Animal and translational models of SARS-CoV-2 infection and COVID-19.

COVID-19 is causing a major once-in-a-century global pandemic. The scientific and clinical community is in a race to define and develop effective preventions and treatments. The major features of disease are described but clinical trials have been hampered by competing interests, small scale, lack of defined patient cohorts and defined readouts. What is needed now is head-to-head comparison of existing drugs, testing of safety including in the background of predisposing chronic diseases, and the development of new and targeted preventions and treatments. This is most efficiently achieved using representative animal models of primary infection including in the background of chronic disease with validation of findings in primary human cells and tissues. We explore and discuss the diverse animal, cell and tissue models that are being used and developed and collectively recapitulate many critical aspects of disease manifestation in humans to develop and test new preventions and treatments.

Tracing the fate of limbal epithelial progenitor cells in the murine cornea.

Stem cell (SC) division, deployment, and differentiation are processes that contribute to corneal epithelial renewal. Until now studying the destiny of these cells in a living mammal has not been possible. However, the advent of inducible multicolor genetic tagging and powerful imaging technologies has rendered this achievable in the translucent and readily accessible murine cornea. K14CreER(T2)-Confetti mice that harbor two copies of the Brainbow 2.1 cassette, yielding up to 10 colors from the stochastic recombination of fluorescent proteins, were used to monitor K-14(+) progenitor cell dynamics within the corneal epithelium in live animals. Multicolored columns of cells emerged from the basal limbal epithelium as they expanded and migrated linearly at a rate of 10.8 µm/day toward the central cornea. Moreover, the permanent expression of fluorophores, passed on from progenitor to progeny, assisted in discriminating individual clones as spectrally distinct streaks containing more than 1,000 cells within the illuminated area. The centripetal clonal expansion is suggestive that a single progenitor cell is responsible for maintaining a narrow corridor of corneal epithelial cells. Our data are in agreement with the limbus as the repository for SC as opposed to SC being distributed throughout the central cornea. This is the first report describing stem/progenitor cell fate determination in the murine cornea using multicolor genetic tracing. This model represents a powerful new resource to monitor SC kinetics and fate choice under homeostatic conditions, and may assist in assessing clonal evolution during corneal development, aging, wound-healing, disease, and following transplantation.

Advanced glycation end-products (AGEs) and functionality of reverse cholesterol transport in patients with type 2 diabetes and in mouse models.

AIMS/HYPOTHESIS: We investigated the contribution of AGEs to the impairment of reverse cholesterol transport (RCT) variables in diabetic individuals and in two animal models of diabetic obesity and of renal impairment. METHODS: The capacity of plasma and HDL from 26 individuals with moderately controlled type 2 diabetes to support cholesterol efflux was compared with 26 age- and sex-matched individuals without diabetes. We also compared the rates of RCT in vivo in two animal models: db/db mice and mice with chronic renal failure. RESULTS: Diabetic individuals had characteristic dyslipidaemia and higher levels of plasma AGEs. The capacity of whole plasma, ApoB-depleted plasma and isolated HDL to support cholesterol efflux was greater for diabetic patients compared with controls despite their lower HDL-cholesterol levels. The capacity of plasma to support cholesterol efflux correlated with plasma levels of cholesteryl ester transfer protein and levels of ApoB, but not with levels of AGE. RCT was severely impaired in db/db mice despite elevated HDL-cholesterol levels and no change in AGE concentration, whereas RCT in uraemic mice was unaffected despite elevated AGE levels. CONCLUSIONS/INTERPRETATION: AGEs are unlikely to contribute significantly to the impairment of RCT in type 2 diabetes.

BRM and BRG1 subunits of the SWI/SNF chromatin remodelling complex are downregulated upon progression of benign skin lesions into invasive tumours.

