Collagenous substrata regulate the nature and distribution of glycosaminoglycans produced by differentiated cultures of mouse mammary epithelial cells

We have investigated the influence of culture substrata upon glycosaminoglycans produced in primary cultures of mouse mammary epithelial cells isolated from the glands of late pregnant mice. Three substrata have been used for experiments: tissue culture plastic, collagen (type I) gels attached to culture dishes, and collagen (type I) gels that have been floated in the culture medium after cell attachment. These latter gels contract significantly. Cells cultured on all three substrata produce hyaluronic acid, heparan sulfate, chondroitin sulfates and dermatan sulfate but the relative quantities accumulated and their distribution among cellular and extracellular compartments differ according to the nature of the culture substratum. Notably most of the glycosaminoglycans accumulated by cells on plastic are secreted into the culture medium, while cells on floating gels incorporate almost all their glycosaminoglycans into an extracellular matrix fraction. Cells on attached collagen gels secrete approx. 30% of their glycosaminoglycans and assemble most of the remainder into an extracellular matrix. Hyaluronic acid is produced in significant quantities by cells on plastic and attached gels but in relatively reduced quantity by cells on floating gels. In contrast, iduronyl-rich dermatan sulfate is accumulated by cells on floating gels, where it is primarily associated with the extracellular matrix fraction, but is proportionally reduced in cells on plastic and attached gels. The results are discussed in terms of polarized assembly of a morphologically distinct basal lamina, a process that occurs primarily when cells are on floating gels. In addition, as these cultures secrete certain milk proteins only when cultured on floating gels, we discuss the possibility that cell synthesized glycosaminoglycans and proteoglycans may play a role in the maintenance of a differentiated phenotype.     

Expression of extracellular matrix components is regulated by substratum

Reconstituted basement membranes and extracellular matrices have been demonstrated to affect, positively and dramatically, the production of milk proteins in cultured mammary epithelial cells. Here we show that both the expression and the deposition of extracellular matrix components themselves are regulated by substratum. The steady-state levels of the laminin, type IV collagen, and fibronectin mRNAs in mammary epithelial cells cultured on plastic dishes and on type I collagen gels have been examined, as has the ability of these cells to synthesize, secrete, and deposit laminin and other, extracellular matrix proteins. We demonstrate de novo synthesis of a basement membrane by cells cultured on type I collagen gels which have been floated into the medium. Expression of the mRNA and proteins of basement membranes, however, are quite low in these cultures. In contrast, the levels of laminin, type IV collagen, and fibronectin mRNAs are highest in cells cultured on plastic surfaces, where no basement membrane is deposited. It is suggested that the interaction between epithelial cells and both basement membrane and stromally derived matrices exerts a negative influence on the expression of mRNA for extracellular matrix components. In addition, we show that the capacity for lactational differentiation correlates with conditions that favor the deposition of a continuous basement membrane, and argue that the interaction between specialized epithelial cells and stroma enables them to create their own microenvironment for accurate signal transduction and phenotypic function.     

Extracellular matrix control of collagen production by rat mammary epithelial and myoepithelial-like cells in vitro

A rat mammary myoepithelial cell line (Rama 401) grown on plastic produces 5 times more collagen (largely type IV) than a mammary epithelial cell line (Rama 704) grown on the same surface. When the cells are grown on collagen gels, the amount of collagen produced by Rama 704 cells increases 3.3 times, whereas there is no increase in collagen production by Rama 401 cells. Increased production of collagen by Rama 704 cells is due to both an increased rate of synthesis and a decreased rate of degradation. These results indicate that for mammary epithelial cells, unlike myoepithelial cells, the rate of production of collagen can be regulated by the extracellular matrix.     

Biosynthesis of sulphated macromolecules by rabbit lens epithelium. II. Relationship to basement membrane formation

Rabbit lens epithelial cells display a similar "cobblestone" morphology and produce the same complement of sulphated macromolecules (also see Heathcote, J.G., and R.W. Orkin, 1984, J. Cell Biol., 99:852-860) whether grown on plastic or glass, dried films of gelatin or type IV collagen with laminin, or on gels of type I collagen. There was no evidence of basement membrane formation by these cells when they were grown on plastic, glass, or dried films. In contrast, cultures that had been grown on gels deposited a discrete basement membrane that followed the contours of the basal surfaces of the cells and in addition, they secreted amorphous basement membrane-like material that diffused into the interstices of the gel and associated with the collagen fibrils of the gel. A significant proportion (approximately 70%) of the heparan sulphate proteoglycan fraction that was secreted into the culture medium (fraction MI) when the cells were grown on plastic became associated with the cell-gel layer in the gel cultures. Further, when basement membrane was isolated by detergent extraction, greater than 90% of the 35S-labeled material present was in this heparan sulphate proteoglycan.     

