Matrix Vol. 12/1992, pp. 352 - 361 © 1992 by Gustav Fischer Verlag, Stuttgart
Type IV Collagen Synthesis and Accumulation in Neonatal Rat Aortic Smooth Muscle Cell Cultures ANN C. HOSPELHORN 1 , BERNICE M. MARTIN and CARL FRANZBLAU Boston University School of Medicine, Department of Biochemistry, 80 East Concord Street, Boston, MA 02118, USA.
Abstract The production of type IV collagen by cultured neonatal rat aortic smooth muscle cells was monitored over a three-week period to further characterize the extracellular matrix of this unique culture system. Type IV collagen was quantified using a dot immunobinding assay and was found to represent 1 % or less of the total collagen produced by these cells in culture. Total collagen represented up to 33 % of the total protein. The pattern of type IV collagen production in the media and the cell layer suggests that although these cells synthesize and secrete type IV collagen from the onset of culture, type IV collagen deposition only occurs after the cells have reached confluence. In the presence of ascorbate the amount of type IV collagen peaked in the media in preconfluent cultures. In the absence of ascorbate, little type IV collagen was detected in the media. On the other hand, the presence or absence of ascorbate made little differehce in the amount of the total collagen detected in the media, although hydroxylation was affected. Remarkably, in the absence of ascorbate type IV collagen accumulation in the cell layer was similar by the end of the culture period to that in cultures treated with ascorbate. Laminin was not affected by the presence or absence of ascorbate. When these cells were exposed to ascorbate for 24 hours, a peak of soluble elastin was detected in the media. However, soluble elastin was not detected in the media in the absence of ascorbate or in cultures which were maintained in the presence of ascorbate. Modulation of the extracellular matrix with ascorbic acid indicated that type IV collagen deposition did not depend on the presence of ascorbic acid and that there was no discernable interaction between type IV collagen, laminin, and elastin. Key words: ascorbic acid, dot immunobinding, extracellular matrix, smooth muscle cells, type IV collagen.
Introduction Production of extracellular matrix components is a key function of vascular smooth muscle cells, and one that may contribute to the pathology of diseases such as atherosclerosis, diabetes, and hypertension (Barnes, 1985; Chobanian, 1990; Mayne, 1987; Perejda, 1987; Ross, 1986). An important model for the study of extracellular matrix production has been the cultured smooth muscle cell 1 Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy, Boston University, 1990.
(Blumenfeld et aI., 1983; Burke and Ross, 1979). Not all such cells, however, are able to produce large amounts of insoluble elastin, a major component of the extracellular matrix of aortic smooth muscle cells in vivo. The ability of neonatal rat aortic smooth muscle cell cultures to produce an extensive insoluble elastin matrix makes this system an invaluable model for both normal and abnormal extracellular matrix production and assembly. Although elastin production has been fairly well characterized in this system (Barone et aI., 1985; Barone et aI., 1988; Chipman et aI., 1985; Oakes et aI., 1982), the production of other extracellular matrix components such as the collagens and glyco-
Type IV Collagen proteins remain to be examined. This study focuses on the production of type IV collagen by neonatal rat aortic smooth muscle cells in culture. In addition some comparisons are made with laminin and soluble elastin, with the aim of defining mechanisms of extracellular matrix development and possible relationships between the various matrix components. Type IV collagen is a major component of basement membranes, which surround smooth muscle cells in vivo, and may playa critical role in smooth muscle cell function (Herbst et al., 1988; Kleinman et al., 1982; Murray et al., 1979; Paralkar et al., 1991). Because type IV collagen is a quantitatively minor component of the total extracellular matrix of aortic smooth muscle cells (Sankey and Barnes, 1984), a simple and sensitive quantitative assay was required in order to examine the production of this protein over extended periods of time. A dot immunobinding assay was developed for this purpose. By monitoring the appearance and accumulation of type IV collagen from the time of plating to post-confluence and using ascorbic acid as a modulator of extracellular matrix production, we have further characterized the development of the extracellular matrix in this model system.