Background Nonmelanoma skin cancer is caused by exposure to ultraviolet radiation within sunlight. Actinic keratoses (AKs) are benign precursor lesions that can develop into invasive squamous cell carcinoma (SCC). Little is known about the molecular events that lead to human skin cancer progression from benign to invasive. Objectives To determine novel genes that may be involved in skin cancer progression based on data from an initial microarray screen of human skin cancers. Methods The SWI/SNF chromatin remodelling ATPase subunit BRM was identified as being downregulated in SCC but not AK compared with normal skin in our microarray screen. Therefore reverse transcription-polymerase chain reaction, gene methylation and protein expression was used to study BRM and its alternative ATPase subunit BRG1 in a range of human skin cancers. Results We found reduced levels of mRNA coding for BRM but not BRG1 in SCC. BRM mRNA levels in AK were similar to those in normal skin. Deregulation of BRM did not result from hypermethylation of CpG regions in the promoter of these genes. Both BRM and BRG1 protein was reduced by about 10-fold in 100% of SCC and basal cell carcinoma, but not in AK specimens examined. Conclusions BRM protein may be decreased due to low levels of mRNA, while BRG1 protein loss appears to be post-translational. BRM and BRG1 may be novel tumour suppressor genes for human skin cancer. They appear to be involved after development of benign lesions, and are downregulated during progression towards invasion.

Effects of divergent selection for serum insulin-like growth factor-I concentration on performance, feed efficiency, and ultrasound measures of carcass composition traits in Angus bulls and heifers.

Angus bulls and heifers from lines divergently selected for serum IGF-I concentration were used to evaluate the effects of IGF-I selection line on growth performance and feed efficiency in 2 studies. In study 1, bulls (low line, n = 9; high line, n = 8; initial BW = 367.1 +/- 22.9 kg) and heifers (low line, n = 9; high line, n = 13; initial BW = 286.4 +/- 28.6 kg) were adapted to a roughage-based diet (ME = 1.95 Mcal/kg of DM) for 24 d and fed individually for 77 d by using Calan gate feeders. In study 2, bulls (low line, n = 15; high line, n = 12; initial BW = 297.5 +/- 34.4 kg) and heifers (low line, n = 9; high line, n = 20; initial BW = 256.0 +/- 25.1 kg) were adapted to a grain-based diet (ME = 2.85 Mcal/kg of DM) for 32 d and fed individually for 70 d by using Calan gate feeders. Blood samples were collected at weaning and at the start and end of each study, and serum IGF-I concentration was determined. Residual feed intake (RFI) was calculated, within study, as the residual from the linear regression of DMI on midtest BW(0.75), ADG, sex, sex by midtest BW(0.75) and sex by ADG. In study 1, calves from the low IGF-I selection line had similar initial and final BW and ADG, compared with calves from the high IGF-I selection line. In addition, DMI and feed conversion ratio were similar between IGF-I selection lines; however, calves from the low IGF-I selection line tended (P < 0.10) to have lesser RFI than calves from the high IGF-I selection line (-0.26 vs. 0.24 +/- 0.31 kg/d). In study 2, IGF-I selection line had no influence on performance or feed efficiency traits. However, there was a tendency (P = 0.15) for an IGF-I selection line x sex interaction for RFI. Bulls from the low IGF-I selection line had numerically lesser RFI than those from the high IGF-I selection line, whereas in heifers, the IGF-I selection line had no effect on RFI. In studies 1 and 2, weaning and initial IGF-I concentrations were not correlated with either feed conversion ratio or RFI. However, regression analysis revealed a sex x IGF-I concentration interaction for initial IGF-I concentration in study 1 and weaning IGF-I concentration in study 2 such that the regression coefficient was positive for bulls and negative for heifers. These data suggest that genetic selection for postweaning serum IGF-I concentration had a minimal effect on RFI in beef cattle.

The interaction of metal ions and Marimastat with matrix metalloproteinase 9.