Characterization of the intermediate filament proteins of Murine Mammary Gland epithelial cells : Response to collagen substratum

The insoluble cytoskeletal material remaining after detergent lysis of 'Normal' Murine Mammary Gland (NMuMG) cells, growing on plastic or collagen gel substrata, was analyzed by two-dimensional gel electrophoresis. The identity of the cytoskeletal elements was determined by their solubility properties, electrophoretic separation pattern, and immunoreactivity using monoclonal antibodies against intermediate filament proteins (AIF), keratins (AE1 and AE3) and actin. The electrophoretic pattern of the cytoskeletal elements from the NMuMG cell strain was found to be very similar to that of primary mouse mammary epithelial cells. Both NMuMG and primary mammary epithelial cells when grown on collagen exhibited an increased expression of a 49 kD protein with a pI of 5.6, that appeared to be a cytokeratin. Many of the cytoskeletal proteins remained tightly attached to the collagen gel substratum after cell lysis. These results demonstrate that the NMuMG cell strain has retained a stable expression of cytokeratins that remains responsive to the presence of extracellular matrix material.     

Effect of the extracellular matrix on plasminogen activator isozyme activities of human mammary epithelial cells in culture.

Multiple molecular forms of plasminogen activator were detected in normal human mammary epithelial cells in culture. Cells derived from (normal) breast mammoplasty specimens and grown on the surface of collagen gels exhibited three major classes of plasminogen activator isozymes (Mr = 100,000 [100K], 75,000 [75K], and 55,000 [55K]). The activity of the 100K and 75K isozymes was greatly reduced when the cells were grown on conventional tissue-culture-grade plastic surfaces. MCF-7, a human mammary carcinoma cell line, exhibited predominantly or exclusively the 55K isozyme, irrespective of the cell growth substratum. The activity of the 55K isozyme was more than twofold higher for MCF-7 cells grown on collagen gels than for cells grown on plastic. Progesterone, diethylstilbestrol, and estrogen stimulated the activity of the 55K isozyme of MCF-7 cells, but only when the cells were grown on a plastic surface. The plasminogen activator activities of the normal human mammary epithelial cells were not stimulated by these hormones, irrespective of the growth substratum. These results show that the expression of plasminogen activator isozymes by human mammary epithelial cells is subject to modulation by the extracellular matrix. Normal and malignant cells may differ in their responsiveness to these effects.     

Characterization of human mammary cell types in primary culture: Immunofluorescent and immunocytochemical indicators of cellular heterogeneity

Summary Parenchymal organoidal structures that were obtained from collagenase digestion of reduction mammoplasty specimens of apparently normal human breasts have been grown in short-term primary cultures, either on plastic or on floating gels of polymerized rat-tail collagen. Three morphologically distinct major cell types are readily observed in both systems: cuboidal cells, which occupy apical positions on collagen gels; larger, epithelioid, or basal cells on gels; and elongated cells which penetrate into the gel. In addition, a fourth cell type, that of a large, flat cell, is observed less readily by phase contrast microscopy on the surface of cultures grown on plastic. Immunofluorescent and immunocytochemical staining of cultures on plastic or histologic sections of cultures on gels have been undertaken with antisera and other histochemical reagents that stain the different parenchymal cell types in vivo. Thus antisera to epithelial membrane antigen(s), monoclonal antibodies (MABs) to the defatted mammary milk fat globule membrane, peanut lectin, and keratin MAB LE61, which preferentially stain the epithelial cells of ducts in vivo, also stain the cuboidal/apical cells in vitro. The large, flat cells are stained intensely by the first three reagents but not by the last one. Antisera to collagen IV, laminin, fibronectin, actin, keratin MAB LP34, MABs to the common acute lymphoblastic leukemia antigen, and MAB LICR-LON-23.10, which showed enhanced staining for the ductal myoepithelial cells in vivo, also stain the epithelioid/elongated cells in vitro. However, the effect of the last four reagents is reduced considerably in most elongated cells, and MAB LP34 stains the large, flat cells intensely. Heterogeneous cells of intermediate morphologies and staining patterns between the cuboidal/flat cells and large epithelioid cells have also been identified. The results suggest that the cuboidal cells and large, flat cells are related to mammary epithelial cells, whereas the large epithelioid/elongated cells have some characteristics of myoepithelial cells, and that intermediate forms may exist in culture between the two parenchymal cell types. This work was supported in part by the Ludwig Institute for Cancer Research and the Cancer and Polio Research Fund. Dr. M. J. Warburton is supported by the Cancer Research Campaign.     

