Raven Press, Ltd., New York ... protease inhibitors and antibodies against tPA and plas- ...... Snow D., Lemmon V., Carrino D., Caplan A., and Silver J. (1990) ...
Journal of Neurochemistry Raven Press, Ltd., New York © 1995 International Society for Neurochemistry
PC 12 Cells Overexpressing Tissue Plasminogen Activator Regenerate Neurites to a Greater Extent and Migrate Faster than Control Cells in Complex Extracellular Matrix Randall N. Pittman and Angela J. DiBenedetto Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania, U.S.A .
Abstract : PC12 cells were stably transfected with expression vectors containing rat tissue plasminogen activator (tPA) under control of either a cytomegalovirus or rous sarcoma virus promoter . Cell lines were characterized using protease assays, ELISAs, immunoblots, northern blots, and Southern blots . Control PC12 cells or cells containing vectors alone released about 1 pg tPA/cell/ 24 h, whereas cells stably transfected with a tPA cDNA released 2-5 pg tPA/cell/24 h . A strong correlation existed between the amount of tPA released and the ability of cells to degrade extracellular matrix . Experiments with protease inhibitors and antibodies against tPA and plasminogen indicated that degradation of matrix involved tPA-generated plasmin and that the amount of matrix degraded was dependent on the amount of tPA released . Cells expressing high levels of tPA migrated on a threedimensional matrix about twice as fast as control cells and regenerated neurites within three-dimensional gels of Matrigel to a greater extent than control cells . Antibodies that inhibited tPA and plasminogen decreased migration and neurite regeneration, indicating that tPA was involved in both events . PC12 cells overexpressing tPA should provide a useful model system for investigating neural functions of tPA including its role in migration and regeneration . Key Words : Neurite outgrowth-Regeneration-Tissue plasminogen activator- Migration -Protease-PC12 cells . J. Neurochem. 64, 566-575 (1995) .
et al ., 1993), and regeneration in the PNS (Bignami et al ., 1982) . The present study was initiated to establish and characterize an in vitro model system that could be used to investigate and test possible functions of tPA in the nervous system . In addition, it was designed to determine if a protease could be overexpressed in neural cells, improve neurite regeneration, and not alter normal cellular functions ; this was viewed as the initial step to determine if proteases may eventually be useful adjuvants for improving regeneration in the nervous system. PC 12 cell lines stably transfected with expression vectors containing rat tPA under control of two different promoters have been established, and data indicate that overexpressing tPA increases migration and neurite regeneration of PC 12 cells in a complex extracellular matrix . MATERIALS AND METHODS Materials Tissue culture plastic was purchased from Corning (100mm dishes) or Costar (24-well dishes) . RPMI, Dulbecco's modified Eagle medium (DMEM), and F12 media were from JRH Biosciences and horse and fetal calf serum were from Hyclone Laboratories . Matrigel and mouse 2 .5S nerve growth factor (NGF) were purchased from Collaborative Research and [ 3H ] collagen and GeneScreen were purchased from New England Nuclear. Spectrozyme PL, DESAFIB fibrin monomer, and Glu-plasminogen were obtained from American Diagnostica . Polyvinylidene difluoride (PVDF) and streptavidin-horseradish peroxidase (streptavidin-
The nervous system contains one of the highest concentrations of tissue plasminogen activator (tPA) in the body (Danglot et al ., 1986) . Although its function in the nervous system is unknown, tPA mRNA is temporally and spatially regulated in the developing nervous system (Sumi et al ., 1992), and neurons express cell surface receptors for tPA (Pittman et al ., 1989 ; Verrall and Seeds, 1989) . It has been suggested that tPA may be involved in a number of events in the nervous system including migration (Moonen et al ., 1982 ; Seeds et al ., 1990), astrocyte differentiation (Kalderon et al ., 1990), neuron-Schwann cell interactions (Clark et al ., 1991), long-term potentiation (Qian