The effect of a range of metal ions on the ability of Marimastat to inhibit matrix metalloproteinase 9 (MMP-9) was examined in a fluorescence based proteolytic assay. Whilst none of the metals examined significantly affected the inhibitory ability of Marimastat, several metal ions did have a significant effect on MMP-9 activity itself. In the absence of Marimastat, Zn(II) and Fe(II) significantly inhibited MMP-9 activity at metal ion concentrations of 10 and 100 microM, respectively. In both the absence and presence of Marimastat, Cd(II) significantly inhibited MMP-9 at 100 microM. In contrast, 1 mM Co(II) significantly upregulated MMP-9 proteolytic activity.

Decreased matrix degradation in diabetic nephropathy: effects of ACE inhibition on the expression and activities of matrix metalloproteinases.

AIMS/HYPOTHESIS: Extracellular matrix accumulation is thought to be involved in the pathogenesis of diabetic nephropathy. Increased matrix synthesis has been well documented but the effects of diabetes on degradative pathways, particularly in the in vivo setting, have not been fully explored. Furthermore, the effect of renoprotective therapies on matrix accumulation through these pathways has not been examined. We investigated the degradative pathway of type IV collagen and the effects of ACE inhibition in experimental diabetic nephropathy. METHODS: Diabetes was induced in 16 rats by administrating streptozocin; 8 of the diabetic rats were allocated at random to receive the ACE inhibitor perindopril (2 mg/l) in their drinking water and 8 age and weight matched rats served as controls. Gene expression of matrix metalloproteinase ( MMP) and tissue inhibitor of metalloproteinase ( TIMP) was measured by RT-PCR and type IV collagen content by immunohistochemistry. MMP activities were determined by degradation of a radiolabelled substrate and by zymography. RESULTS: Six months of diabetes was associated with a decrease in mRNA and enzymatic activity of MMP-9 (21 % and 51 % respectively, p < 0.05 vs control) and a 51 % increase in TIMP-1 mRNA ( p < 0.05 vs control). By contrast, MMP-2 mRNA was increased but its activity decreased (43 % and 43 % respectively, p < 0.05 vs control). Total degradative capacity of kidney tissue from diabetic rats was also lower (Control: 48 +/- 7 %, Diabetic: 33 +/- 6 %, p < 0.05). Activation of latent MMPs with amino-phenylmercuric acetate increased matrix degradation by two-fold. However the relative decrease associated with experimental diabetes still remained. All diabetes-associated changes in MMP and TIMP mRNA and activities were attenuated by perindopril treatment in association with reduced type IV collagen accumulation. CONCLUSIONS/INTERPRETATION: These results indicate that the impairment of matrix degradation contributes to matrix accumulation in diabetic nephropathy and that the beneficial effects of ACE inhibition could in part be mediated by modulation of changes in matrix degradative pathways.

Effects of glucose on matrix metalloproteinase and plasmin activities in mesangial cells: possible role in diabetic nephropathy.

Diabetic nephropathy is characterized by an accumulation of mesangium matrix that correlates well with the loss of kidney function. High glucose concentration is known to increase the synthesis of many matrix components. Recently, we have shown that degradation of matrix also decreases in diabetes. The major enzymes responsible for matrix degradation are the matrix metalloproteinases. The physiology of these enzymes is complex and their activity is tightly regulated at many levels. At the transcriptional level matrix metalloproteinase (MMP) expression is increased by protein kinase C (PKC) agonists, and some growth factors. In contrast transforming growth factor (TGF)-beta can decrease MMP expression. Once synthesized, MMPs are secreted as inactive pro-enzymes that are activated by other MMPs or plasmin. To effect this, plasmin must be liberated from plasminogen in the pericellular environment. In turn, activated MMPs can be inhibited by binding to specific inhibitors known as tissue inhibitor of metalloproteinases (TIMP). Cell culture and animal studies have shown that high glucose (HG) decreases expression of MMPs and increases expression of TIMPs. HG can also affect MMP activation by decreasing plasmin availability and reducing expression of a membrane-bound MMP called MT1-MMP. How HG induces these changes remains to be fully elucidated. One possibility is that HG can increase TGF-beta. which may in turn alter MMP promoter activity: this area is currently being studied in our laboratory.