Synthesis of basement membrane proteins by rat mammary epithelial cells

A mammary epithelial cell line, Rama 25, growing on plastic, deposits fibronectin, type IV collagen, and laminin in punctate structures located beneath the basal surface of the cells. When grown on the surface of collagen gels, Rama 25 cells deposit these basement membrane proteins in a continuous layer between the basal surface of the cells and the surface of the collagen matrix. Rama 25 cells also penetrate the collagen matrix forming rudimentary duct-like structures. These structures are surrounded by a discontinuous layer of basement membrane proteins. The ducts of fetal and neonatal rat mammary glands contain few mature myoepithelial cells and our results suggest that some mammary epithelial cells, in contact with a collagenous stroma, are capable of synthesizing a basal lamina-like structure.     

Loss of growth inhibitory activity of TGF-beta toward normal human mammary epithelial cells grown within collagen gel matrix

TGF-beta at concentrations in the range from 0.1 to 10 ng/ml gave significant growth inhibition of nonmalignant human mammary epithelial cells (HMEC) but not of malignant HMEC grown in monolayer cultures in serum-free medium. However, no growth inhibition of the nonmalignant cells was observed when the cells were cultivated within a type-I collagen gel matrix either adhering to a plastic substratum or floating on the medium. Within floating collagen gels, both nonmalignant and malignant HMEC formed a cell mass having radial extensions, and TGF-beta at 1 or 10 ng/ml prevented the formation of extensions only in the nonmalignant HMEC.     

Rapid communication Improved growth of cultured brain microvascular endothelial cells on glass coated with a biological matrix

An improved method for culturing primary rat brain capillary endothelial cells on glass has been developed, using a corneal extracellular matrix coat. Since the collagen-coated plastic attachment surface conventionally used for primary cultures of brain microvascular endothelium gives a high level of background fluorescence in microfluorimetric studies, an alternative attachment surface was tested involving no plastic element. Five substrata combinations were examined and a new combination of glass and corneal endothelial extracellular matrix coat was found to provide excellent cell adhesion, culture growth and purity. Other established substrata combinations tested for comparison, either involved plastic, or used glass with collagen or carbodiimide and collagen coating but the last two gave poor endothelial cell adhesion and growth. Our method using this new attachment surface combination results in stable and pure endothelial cultures, as verified by immunocytochemistry, which are suitable for fluorimetric investigations.     

A comparison of the effects of different substrata on chondrocyte morphology and the synthesis of collagen types IX and X

Summary Embryonic chick sternal chondrocytes were cultured either within three dimensional gels of type I collagen, type II collagen or agar, or as monolayers on plastic dishes coated with air-dried films of these matrix macromolecules. It was observed that cell shape and cell growth varied markedly between the different culture conditions. Flattened monolayers of cells on plastic or films of type I or type II collagen, proliferated more rapidly and reached a higher final cell density per culture than the more rounded cells found in the cultures on agar films or within three-dimensional gels. Biosynthetic studies demonstrated that in addition to the synthesis of type II collagen, all the cultures were producing collagen types IX and X. Chondrocytes cultured on plastic or films of the different matrix macromolecules all showed a similar expression of types IX and X collagen, independent of whether they displayed a flattened or round cell morphology. In contrast, marked variations in the proportions of the minor collagens, particularly type X collagen, were observed when the cells were cultured within three-dimensional gels. The data suggest that direct interaction of the cell surface with matrix constituents displaying a particular spatial array could be an important aspect in the control of type IX and X collagen expression by chondrocytes. The financial support of the Arthritis & Rheumatism Council and the Medical Research Council is gratefully acknowledged.     