Materials and Methods Collagen isolation
Types I, III, IV and V collagens were pepsin-extracted from human placenta and separated by salt precipitation from acidic and neutral salt solutions (Kresina and Miller, 1979; Miller and Rhodes, 1982). Type IV collagen was purified further by incubation with DEAE cellulose (Kresina and Miller, 1979). Collagen fractions were monitored by amino acid analysis and SDS polyacrylamide gel electrophoresis (Laemmli, 1970; Miller and Rhodes, 1982). Antibody production and purification
Rabbit anti-human type IV collagen antibodies were generated and purified based on the methods of Furthmayr (1982). Immune sera were assayed for the presence of antibody employing a passive hemagglutination assay, and high titer (> 1: 80 000) antisera were affinity-purified. The IgG fraction of whole serum was eluted from Protein Asepharose CL-4B (Pharmacia, Inc., Uppsala, Sweden) with .58% (v:v) glacial acetic acid containing 0.15 M NaCl (Goding, 1976) and dialyzed agains PBS. This IgG fraction was sequentially passed over a series of extracellular matrix components coupled to CNBr-activated Sepharose 4B (Pharmacia): fibronectin (human plasma), a mixture of types I and III collagens (human placenta), type V collagen (human placenta), and laminin (murine EHS, Collaborative Research, Inc., Bedford, MA). The IgG specific for type IV collagen was then eluted from a type IV collagen affinity
353
column with 1 M acetic acid containing 0.15 M NaCl, neutralized, dialyzed against PBS, and concentrated. Rabbit anti-mouse laminin antibodies were purchased from Collaborative Research. The preparation of rabbit anti-bovine alpha elastin antibodies was described previously (Chipman et al., 1985). These antibodies did not cross-react with types I, III, IV and V collagens or fibronectin; nor did they cross-react with each other. Passive hemagglutination
Fixed human red blood cells were incubated with tannic acid and coated with type IV collagen (5 f.lg!ml). Titer and inhibition assays were carried out as described by Furthmayr (1982). A constant amount of antibody was preincubated with various concentrations of antigen in the inhibition assay. Dot immunobinding assay
The method used in these studies was a composite of several procedures (Hawkes et al., 1982; ]ahn et al., 1984; Towbin and Gordon, 1984). Grids (1 cm 2 ) were drawn on nitrocellulose with a pencil. The nitrocellulose was soaked for 5 -10 min in distilled water and dried completely before protein application. Protein was applied to nitrocellulose in 1-3 f.ll aliquots. In some cases successive applications of protein were made to the same area, allowing the nitrocellulose to dry completely between each application. If samples contained little protein, bovine serum albumin (1 mg! ml, IgG-free) was added as a carrier. Blots were incubated for 15 min in 10% (voVvol) acetic acidl25% (voVvol) isopropyl alcohol to fix proteins to the nitrocellulose, rinsed several times with distilled water, and equilibrated in TrisTween buffer for 15 min (0.15 M NaCl, 0.05 M Tris, 0.2% Tween, pH 7.2). This was followed by incubation for 1 h with blocking buffer (Tris-Tween containing 3% nonfat dry milk). The appropriate antibodies were added to incubation buffer (Tris-Tween containing 1% nonfat dry milk): 1 f.lg!ml anti-type IV collagen IgG, 5 f.lg!ml anti-laminin IgG, or 1/200 dilution of anti-elastin serum. Blots were incubated in the appropriate antibody solution for 1 h followed by three 10-min washes in Tris-Tween buffer. [125 1]_ protein A was added to incubation buffer (500,000 cpm! ml, 70-100 f.lCilf.lg), and blots were incubated in this solution for 1 h. This was followed by a quick rinse in TrisTween buffer, four quick rinses in distilled water, and four 20-min washes in distilled water. Blots were dried at room temperature. Blocking, antibody incubation, and 25 1]_ protein A incubations were done at 37°C. All other steps were done at room temperature. Kodak XAR film was preflashed and exposed to the blots for 4 to 48 h at - 80°C in cassettes containing enhancing screens. Quantification was based on densitometry of the exposed film. Dots were scanned using a Shimadzu Dual-
e
354
A. C. Hospelhorn et ai.
Wavelength Thin-Layer Chromato Scanner Model CS-930 densitometer. Two-fold dilutions (1600-6 pg) of standard proteins were included on each blot and were used to generate a standard curve. Two or three concentrations of each sample, and duplicate or triplicate samples were blotted. The amount of a specific protein in each sample was determined from the standard curve generated under the same conditions, i. e. standard antigen diluted in media or acetic acid (cell layer), and all values for each time point were averaged. When necessary, multiple exposures of the same blot were scanned to obtain values on the linear portion of the standard curve.