Received February 21, 1994 ; revised manuscript received June 28, 1994 ; accepted June 29, 1994 . Address correspondence and reprint requests to Dr . R . N . Pittman at Department of Pharmacology, University of Pennsylvania School of Medicine, Philadelphia, PA 19104-6084, U .S .A . Abbreviations used: BSA, bovine serum albumin; DMEM, Dulbecco's modified Eagle medium ; G418, Geneticin sulfate ; HRP, horseradish peroxidase; NGF, nerve growth factor ; PAI-1, plasminogen activator inhibitor 1 ; PGB, p-nitrophenyl-p'-guanidinobenzoate ; PVDF, polyvinylidene difluoride ; SDS, sodium dodecyl sulfate ; SSPE, saline-sodium phosphate-EDTA buffer ; tPA, tissue plasminogen activator . 566
PLASMINOGEN ACTIVATOR AUGMENTS REGENERATION HRP) were from Bio-Rad and Nytran was from Schleicher and Schnell. Biotinylated goat anti-rabbit IgG was from Vector, goat anti-mouse IgG-coupled Sepharose was from Zymed, and ELISA plates were from Falcon . Geneticin sulfate (G418) was obtained from GIBCO BRL and dansyl Glu-Gly-Arg-CMK was obtained from Calbiochem . Restriction endonucleases and Lipofectin were purchased from BRL and the pcDNAl-Neo vector was purchased from Invitrogen . All other reagents includingp-nitrophenyl-p'-guanidinobenzoate (PGB) and protein A-Sepharose were obtained from Sigma . Cell culture The PC2 subline of PC 12 cells used in the present study was originally isolated in 1986 in our laboratory . This subline exhibits more migratory behavior and a more rapid response to NGF than the PC12 parent line present in the laboratory . Stably transfected cell lines were routinely cultured in RPMI medium containing 5% fetal calf serum and 10% horse serum, 60 pg/ml penicillin, 100 Mg/ml streptomycin, 5 Mg/ml PGB, and 0.4 mg/ml G418 . PGB was included to inhibit tPA and plasmin; cells secreting high levels of tPA were often lightly attached and more rounded when grown on tissue culture plastic due to local plasmin generation . G418 and PGB were omitted from culture medium used in experiments . In some experiments, cells were grown in serum-free RPMI medium containing 60 Mg /ml penicillin, 100 pg/ml streptomycin, 5 jug/ml insulin, 10 pg/ml transferrin, 30 pM selenium, 20 nM progesterone, 16 pg/ml putrescine, and 100 pg/ml bovine serum albumin (BSA) . To generate two-dimensional substrates, Matrigel was diluted 1 :7 with serum-free RPMI medium and 0.25 ml was added to each well of a 24-well culture dish (2 .2 cm'/well) . Dishes were incubated at room temperature for 2 h and the solution removed and wells rinsed once with serum-free medium before plating cells. To generate three-dimensional gels, Matrigel was diluted 1 :1 with serum-free RPMI medium and 100 pl spread onto 2.2-cm2 wells in a 24-well culture dish . In some cases, [3H]collagen, antibodies, protease inhibitors, and/or cells were present in the serum-free RPMI medium that was mixed with the Matrigel . The mixture was gelled by incubating dishes in a 37°C incubator for 30 min followed immediately by adding serum-containing medium and, in some cases, antibodies, protease inhibitors, and cells. When [3H]collagen was incorporated into Matrigel, gels were washed three times (over a 30-min period) with serum-containing medium before use. Statistical analyses were performed using an unpaired, two-tailed Student's t test . tPA vector constructs pCMV-tPA (Fig . 1) was constructed by digesting pcDNAl-Neo with BamHI and filling in ends to generate a blunt-end pcDNAl-Neo vector . A full-length (2.6 kb) rat tPA cDNA isolated in our laboratory was excised from pGem3Zf(-) using Hind11I and Xhol, filled in to generate blunt ends, and inserted into blunt-ended pcDNAl-Neo. pRSV-tPA (Fig . 1) was generated similarly by digesting pRldn (Conners et al ., 1988) with HindIII, followed by blunt-end ligation of the full-length tPA cDNA into this blunted site . Ligation sites, reading frames, and orientation of tPA cDNA inserts were confirmed by nucleotide sequencing . When a cell line already expresses a low level of an mRNA, it is easier to detect mRNA/protein derived from a
567