Improved growth of cultured brain microvascular endothelial cells on glass coated with a biological matrix

An improved method for culturing primary rat brain capillary endothelial cells on glass has been developed, using a corneal extracellular matrix coat. Since the collagen-coated plastic attachment surface conventionally used for primary cultures of brain microvascular endothelium gives a high level of background fluorescence in microfluorimetric studies, an alternative attachment surface was tested involving no plastic element. Five substrata combinations were examined and a new combination of glass and corneal endothelial extracellular matrix coat was found to provide excellent cell adhesion, culture growth and purity. Other established substrata combinations tested for comparison, either involved plastic, or used glass with collagen or carbodiimide and collagen coating but the last two gave poor endothelial cell adhesion and growth. Our method using this new attachment surface combination results in stable and pure endothelial cultures, as verified by immunocytochemistry, which are suitable for fluorimetric investigations.     

Effect of cell shape change on the function and differentiation of rabbit mammary cells in culture

We examined the role of cell shape, cytodifferentiation, and tissue topography on the induction and maintenance of functional differentiation in rabbit mammary cells grown as primary cultures on two-dimensional collagen surfaces or in three-dimensional collagen matrices. Mammary glands from mid-pregnant rabbits were dissociated into single cells, and epithelial cells were enriched by isopycnic centrifugation. Small spheroids of epithelial cells (approximately 50 cells) that formed on a rotary shaker were plated on or embedded in collagen gels. The cells were cultured for 1 d in serum-containing medium and then for up to 25 d in chemically defined medium. In some experiments, epithelial monolayers on gels were mechanically freed from the dishes on day 2 or 5. These gels retracted and formed floating collagen gels. On attached collagen gels, flat monolayers of a single cell type developed within a few days. The cells synthesized DNA until the achievement of confluence but did not accumulate milk proteins. No morphological changes were induced by prolactin (PRL). On floating gels, two cell types appeared in the absence of cell proliferation. The cells in direct contact with the medium became cuboidal and developed intracellular organelles typical of secretory cells. PRL-induced lipogenesis, resulting in large fat droplets filling the apical cytoplasm and accumulation of casein and α-lactalbumin in vesicles surrounding the fat droplets. We detected tranferrin in the presence or absence of PRL intracellularly in small vesicles but also in the collagen matrix in contact with the cell layer. The second cell type, rich in microfilaments and reminiscent of the myoepithelial cells, was situated between the secretory cell layer and the collagen matrix. In embedding gels, the cells formed hollow ductlike structures, which grew continuously in size. Secretory cells formed typical lumina distended by secretory products. We found few microfilament-rich cells in contact with the collagen gels. Storage and secretion of fat, caseins and alpha-lactalbumin required the presence of PRL, whereas the accumulation and vectorial discharge of transferrin was prolactin independent. There was no differentiation gradient between the tip and the cent of the outgrowth, since DNA synthesis and milk protein storage were random along the tubular structures. These results indicate that establishment of functional polarity and induction of cytodifferentiation are influenced by the nature of the interaction of the cells with the collagen structure. The morphological differentiation in turn plays an important role in the synthesis, storage, and secretion of fat and milk proteins.     

Influence of collagen gel on the orientation of epithelial cell polarity: follicle formation from isolated thyroid cells and from preformed monolayers

The influence of collagen gels on the orientation of the polarity of epithelial thyroid cells in culture was studied under four different conditions. (a) Isolated cells cultured on the surface of a collagen gel formed a monolayer. The apical pole was in contact with the culture medium and the basal membrane was attached to the substratum. (b) Isolated cells embedded inside the gel organized within 8 into follicles. The basal pole was in contact with collagen and the apical pole was oriented towards the interior of the follicular lumen. (c) Cells were first organized into floating vesicles, structures in which the apical surface is in contact with the culture medium, and the vesicles were embedded inside the collagen gel. After 3 d, cell polarity was inverted, the apical pole being oriented towards the cavity encompassed by cells. Vesicles had been transformed into follicles. (d) Monolayers formed on collagen gels as in a were overlaid with a second layer of collagen, which was polymerized in contact with the apical cell surface. A disorganization of the continuous pavement occurred within 24 h; cells attached to the upper layer of collagen and reorganized into follicles in the collagen sandwich within 4-8 d. A similar process occurred when the monolayer was grown on plastic and overlaid with collagen, or grown on collagen and covered with small pieces of glass cover slips. No reorganization was observed between two glass surfaces. In conclusion, first, a basal pole was always formed in the area of contact between the cell membrane and an adhesive surface and, second, the interaction of a preformed apical pole with an adhesive surface was not compatible with the stability of this domain of the plasma membrane. The interaction of the cell membrane with extracellular components having adhesive properties appears to be a determinant factor in the orientation and stabilization of epithelial cell polarity.     