Absolute quantitation of specific mRNAs in cell and tissue samples by comparative PCR.

A comparative PCR assay, for the absolute quantitation of specific mRNAs in cell and tissue samples, has been designed to overcome problems with previous techniques. cDNAs made from the RNAs are co-amplified with "competitor" plasmid templates under conditions in which reagents are not limiting at the equivalence point, thereby preventing competition between target and competitor templates and distinguishing the assay from competitive PCR assays. The cDNAs are serially diluted, and competitor templates concentrations are kept constant, rather than vice versa, as occurs in competitive PCR assays. Products from target and competitor templates are resolved by electrophoresis and measured by phosphorescent or fluorescent imagery. Both products are measured to minimize errors in the competitor:target ratio. A synthetic external standard RNA is included in the tissue lysis solution and co-purified with endogenous mRNAs, thereby being subjected to identical losses of yield during subsequent procedures. The determination of the number of copies of external standard cDNA allows inefficiencies of RNA extraction and cDNA synthesis to be taken into account. Standard concentrations of plasmids containing the endogenous target sequences are also measured, so that corrections can be made for discrepancies due to unequal amplification of target and competitor sequences. These corrections, together with the use of an external standard and the PCR conditions chosen, allow for the accurate, specific and sensitive determination of the absolute number of mRNA copies in a sample.

Epithelial cells up-regulate matrix metalloproteinases in cells within the same mammary carcinoma that have undergone an epithelial-mesenchymal transition.

A metastatic rat mammary carcinoma cell line, BC1, contains cells that have retained epithelial differentiation characteristics and metaplastic cells that have undergone an epithelial-mesenchymal transition. These two subpopulations cooperate to degrade collagen. We have used novel PCR assays to quantitate, for the first time, absolute levels of the mRNAs encoding matrix metalloproteinases (MMPs) and tissue inhibitors of metalloproteinases (TIMPs) in cell and tumor samples. BC1 tumors expressed high levels of the collagenase-3, TIMP-2, stromelysin-1, and gelatinase B genes and low levels of the stromelysin-2 and TIMP-1 genes. This pattern of expression was repeated in cultures of BC1 and cultures containing mixed clones of epithelial cells and metaplastic cells. In both BC1 and the biclonal cultures, metaplastic cells were the main source of collagenase-3, stromelysin-1 and stromelysin-2, whereas TIMPs were equally distributed and epithelial cells were the main source of gelatinase B. High levels of all four MMP mRNAs in metaplastic cells were dependent on coculture with epithelial cells, suggesting the production of an inducing factor by the epithelial cells. In contrast, gelatinase B mRNA was produced at a high level by epithelial cells in the absence of metaplastic cells. TIMP-2 mRNA was abundant in both subpopulations grown alone and did not change substantially upon coculture. Thus, the interclonal cooperativity to degrade collagen in BC1 cells required the induction of MMPs in metaplastic cells by epithelial cells. Interclonal cooperativity may be important to the progression of neoplastic tumors, a feature of which is phenotypic heterogeneity.

Direct activation and anti-repression functions of GAL4-VP16 use distinct molecular mechanisms.