transfected cDNA if a truncated cDNA or cDNA from another species is used for transfections. This was not done in the current study for the following two reasons: (1) Normal translational control is often lost when 3'- and 5'-untranslated regions are deleted (Jackson, 1993) ; the 3'-untranslated region of tPA has been shown to be important for translational control and for mRNA stability (Strickland et al ., 1988 ; Huarte et al ., 1992) ; and (2) tPA from other species may not have the full range of biological functions in PC12 cells. Because a primary goal of the present study was to establish a model system to investigate potential roles of tPA in the nervous system, a full-length tPA cDNA including 5'- and 3'-untranslated regions was used for transfections . Generating stable cell lines PC12 cells were transfected with tPA expression vectors or parent vectors by incubating 10 6 cells in 2 ml of serumfree DMEM/F12 medium containing 10 pg of the appropriate plasmid and 20 pg of Lipofectin for 5 h at 37°C . Transfection medium was removed and replaced with plasminogen-free medium containing 7% horse and 7% fetal calf serum (plasminogen was removed from serum using lysineSepharose; Deutsch and Mertz, 1970) . Cells were allowed to recover for 3 days and then subcultured into 4 x 100mm dishes/transfection . Cells were then fed a 1 :1 mixture of conditioned/ fresh medium (plasminogen free) containing 7% horse serum, 7% fetal calf serum, and 0.4 mg/ml G418 . After 2 weeks, the G418 concentration was increased to 0.8 mg/ml. Colonies were picked after 4-6 weeks and expanded . Antibodies Antiserum used for two-site ELISAs, immunoblots, and to inhibit tPA activity was made against rat tPA by immunizing rabbits with 250 pg of recombinant rat tPA (made and purified from Chinese hamster ovary cells stably transfected with a full-length rat tPA cDNA) followed by boosting twice with 100 pg of recombinant rat tPA 6 weeks apart. A similar protocol was used to generate antiserum against plasminogen purified from rat serum according to the method of Deutsch and Mertz (1970) . IgG fractions were purified using protein A-Sepharose . Monoclonal antibody N1-20 was generated by injecting a female BALB/c mouse with 100 pg of recombinant rat tPA followed by two injections of 50 pg of rat tPA 4 weeks apart. The spleen was removed 3.5 days after the third injection and fused with NS 1 myeloma cells using PEG 4000 . Hybridomas were selected in HAT (hypoxanthine, aminopterin, and thymidine) medium . Assays tPA activity. A two-step chromogenic assay was used to determine tPA activity in serum-free medium . The assay was performed in 96-well ELISA plates and consisted of 4 pg of fibrin monomer, 1-40 pl of serum-free conditioned medium (source of released tPA), 200 pM Spectrozyme PL, 0.5 pg Glu-plasminogen in a total volume of 125 pl (buffer: 50 mM Tris, pH 7.4, 0.02% Triton X-100, and 5 mM EDTA) . Samples were incubated 1 .5-2 h at 37 °C and the reaction stopped by adding 15 pl of acetic acid and the absorbance determined at 405 nm . Routinely, three different concentrations of serum-free conditioned medium were tested in duplicate for each sample so that at least one value was in the linear part of the standard curve (3-100 pg/well) . tPA acJ. Neurochem ., Vol. 64, No. 2, 1995
568
R. N. PITTMAN AND A. J. DIBENEDETTO
FIG. 1 . Vectors used to stably transfect PC12 cells. Expression vectors pRSV-tPA and pCMV-tPA were constructed as described in Materials and Methods. Both vectors carry rat tPA cDNAs containing 5' sequences and the entire coding and 3' untranslated sequences under rous sarcoma virus long terminal repeat (RSV pro) or cytomegalovirus (CMV pro) promoter control, respectively . pRSV-tPA also carries the E. coli /3-lactamase gene (AMPr) and both the neomycin resistance gene (NEO) and dihydrofolate reductase gene (DHFR) under ,Q-globin promoter control (betaglopro) . pCMV-tPA carries neomycin resistance under RSV promoter control and confers Sup F phenotype, restoring ampicillin resistance to mutant host strains. Digestion of either expression vector with EcoRl yields a 1 .7-kb fragment from the 3' end of the tPA cDNA ; this same 1 .7-kb fragment excised from a tPA cDNA in pGEM 3Zf(-) was used to probe both northern and Southern blots of transfected cells.
tivity was measured as the difference between total PA activity (tPA + uPA measured in the presence of fibrin monomer) and uPA activity (activity in the absence of fibrin monomer) . Data were converted to protein values by comparing absorbance values to a rat tPA standard curve. uPA activity comprised