Effects of matrix components on aromatase activity in breast stromal cells in culture

Local estradiol production within breast tissue is maintained by the aromatase cytochrome P450_arom complex, which has been localized primarily to the stromal component of tumors but also has been detected in the breast epithelial cells. Paracrine interactions between stromal and epithelial components of the breast are critical to the sustained growth and progression of breast tumors. Maintenance of the differentiated state, including hormone and growth factor responsiveness, requires extracellular matrix proteins as substrata for cells. This research has focused on developing a cell culture system that more closely mimics in vivo interactions in order to dissect actual paracrine signaling between these two cell types. Human fibroblasts were isolated from breast tissue and were maintained in a cell culture system grown on plastic support or on a collagen I support matrix. The collagen I matrix model supports cell maintenance and subsequent differentiation on collagen rather than maximal proliferation, therefore allowing for a more accurate environment for the study of hormonal control and cellular communication. Initial experiments compared aromatase activity of patient fibroblasts grown on plastic versus collagen I using the tritiated water release method. Constitutive aromatase activity was found to be lower when cells were grown on a collagen gel for 4–7 days (7.7 fold lower) using DMEM/F12 containing 10% dextran coated charcoal stripped serum. However, fibroblasts grown on collagen I appeared to be significantly more responsive to stimulation by 100 nM dexamethasone (plastic: 6.0 fold induction, collagen: 33.2 fold induction) when pretreated for 12 h prior to measurement of aromatase activity. In an effort to examine paracrine interactions between the stromal and epithelial cells in breast tissue, experiments using conditioned media from fibroblast cultures were performed. Testosterone administration to fibroblasts results in the production of estradiol into the media in sufficient concentrations to elicit an increase in pS2 expression when the conditioned media is administered to MCF-7 cells. The addition of a potent aromatase inhibitor resulted in a complete suppression of fibroblast-derived estrogens and showed only a modest increase in pS2 expression. Culturing breast fibroblasts and epithelial cells on extracellular matrix allows for a more meaningful examination of the paracrine interactions between these cell types within the context of an appropriate extracellular environment. This study highlights the need for evaluation of gene expression in cell culture systems that accurately reflect the tissue microenvironment.     

Reconstituted basement-membrane matrix modulates fibroblast activities in vitro

Cellular growth and collagen biosynthesis were compared in dermal calf fibroblasts cultured on plastic or on a reconstituted basement membrane gel, termed matrigel. This matrix, extracted from Engelbreth-Holm-Swarm tumors, consists mainly of laminin, entactin, type IV collagen, and heparan sulfate proteoglycan. The multiplication rate of fibroblasts grown on matrigel was stimulated compared to that of monolayered cells cultured on plastic, and these cells formed multilayers after 4 days. Protein and collagen biosynthesis was reduced in fibroblasts cultured on matrigel. A higher proportion of the newly synthesized collagen (40%) was incorporated to the extracellular matrix in cultures grown on matrigel than in those grown on plastic (14%). Type III collagen was the preferential collagen type deposited on matrigel, and the ratio of type III:type I collagens secreted in the medium was also slightly higher in cultures grown on matrigel. Partially processed collagen was more abundant in fibroblasts grown on matrigel than in cells cultured on plastic. Finally, cells grown on matrigel exhibited a higher catabolic activity than cells grown on plastic. In this experimental model, the reconstituted basement-membrane matrix seems to influence the activities of fibroblasts significantly.     

A novel artificial substrate for cell culture: Effects of substrate flexibility/malleability on cell growth and morphology

Summary Gels of glyoxyl agarose (GA) are evaluated as a novel flexible substrate for cell culture with physical properties comparable to extracellular matrix (ECM) gels. We show here that cells adhere well to pure GA gels; in addition, specific interactions involving matrix receptors can be studied when individual matrix molecules are bound to the gel covalently. When cells are grown on such substrates, morphology is comparable to that observed on “natural” matrix gels (reconstituted gels of collagen type I or of Matrigel): rather than being flattened as in monolayer cultures on tissue culture plastic the cells assume a rounded morphology and tend to form tissue-like aggregates. The effects of the artificial matrix gels are discussed in the context of previous publications on cell interactions with the extracellular matrix, suggesting that in addition to specific recognition of matrix molecules the physical properties of ECM by themselves can be decisive for cell differentiation. We conclude that gels of glycoxyl agarose a) provide a useful model to mimic the physical properties of matrix gels without the presence of specific adhesion factors; b) may be useful as a general, non-specific ECM allowing cells to be cultured in vitro under conditions favorable for differentiation; and c) allow to design a variety of “synthetic” ECM models composed of a chemically defined gel matrix, which can be supplemented with covalently bound molecules to be recognized by cell surface receptors.     