Smooth muscle cell culture Neonatal rat smooth muscle cells were isolated from aortae of 3-day-old rats (Oakes et aI., 1982). First passage cells were used for all experiments. Cells were seeded in 35mm dishes (Falcon) at 1.5 x 10 4 cells/cm 2 and maintained in 2 ml of media containing 10% fetal bovine serum. Media were changed twice a week. Cells were cultured up to 20 days, with groups of cells harvested every other day starting with day 2. Twenty-four hours prior to each harvest, cells were washed (2 x) with Puck's saline and changed to 1 ml of the appropriate incubation media with or without serum and with or without ascorbate (50 f.A.g/ml). Cells grown continuously in the presence of ascorbate received ascorbate (20 f.A.g/ml) the day after plating, with each media change (twice a week), and 24 h prior to harvest (designated continuous ascorbate). When ascorbate was added only during the 24-h test period, this was denoted as short-term ascorbate treatment. DNA was analyzed using diphenylamine (Burton, 1956; Patterson, 1979). Cell counts were obtained using a Coulter counter after cells were released from culture dishes with a trypsin-collagenase solution (3 mg trypsin (Type III, Sigma Chemical Co., St. Louis, MO)-10mg collagenase (Type lA, Sigma)/12ml Dulbeccos's modified Eagle's medium) and diluted with Isoton (Coulter Electronics, Inc., Hialeah, FL).
Analysis ofcell cultures At each harvest the media were transferred to microfuge tubes and centrifuged at 16000 x g for 30 sec to remove cellular debris. Some media samples were dialyzed against distilled water, and total protein and hydroxyproline were determined by amino acid analysis. Other aliquots of media were applied directly to nitrocellulose after the addition of carrier BSA and analyzed for specific matrix proteins by dot immunobinding assays. Cell layers were washed twice with Puck's saline G, and the washes were discarded. Some cell layers were analyzed for DNA and total protein. Total protein was determined by amino acid analysis of the residue remaining after DNA extraction (Patterson, 1979). Other cell layers were incu-
bated with 1 ml of cold 0.5 M acetic acid containing 100 f.A.g of pepsin (Sigma 1:60000) for 20 hat 4°C on a shaker to release matrix proteins. The supernatant was transferred to a microfuge tube and mixed with 50 f.A.I of a pepstatin A (Sigma) solution (100 f.A.g/ml 0.5 M acetic acid). Aliquots were applied directly to nitrocellulose after the addition of carrier BSA and assayed by the dot immunobinding procedure.
