathways. In addition, ectopic expression of wild-type ADAM 13 results in a gain-of-function phenotype in embryos, namely the abnormal positioning of trunk neural-crest cells. We further show that explanted embryonic tissues expressing wild-type, but not protease-defective, ADAM 13 display decreased cell-matrix adhesion. Purified ADAM 13 can cleave fibronectin, and tissue culture cells that express wild-type, but not protease-defective, ADAM 13 can remodel a fibronectin substrate. CONCLUSIONS: Our findings support the hypothesis that the protease activity of ADAM 13 plays a critical role in neural crest-cell migration along defined pathways. We propose that the ADAM 13-dependent modification of ECM and/or other guidance molecules is a key step in the directed migration of the CNC.     

Quantitative analyses of changes in cell shapes during bending of the avian neural plate

It is widely believed that changes in cell shapes play important roles in the bending or folding of epithelial sheets, but few studies have actually examined cell shapes in such systems. We have determined the percentages of four types of neuroepithelial cells (i.e., spindle, flask, inverted flask, and globular) present during bending of the avian neural plate. Serial transverse plastic sections through seven craniocaudal levels of the neuroepithelium were examined. Four distinct periods of bending were chosen based on the morphology of the neuroepithelium: period I, flat neural plate; period II, midline furrow without elevation of the neural folds; period III, midline furrow with elevation; and period IV, bilateral furrows with convergence of the neural folds. We compared statistically the percentages of different cell types in bending (furrowed) and nonbending regions of the neuroepithelium, as well as changes in cell shapes with time. Our results demonstrate that dramatic changes in cell shapes occur in the midline and bilateral furrows during bending of the neural plate, such that as many as 70% of the neuroepithelial cells in the midline and 55% in the bilateral furrows are wedge shaped by the end of bending. In contrast, less than 35% of the neuroepithelial cells are wedge shaped outside of the three morphological loci of bending. These results support the hypothesis that localized changes in cell morphologies have roles in bending and shaping of the neural plate, but exactly how cells change shapes and what precise roles such changes play in bending remain to be determined.     

Abnormal elevation of the neural folds in the loop-tail mutant mouse.

The processes of elevation and convergence of the spinal neural folds were analyzed in normal (+/+; Lp/+) and abnormal (Lp/Lp) embryos of the loop-tail mutant mouse in order to determine possible mechanisms underlying the dysraphic defect characterized by a failure of the neural fold to close in this mutant. The results indicate that the neural folds are already defective during very early phases of elevation, with greater distances between the apical points of the paired walls of the neural groove, larger ventral angles and higher ratios of luminal/basal linear distances occurring in the abnormal embryos relative to those in normal embryos. The cross-sectional area of the neuroepithelium is also greater in abnormals, suggesting that faulty elongation of the neuraxis may contribute to the dysraphic condition.     

Aberrant differentiation of neuroepithelial cells in developing mouse brains subsequent to retinoic acid exposure in utero

All-trans-retinoic acid induced 2 types of disorganized neuroepithelium, localized and continuous, in the exencephaly of 9-day-old mouse embryos exposed to 60 or 40 mg/kg for 27 to 30 hr in utero. The localized effect appeared as a protuberance in the wall of the telencephalon and thick neural folds in the mesencephalon with the discontinuity of the apical terminal sheet. The continuous disorganization was seen from the olfactory placode to the myelencephalon with rosettes of cells and many dense bodies in the neuroepithelium. Ultrastructurally, cells in the localized disorganizations showed swelling of Golgi complexes, coated vesicles, and rough endoplasmic reticulum resulting in degeneration. The continuous disorganizations consisted of undifferentiated homogeneous cells in which the nuclei exhibited expansion of nucleolar granular portions and coagulated heterochromatin, and cytoplasm showed monosomal dispersion. In both types of disorganized neuroepithelium, junctional complexes were seen focally at the apical side or apical processes of the rosette, with few or no microfilament bundles. A layer of microfilaments at the base of the neuroepithelial cells in controls, just above the basal lamina, was not present in the monosome dispersed cytoplasm. In the neuroepithelium of controls, one phagosome was seen in the perinuclear region in 0.8% of the cells examined, whereas in the experimental neuroepithelium 2 or more phagosomes were seen in a cell, and phagocytosis occurred by pseudopods. These findings suggest that all-trans-retinoic acid induces not only cytotoxicity but also dedifferentiation in the neuroepithelial cells leading to more cell death, which activates the phagocytosis. These lesions in the neuroepithelium may be a cause of exencephaly.     

Aberrant convergence of the neural folds in the mouse mutant vl.

Progressive changes in the dorsolateral angles (DA) and ventral angle (VA) during elevation and convergence of the caudal neural folds were morphometrically analyzed in normal and dysraphic abnormal embryos of the mouse mutant vacuolated lens (vl), and correlations with the configuration of microfilaments in the apices of neuroepithelial cells were made by means of ultrastructural cytochemistry. In 22-28 somite stage abnormal (vl/vl) embryos, the DA and VA are larger than those in their normal counterparts at each comparable level of the caudal neural folds, suggesting that defective convergence involves both the DA and VA in this mutant. In 30-35 somite stage abnormal embryos, the VA is likewise larger than that in normal embryos in which the neural folds have converged and closed; however, the DAs are much smaller, indicating that a medial collapse of the dorsal ends of the neural folds may occur secondary to the closure failure. At the DA, the ultrastructural configuration of microfilaments is similar in abnormal and normal embryos in terms of their circumferential arrangement around the perimeters of the neuroepithelial cell apices. In abnormal embryos, however, the bundles of microfilaments are more delicate and less prominent than in normal embryos; thus it is possible that a quantitative and/or functional deficiency in these elements may be involved in the failure of the abnormal neuroepithelium to bend properly during convergence of the neural folds.     