Control of type IV collagen production in rat mammary epithelial and myoepithelial-like cells

A rat mammary myoepithelial-like cell line (Rama 401) produces 3.5 times more type IV collagen than a mammary epithelial cell line (Rama 25), as measured by the formation of protein hydroxyproline. However, using quantitative "dot" hybridization techniques, the level of poly (A)-containing mRNA hybridizing to a type IV collagen cDNA probe is only 50% higher in Rama 401 cells than in Rama 25 cells. The total amount of hydroxyproline synthesized per cell by the two cell lines is similar. However, in the Rama 25 cells approximately 70% of the hydroxyproline is found as free hydroxyproline against 13% for Rama 401 cells. When Rama 25 cells are grown on collagen gels, they accumulate 2.5-fold more type IV collagen. However, type IV collagen mRNA levels are only 30% higher in Rama 25 cells grown on collagen. The total amount of hydroxyproline synthesized is the same as cells grown on plastic, whereas the extent of collagen degradation is reduced from 71% to 30% in cells grown on collagen gels. No degradation of type IV collagen can be detected in the culture medium of Rama 25 cells. These results indicate that the increased accumulation of type IV collagen in Rama 401 cells is not due to increased synthesis but to a decreased rate of intracellular degradation, and that for Rama 25 cells, the extracellular matrix modulates type IV collagen production by regulating the rate of intracellular collagen degradation.     

Establishment and characterization of a caprine mammary epithelial cell line (CMEC)

Summary We describe the establishment of a continuous, nontransformed cell line obtained from primary culture of a lactating (114 days postparturition) Anglo-Nubian (Capra hircus) goat mammary gland biopsy. These cells (CMEC), have been cultured in the presence of supraphysiologic concentrations of insulin and hydrocortisone for more than 560 population doublings (over 80 passages) without any sign of senescence while maintaining a normal/near-normal diploid chromosome modal number of 2n=60 and are responsive to contact inhibition of proliferation. Cytoskeletal analysis indicates that CMECs are epithelial, without detectable fibroblastic or myoepithelial cells. When grown at low density on plastic substratum, the cells tend to form island monolayer aggregates with the characteristics cobblestone morphology of epithelial cells. With increasing density, the cells organize into lumen-like structures with various morphology consisting of large and small vacuolized and nonvacuolized cells. Postconfluent cultures form epithelial raised dome-like structures, implying a process of contact-induced differentiation. This is corroborated by positive immunocytochemistry to lactation-specific proteins: β-casein and α-lactalbumin, which were predominantly expressed in dome-forming cells. We also observed an overall modulation of cytokeratin 18/19 expression associated with number of days post subculture and with the expression of lactation-specific proteins. Postconfluent cultures which contain lactation-specific, antibody-reactive, dome-like structures showed a decreased expression of keratin 18 and no (null) expression for keratin 19. Lastly, cells cultured within a collagen matrix show morphological differentiation with the organization of branching duct-like and acini-like structures. This study suggests that CMECs are a useful in vitro model for study of mammary gland development and differentiation, in particular, direct modulation of epithelial cells grown on plastic substratum or extracellular matrix without the influence of stromal elements or the necessity and variability associated with primary cell culture or tissue explants.     

Collagen processing in ras-transfected mouse mammary epithelial cells

A mouse mammary epithelial cell line (NMuMG), after transfection with the c-rasH oncogene, forms invasive tumors in nude mice. NMuMG and NMuMG/p-rasH cells produce similar amounts of collagen (mostly type IV) when grown on plastic. NMuMG cells respond to growth on collagen gels by increasing the rate of collagen synthesis and deposition by 100%, unlike NMuMG/p-rasH cells which synthesize similar amounts of collagen whether grown on plastic or collagen gels. These results suggest that ras transformation partially inhibits the interaction between epithelial cells and the surrounding stroma that is necessary for basement membrane deposition in vivo and consequently may facilitate the invasion of the stroma by transfected cells.