Collagenase assay To assess collagenase susceptible protein, radiolabeled proline (10 f.A.Ci/ml, L-[2,3,4,5- 3H] proline, 111 Ci/mmol) was added to the cell culture media for 24 h. Radiolabeled proteins from cell cultures were then incubated with bacterial collagenase (Form III, Advanced Biofactures) (Peterkofsky et aI., 1982; Freiberger et aI., 1980). Media and sonicated cell layer samples were dialyzed extensively against 1 mM proline followed by collagenase buffer [0.15 M NaCl, 0.05 M Tris, and 0.005 M CaCh, pH 7.5 (Sage et aI., 1979)] containing 2.5 mM n-ethylmaleimide. Duplicate 200 f.A.I control and digest samples were aliquoted. Bovine serum albumin was added as carrier (200 f.A.I, 0.5 mg/200 f.A.I), and 5 BTC units of collagenase were added to digest samples. Samples were brought to 500 f.A.I with buffer and agitated at 37°C for 20 h. Proteins were precipitated with an equal volume of cold 10% trichloroacetic acid (TCA), 0.5% tannic acid and centrifuged at 16000 x g for 10 min in a microfuge. The supernatant was transferred to another tube, and the pellet was washed with 5% TCA, 0.25% tannic acid and centrifuged. The wash was combined with the supernatant, and an aliquot was added to scintillation fluid and counted. The pellets of the digested samples were hydrolyzed in 6 N HCI, and an aliquot of the hydrolysate was added to scintillation fluid and counted. Percent collagen was calculated (Peterkofsky et aI., 1982): _ _ _ _ _ _ _d~p~m_in_c_o_ll~ag~e_n (dpm in noncollagen protein x 5.4) + dpm in collagen
x 100
Isolation oftype IV collagen from cell culture medium and determination ofhydroxylation Cells were seeded in 100-mm dishes at 1.5 x 10 4 cells/ cm 2 and maintained in 15 ml of media containing 10% fetal bovine serum. On day 3 the media were aspirated. The cell layers were washed twice with serum-, proline- and lysinefree media and then were incubated with 5 ml of media containing [3H]-proline (20 f.A.Ci/ml, L-[2,3,4,5-3H] proline, 111 Ci/mmol) and [3H]-lysine (20 f.A.Ci/ml, L-[4,5- 3H]lysine, 76 Ci/mmol) with and without 50 f.A.g/ml ascorbate for 24 h. Media from three dishes were pooled, and proteolytic inhibitors were added to a final concentration of 2.5 mM EDTA, 2 mM phenylmethylsulfonyl fluoride, and 10 mM n-ethylmaleimide. The media were stored at - 20°C. After
Type IV Collagen thawing, 2 ml of media were incubated with 20 Ilg of affinity-purified type IV collagen antibody for 2 h at room temperature with gentle mixing. This solution was then loaded on a protein A-sepharose CL-4B column which had been washed with PBS containing 1 mM proline and 1 mM lysine. After loading, the column was eluted with 0.15 M NaCl, 0.58% (v:v) acetic acid. After dialysis against distilled water, the eluted material was lyophilized and hydrolyzed for amino acid analysis.
355
equally well with native and heat-denatured type IV collagen. Types I, III and V collagens, as well as fibronectin, showed no reaction. Although there was a consistent slight reaction with EHS laminin, incubation of this preparation with bacterial collagenase eliminated the reaction. Since antibodies to EHS laminin reacted equally well with control and collagenase-treated laminin samples, the reactivity of laminin samples with anti-type IV collagen IgG appeared to be caused by collagen contamination of the laminin preparation. These antibodies specifically labeled basement membranes in human lung tissue (not shown). Reaction of pepsin-solubilized human type IV collagen with affinity-purified rabbit IgG against pepsin-solubilized human type IV collagen in the dot immunobinding assay showed a linear dose response (Fig. 1).
Results Antibody specificity and dot immunobinding assay
The specificity of type IV collagen antibodies was determined by passive hemagglutination inhibition and dot immunobinding. Type IV collagen antibodies reacted
A
IN IV 6
13
25
50
100
200 400 800 1600 (pg)
B lil 045
«l
i'-
a:
t
:Ii :II
•
25
+'
a
CI 0=
2D 16
Fig. 1. Dot immunobinding assay. A) Autoradiogram of two-fold dilutions of laminin and type IV collagen incubated with antitype IV collagen IgG followed by [125 1]_ protein A. B) Plot shows linear portion of the type IV collagen standard curve. Relative areas of dots were obtained by scanning densitometry.
10 6 0 0
( 25
50
76
100
T~
126
160
176 200
22fi
IV Collogl!ln (plcograne)
260
276
:Ill
356
« z
c
A. C. Hospelhorn et al. 30
Cl
25
c
20
:::s
-y----------------------, A
CIl
Cl
~
o
15
>
10
()
CIl C.