Netrin‐1 is required for efficient neural tube closure

During neural tube formation, neural plate cells migrate from the lateral aspects of the dorsal surface towards the midline. Elevation of the lateral regions of the neural plate produces the neural folds which then migrate to the midline where they fuse at their dorsal tips, generating a closed neural tube comprising an apicobasally polarized neuroepithelium. Our previous study identified a novel role for the axon guidance receptor neogenin in Xenopus neural tube formation. We demonstrated that loss of neogenin impeded neural fold apposition and neural tube closure. This study also revealed that neogenin, via its interaction with its ligand, RGMa, promoted cell–cell adhesion between neural plate cells as the neural folds elevated and between neuroepithelial cells within the neural tube. The second neogenin ligand, netrin‐1, has been implicated in cell migration and epithelial morphogenesis. Therefore, we hypothesized that netrin‐1 may also act as a ligand for neogenin during neurulation. Here we demonstrate that morpholino knockdown of Xenopus netrin‐1 results in delayed neural fold apposition and neural tube closure. We further show that netrin‐1 functions in the same pathway as neogenin and RGMa during neurulation. However, contrary to the role of neogenin‐RGMa interactions, neogenin‐netrin‐1 interactions are not required for neural fold elevation or adhesion between neuroepithelial cells. Instead, our data suggest that netrin‐1 contributes to the migration of the neural folds towards the midline. We conclude that both neogenin ligands work synergistically to ensure neural tube closure. © 2012 Wiley Periodicals, Inc., 2013     

Developmental study of neural tube closure in a mouse stock with a high incidence of exencephaly

About 17% of embryos and fetuses in the SELH/Bc mouse stock have the anterior neural tube defect, exencephaly. No other malformations are seen. The genetic liability to exencephaly was shown to be probably genetically fixed in the SELH/Bc stock. This means that SELH/Bc embryos with successful neural tube closure are genetically the same as exencephalics. Females were significantly more likely to be affected than males (66% females). The pattern of morphological developmental events during anterior neural tube closure on days 8 and 9 of gestation was compared among 322 ICR/Bc (normal), 304 SWV/Bc (normal), and 265 SELH/Bc embryos. Anterior neural tube closure was found to follow a strikingly different pattern in almost all SELH/Bc embryos than in either of the normal strains or in previous published studies. SELH/Bc embryos lack the initial contact between the anterior folds in the posterior prosencephalon/anterior mesencephalon region (Closure 2). In spite of this, all but 17% manage to close the anterior neural tube by extending caudally the later occurring normal anterior zone of contact and fusion at the most rostral aspect of the prosencephalon (Closure 3) through the region of Closure 2 to meet the zone of closure of the rhombencephalon, Closure 4. Anterior neural tube closure was completed late, and in some SELH/Bc embryos, elevation and fusion in the mesencephalon did not occur at all. In histological sections of six- and eight-somite embryos, elevated numbers of pyknotic cells in the neuroepithelium and mesenchyme, and elevated numbers of unstained inclusions in the neuroepithelium were found; but their relationship, if any, to the abnormal pattern of neural tube closure is not clear.     

Ultrastructural and histochemical differences in cell surface properties of strain-specific and nonstrain-specific TA3 adenocarcinoma cells

Transmission and scanning electron microscopy and histochemical and biochemical methods were used to investigate differences in cell structure and cell surface properties between the strain-specific TA3- St and nonstrain-specific TA3-Ha ascites sublines of the TA3 murine mammary adenocarcinoma. The TA3-St subline is lethal only to the syngeneic strain A mouse (the strain of origin), whereas the TA3-Ha subline is lethal even to foreign species. In contrast to the TA3-St cell surface, which has numerous folds and irregular microprojections, the TA3-Ha cell has abundant long microvilli of uniform dimensions. An extensive cell surface coat which resembles the "fuzz" coat found on microvilli of normal epithelium was present on the TA3-Ha, but not on the TA3-St cells. After routine fixation, the surface coat of the TA3- Ha cell usually appeared as a filamentous network extending 30-50 nm from the plasmalemma; occasionally, longer filamentous or rod-like structures were found extending 200-400 nm from the plasmalemma. The cell coat material was more extensive on the microvilli than on the intermicrovillous membranes. Free virus-like particles associated with TA3-Ha cells have a similar-appearing surface coat on their outer membranes. The density of surface anionic sites, determined with polycationic ferritin, was greater on the TA3-Ha than on the TA3-St cell surface, consistent with the presence at the TA3-Ha cell surface of several-fold more neuraminidase-susceptible sialic acid groups. The observed surface features of the nonstrain-specific TA3-Ha cell, in comparison to the strain-specific TA3-St cell, are consistent with the suggestion that sialic acid-rich glycoproteins at the TA3-Ha cell surface mask histocompatibility antigens and enhance the ability of malignant cells to invade foreign species.