In order to determine whether the molecular mechanisms used for direct activation by GAL4-VP16 are the same as those used for anti-repression, we have employed monoclonal antibodies specific for the VP16 activation domain. In the absence of added repressors, GAL4-VP16 was able to stimulate transcription from a template containing GAL4-binding sites, and the antibodies raised against the VP16 activation domain failed to inhibit this direct activation. GAL4-VP16 also was able to prevent histone H1-mediated repression by a mechanism that was strongly dependent on the presence of specific GAL4-binding elements in the promoter. However, in contrast to the assays conducted in the absence of repressors, the antibodies were strong inhibitors of GAL4-VP16-activated transcription in the presence of histone H1. Thus the binding of the antibodies distinguished between the direct activation and anti-repression functions of GAL4-VP16, indicating that these functions operate through distinct molecular mechanisms. The anti-repression-specific mechanism that is inhibitable by the antibodies acted at an early stage of preinitiation complex formation. Deletions of individual subdomains of the VP16 activation domain demonstrated that there was not a discrete subdomain responsible for the anti-repression function of GAL4-VP16. Thus, the inhibitory effect of the antibodies appeared to be due to the location of the epitope within the activator protein rather than to some inherent biochemical property of that region of the protein that is required specifically for anti-repression. The inhibitory effect of the antibodies also ruled out the possibility that steric exclusion of repressor proteins from the promoter was the sole means of anti-repression by the transcriptional activator.

Anti-integrin antibodies induce type IV collagenase expression in keratinocytes.

During wound healing, pericellular proteolysis is thought to be essential for the detachment of keratinocytes from basement membrane and in their migration into the wound bed. We have characterized integrin-type cell adhesion/migration receptors in human mucosal keratinocytes and examined their function in the regulation of type IV collagenase gene expression. Two major integrins of the beta 1 class, alpha 2 beta 1 and alpha 3 beta 1, were found to function as collagen and fibronectin receptors, respectively. Antibodies against beta 1 and alpha 3 integrin subunits were found to stimulate the expression of the 92 kDa type IV collagenase severalfold in a dose-dependent manner. Keratinocytes expressed also the 72 kDa type IV collagenase, the synthesis of which remained, however, unchanged in keratinocytes treated with anti-integrin antibodies. Stimulation of 92 kDa enzyme was found to be caused directly by antibody binding to integrins, since Fab-fragments of anti-beta 1 antibodies alone were able to induce collagenase expression in the absence of secondary, clustering antibodies. Antibodies against alpha 2 beta 1 integrin caused no stimulation. Keratinocytes seeded on different substrata (plastic, collagen, fibronectin, laminin, or vitronectin) showed equal induction of type IV collagenase expression. Expression of 92 kDa type IV collagenase could not be induced by peptides (GRGDS, GRGES), proteins (fibronectin, laminin, fibrinogen, albumin), or antibodies to fibronectin. We suggest that proteolytic processes around keratinocytes can be regulated by extracellular factors signalling through integrin-type receptors.

Interleukin-1 beta and transforming growth factor-alpha/epidermal growth factor induce expression of M(r) 95,000 type IV collagenase/gelatinase and interstitial fibroblast-type collagenase by rat mucosal keratinocytes.

Rat mucosal keratinocytes serially propagated under permanently serum-free conditions responded to interleukin (IL)-1 beta/IL-alpha and to transforming growth factor (TGF)-alpha/epidermal growth factor (EGF) (as well as to 12-O-tetradecanoylphorbol-13-acetate (TPA)) by upregulation of M(r) 95,000 gelatinase (MMP-9) (M(r) 95K GL) and fibroblast-type collagenase (MMP-1) (FIB-CL), whereas control cells expressed barely detectable levels of either of these enzymes. The cells secreted 8-10 micrograms/10(6) cells/day (M(r) 95K GL) and 2-3 micrograms/10(6) cells/day (FIB-CL) of enzyme protein for at least 24 h when maximally induced. This level was attained only after a 24-h lag period, and the earliest emergence of enzyme protein in the culture medium required 10-14 h. IL-1 beta was by far the most potent cytokine with maximal effect already at 10(-10) M, whereas IL-1 alpha, TGF-alpha, and EGF required 20-100-fold higher concentrations. Pretreatment of the cells with TPA (10(-7) M) abolished the subsequent response to IL-1 beta, TGF-alpha, and EGF and at the same time resulted in > 90% reduction of cytosolic protein kinase C activity. Surprisingly, staurosporine, a potent kinase inhibitor, not only failed to block growth factor/cytokine responses but itself stimulated expression of the enzymes at a magnitude comparable to TPA. The inducing effect of TGF-alpha/EGF was down-regulated by 70-85% by 10(-7) M dexamethasone. Dexamethasone was less effective in ablating the IL-1 beta response yielding 60% reduction M(r) 95K GL and little or no reduction of FIB-CL. Dexamethasone also failed to block the TPA response.

