When primary mouse mammary epithelial cells are cultured on plastic, they rapidly lose their ability to synthesize and secrete most milk proteins even in the presence of lactogenic hormones, whereas cells cultured on released type I collagen gels show greatly enhanced mRNA levels and secretion rates of ,B-casein and of some other milk proteins. We show here that culture on a reconstituted basement membrane from Engelbreth-Holm-Swarm tumor (EHS) allows >90% of cells to produce high levels of 13-casein. By 1-3). Until a decade ago, most investigators cultured cells on plastic surfaces, which led to drastic alterations of morphology and function from the parent tissue (4). Following the example of Emerman and Pitelka (5), we and others have shown that several aspects of functional epithelium can be maintained when primary mouse mammary epithelial cells (PMME) are cultured on "released" (RG or "floating") collagen type I gels instead of plastic. These include polarization of organelles, appearance of apical microvilli, formation of a basal lamina (5, 6), changes in glucose metabolite pattern (7), lumina formation (8), altered synthesis and compartmentalization of extracellular matrix (ECM) components (9, 10), enhancement ofmost milk protein synthesis and secretion (11), and increases in p-casein (12) and transferrin (13,14) mRNA levels. These results signify the importance of cell-substratum interactions in regulating tissue-specific functions and point to a possible regulatory role for ECM in vivo.Cells on released gels (and not on plastic or flat gels) were shown to synthesize an intact basement membrane (5) containing high levels of heparan sulfate and other sulfated glycosaminoglycans (9), type IV collagen, and laminin (10). We reasoned that the released gel may act through the influence of newly synthesized basement membrane components. To test the mechanism of cell-ECM interaction more directly, we have analyzed the consequences of culturing PMME cells on a reconstituted basal lamina derived from Engelbreth-Holm-Swarm (EHS) tumor (15) and on some of its individual components.MATERIALS AND METHODS PMME from 14-to 17-day pregnant BALB/c mice and collagen from rat tail tendon were prepared as described (5, 10, 11). Tumors (EHS) from normal or lathyritic mice were extracted with high salt and 2 M urea as described by Kleinman et al. (15). Dialyzed EHS extract (100-200 1.d) was spread either directly on 35-mm dishes or on top of rat-tail collagen gels. Collagen gels were released 3-4 days after seeding. Matrigel, an EHS preparation from Collaborative Research (Waltham, MA), was used for some experiments. Laminin and type IV collagen (Bethesda Research Laboratories) were spread at 2.5-10 pug/cm2. Heparan sulfate proteoglycan (HSPG; low density form) was prepared from EHS tumors as described by Hassell et al. (16) and was spread at 1-5 ,ug/cm2 or was added to the medium at 5 ,ug/ml every other day. The latter was less toxic but also less effective. EHS (10 ,ug/ml) and individual substrata other than HSPG we...

Two cell lines-Madin-Darby canine kidney (MDCK) and normal murine mammary gland (NMuMG)-growing as monolayers on collagen gels were overlaid with another collagen gel. The cells responded to the overlay by undergoing reorganization resulting in the creation of lumina. MDCK cells formed lumina that coalesced to form large cavities comparable in size with a tubule. NMuMG cells formed clusters surrounding small lumina, which appeared similar to acini of glandular tissue. The characteristic arrangements, described here by light and electron microscopy, resembled the morphology of the tissues of cell line origin. MDCK cells, grown in the presence of serum, formed lumina whether or not serum was removed at the time of overlay, whereas NMuMG cells required either a nondialyzable component of serum or hormonal supplements in serum-free defined media. Lumen formation was delayed by MDCK cells in the presence of the glutamine analog 6-diazo-5-oxo-L-norleucine, but this compound did not affect NMuMG lumen formation. In both cell lines, lumen formation was unaffected by the absence of sulfate, the presence of an inhibitor of sulfated glycosaminoglycan synthesis, or an inhibitor of collagen synthesis. DNA synthesis accompanied lumen formation but was not required.The living stroma is typically separated from the exterior by a delimiting interface of epithelial cells. Epithelial cells can exist as stratified layers that may directly face the exterior, as for example in epidermis, or as a monolayer that may be folded into circular acini or ducts containing lumina continuous with the exterior. Certain specialized epithelial cells, such as those in the thyroid or cornea, border internal fluid-filled cavities that are closed to the exterior. The epithelial cell is functionally and morphologically polarized with respect to the two environments it separates. The basal cell surface is attached to extracellular matrix (ECM) material specialized as a layered basal lamina adjacent to the ECM of the stroma. In most epithelia, the opposite apical surface is free from an apposed extracellular layer. The rim of the apical surface is girdled by tight junctions to adjacent cells, so that the cell layer serves as a boundary to molecular diffusion. From the histological organization of epithelium, it appears that attachment to the ECM has a function in directing the polarity. This is strongly reinforced by the histology ofinvasive tumors. Metastatic epithelial tumor cells secrete hydrolytic enzymes that degrade the basal lamina before they invade the stroma (1). Tumor cells that have degraded the basal lamina lose their polarity, but if they maintain contact with undegraded basal lamina, they remain polar (2).The purpose of the work described here is to establish morphogenetic models in culture in which ECM material, in this case type I collagen, induces epithelial rearrangement and thus provide a system for analysis of the biochemical steps involvedThe publication costs ofthis article were defrayed in part by page charge payment. Thi...

We have successfully generated and characterized a stable packaging cell line for HIV-1-based vectors. To allow safe production of vector, a minimal packaging construct carrying only the coding sequences of the HIV-1 gag-pol, tat, and rev genes was stably introduced into 293G cells under the control of a Tet(o) minimal promoter. 293G cells express the chimeric Tet(R)/VP16 trans-activator and contain a tetracycline-regulated vesicular stomatitis virus protein G (VSV-G) envelope gene. When the cells were grown in the presence of tetracycline the expression of both HIV-1-derived and VSV-derived packaging functions was suppressed. On induction, approximately 50 ng/ml/24 hr of Gag p24 equivalent of vector was obtained. After introduction of the transfer vector by serial infection, vector could be collected for several days with a transduction efficiency similar or superior to that of vector produced by transient transfection both for dividing and growth-arrested cells. The vector could be effectively concentrated to titers reaching 10(9) transducing units/ml and allowed for efficient delivery and stable expression of a GFP transgene in the mouse brain. The packaging cell line and all vector producer clones described here were shown to be free from replication-competent recombinants, and from recombinants between packaging and vector constructs that transfer the viral gag-pol genes. The packaging cell line and the assays developed will advance lentiviral vectors toward the stringent requirements of clinical applications.

No abstract

AAV producer cell lines can be readily scaled to meet the needs of clinical trials. One 500 L bioreactor of these producer cells can produce the equivalent of 2500 high capacity roller bottles or 25 000 T-175 tissue culture flasks.