Transforming growth factor-beta 1 up-regulates type IV collagenase expression in cultured human keratinocytes.

During the wound healing process lysis of basement membranes precedes keratinocyte migration into the wound bed. We studied, in vitro, whether this degradation of basement membranes could be regulated by transforming growth factor-beta 1 (TGF-beta 1), which is known to accelerate wound healing in vivo. Transforming growth factor-beta 1 was found to increase the expression of both 92- and 72-kDa type IV collagenases (gelatinases) in cultured human mucosal and dermal keratinocytes. The 92-kDa enzyme predominated in both unstimulated and stimulated cultures. The 92-kDa form was stimulated over 5-fold, and the other form by a factor of 2-3. This increase in the synthesis of type IV collagenases was associated with a marked increase in the mRNA levels of these enzymes as well. The induction of the 92-kDa enzyme was similar in culture medium containing either 0.15 or 1.2 mM calcium chloride. Rat mucosal keratinocytes secreted only 92-kDa type IV collagenase, the secretion of which was not regulated by TGF-beta 1. Also, TGF-beta 1 did not cause any significant induction (maximum about 1.2-fold) of either type IV collagenase in human gingival fibroblasts. The induction levels of both collagenases in human keratinocytes were independent of the type of the extracellular matrix the cells were grown on. However, the basement membrane matrix (Matrigel) activated about half of the 92-kDa type to its 84-kDa active form. The data suggest that TGF-beta 1 has a specific function in up-regulating the expression of type IV collagenases in human keratinocytes, offering a possible explanation of how keratinocytes detach from basement membranes prior to the migration over the wound bed.

Characteristics of a 95-kDa matrix metalloproteinase produced by mammary carcinoma cells.

A Mr 95,000 matrix metalloproteinase (MMP) produced by rat mammary carcinoma cells has been isolated and characterized. The MMP was secreted in a proteolytically inactive form that was free from bound tissue inhibitor of metalloproteinases. The enzyme was highly glycosylated as evident from an apparent drop of Mr from 95,000 to 83,000 after treatment with N-glycanase. Rotary shadowing electron micrographs of purified proenzyme preparations revealed a uniform set of ellipsoidal molecules. Treatment of the proenzyme with 1% SDS resulted in generation of catalytic activity and exposed a cryptic unpaired Cys residue. The latent proenzyme may be activated in at least three additional ways: either spontaneously upon storage, by treatment with organomercurials, or by limited proteolysis by trypsin. Each mode of activation yielded a distinct pattern of cleavage of the enzyme. The activated enzyme cleaved gelatin (denatured type I collagen) and native type IV and V collagen at 30-37 degrees C. Noncollagenous proteins including alpha 1-proteinase inhibitor, casein, and fibrinogen also were cleaved. The rat mammary carcinoma cell line that produces the Mr 95,000 MMP is composed of two distinct (epithelial- and myoepithelial-like) cell types. The enzyme is expressed constitutively by the epithelial cells. This suggests that expression of the Mr 95,000 MMP is regulated differently from that of interstitial collagenase, which is produced by the epithelial cells only in response to specific inductive factor(s) from the myoepithelial-like cells. Monoclonal antibodies raised against the purified latent Mr 95,000 form of the enzyme bind specifically to the Mr 95,000 MMP and have been used to localize the enzyme to the Golgi region and cytoplasmic granules of the epithelial cells.

Immunolocalization of collagenase in neoplastic epithelial cells.

The distribution of interstitial collagenase in a rat mammary carcinoma model system has been studied by immunocytochemistry. Rabbit antibodies were raised against collagenase from neoplastic epithelial cells which were derived from an anaplastic, invasive, rat mammary carcinoma (BC1). Specificity of the antibodies was determined by Western blot analysis which showed reactivity with the inactive procollagenase from conditioned culture medium of BC1 cells as well as with purified, active BC1 collagenase. Anti-BC1 collagenase antibodies did not recognize BC1 collagenase entrapped by the inhibitor, rat alpha-2-macroglobulin (alpha 2M), or collagenase derived from TPA-stimulated human fibroblasts. Anti-human fibroblast collagenase antibodies did not recognize BC1 collagenase, suggesting that the human-mesenchymal and rat-epithelial enzymes are immunologically distinct molecules. Collagenase was immunolocalized intracellularly in BC1 cells cultured in the presence of monensin. Neither BC1 collagenase, alpha 2M nor enzyme-inhibitor complexes were demonstrated in or around invading tumours by immunostaining of tissue sections of rat mammary carcinomas.

Cellular interactions determining the production of collagenase by a rat mammary carcinoma cell line.

The cellular interactions regulating the production of collagenase by a cell line derived from a spontaneously arising rat mammary carcinoma have been studied. The cell line, BC1, was grown permanently under defined serum-free conditions, so that the poorly characterized and variable effects of serum on collagenase expression were avoided. Two stable subpopulations of cells present in BC1 cultures were defined as epithelioid cells ("E-cells") and myoepithelioid cells ("M-cells"). These subpopulations differed in their morphology, pattern of growth and susceptibility to detachment from culture vessels by trypsin. Seven clones of M-cells and 7 clones of E-cells, obtained by the limiting dilution technique, were used to determine the cellular source of collagenase and the interactions which led to its expression. M-cells displayed an absolute dependence on a soluble factor produced by E-cells for their survival in vitro. The presence of both cellular types in culture was necessary for collagenase secretion to occur, E-cells being the major source of enzyme in mixed cultures. A soluble factor produced by M-cells was largely, if not completely, responsible for the induction of collagenase secretion by E-cells. Clones representative of both subpopulations were tumorigenic in syngeneic host animals. These results suggest that the phenotypic diversity which occurs within populations of neoplastic cells may give rise to subpopulations of cells which display a more aggressive phenotype in coexistence than in isolation.

The collagenase produced by neoplastic rat epithelial cells: modulation of secretion, molecular weight characteristics, and purification.

The modulation of the production of collagenase by an epithelial cell line derived from a spontaneously arising rat mammary carcinoma has been studied. The cell line, BC1, was grown permanently under defined serum-free conditions, thus avoiding the poorly characterized and variable effects of serum on collagenase production. Piperazine-N,N'-bis-(2-ethanesulfonic acid) (Pipes), retinoic acid and cytochalasin B all stimulated collagenase secretion, while dexamethasone inhibited it and progesterone, prolactin, prostaglandin E2, and estrogen had no effect. This profile of response to exogenous compounds was distinct from that of cells of mesenchymal origin and from human keratinocytes. For the production of large quantities of collagenase, culture medium was supplemented with Pipes (30 mM, pH 6.8), and retinoic acid (1 microM, on alternate feeds). The collagenase secreted by BC1 cells grown under these conditions was latent and had a molecular mass of 59 kDa. Treatment of the 59 kDa form with trypsin or APMA caused a progressive decrease in molecular mass via 54 kDa and 52 kDa intermediates, to a 48 kDa form. This form was purified to electrophoretic homogeneity by heparin-Sepharose, zinc-chelate-Sepharose, and Sephacryl S-200 chromatography. Five milligrams of purified collagenase were recovered per litre of culture medium.