>-
~
Cl
C
o
10
« z
2
4
6
2
4
6
8
10
12
14
16
18
20
14
16
18
20
Days In Culture
B
8
C
Cl
:::s
6
c c E ('Q
4
Cl
2
..J
c
0 0
8
10
12
Days In Culture
10
« z c
C
8
Cl
:::s
w
6
c
Cfj
~
Cl
c
4
2
0 0
2
4
6
8
10
12
Days In Culture
14
16
18
20
Fig. 2. Effect of no ascorbate (_), shortterm ascorbate (0), continuous ascorbate (0) and serum + short-term ascorbate (~) on A) type IV collagen, B) laminin and C) soluble elastin detected in the media. See text for details. The amount of laminin detected in the absence of ascorbate was essentially the same as that for each of the ascorbate conditions and is not shown.
TypeIV Collagen
Proliferation ofneonatal rat aortic smooth muscle cells in culture The proliferation of first passage neonatal rat aortic smooth muscle cells was monitored by cell counts, DNA content, and phase-contrast microscopy. There was little difference in cell number or DNA content between control and treated cultures. The most rapid proliferation occurred between days 2 and 4, and phase-contrast microscopy demonstrated that cells were confluent by day 4.
Type IV collagen detected in the media Neonatal rat aortic smooth muscle cell cultures produced more type IV collagen in the presence of ascorbate than in its absence. All cell cultures incubated in the presence of ascorbate, regardless of the time of exposure to ascorbate (short-term vs continuous) or the presence or absence of serum during the 24-h incubation period, exhibited a similar pattern of type IV collagen accumulation in the media (Fig.2A). Type IV collagen was detected in the media as early as day 2. The greatest amounts were secreted into the media prior to detection of type IV collagen in the cell layer (Fig. 3). Although the pattern of type IV collagen production was similar under the conditions examined, the data suggest that type IV collagen production was influenced by the culture conditions during the early phase ~f culture. The presence of serum in conjunction with ascorbate during the 24-h incubation period appeared to stimulate type IV collagen production. Continuous exposure to ascorbate, on the other hand, appeared to attenuate type IV collagen production. Although cultures from different primary isolates were used for the "continuous ascorbate" and "short-term ascorbate" experiments, the difference in the amount of type IV collagen found during the early phase of culture versus later time points was greatly diminished in cultures continuously exposed to ascorbate. In the absence of ascorbate, only a small amount of type IV collagen could be detected in the media after the first week in culture. As expected, in the absence of ascorbate, type IV collagen isolated from the media was found to be underhydroxylated compared with that isolated from media of cultures exposed to ascorbate (25% of the total proline was hydroxylated in the absence of ascorbate, 57% in the presence of ascorbate).
357
when the cells were incubated with ascorbate for only 24 h (Fig. 3A) and when cells were maintained in the presence of ascorbate (Fig. 3B).
Total collagen detected in the media Fig. 4 compares percent collagenase-susceptible protein (4A) to total protein (4B) detected in the media in the presence and absence of ascorbate. In each 24-h period, approximately 90% of the counts in collagen were associated with the media fraction. Based on these data the amount of type IV collagen represented less than 1% of the total collagen. Overall, the presence or absence of ascorbate made little difference in the amount of total collagen detected in the media, although preconfluent cell cultures produced slightly more collagen in the presence of ascorbate than in its absence. Hydroxyproline could be detected in the media in the presence of ascorbate but not in the absence of ascorbate, indicating that the collagen produced in the absence of ascorbate was underhydroxylated (data not shown). Compared with total collagen, type IV collagen accumulation in the media appeared to be more affected by the presence or absence of ascorbate (Fig. 2A).
Total collagen detected in the cell layer Again, somewhat surprisingly, the presence or absence of ascorbate did not appear to affect the amount of collagen detected in the cell layer during the 24-h period examined (data not shown). The level of hydroxylation in the cell layer remains to be determined. Both type IV collagen and total collagen were detected in the cell layer only after the cells had reached confluence.
Laminin detected in the media and cell layer Laminin was detected in the media by day 4 (Fig. 2B). In contrast to type IV collagen production, laminin production was not affected by the presence or absence of ascorbate. Laminin levels remained relatively constant throughout the culture period, although additional exposure to serum 24 h prior to harvest slightly enhanced laminin production on days 4-8. Small but increasing amounts of laminin were detected in the cell layer from day 12 (0.44 ng! !!g DNA) to day 20 (5.09 ng!!!g DNA) in both the presence and absence of ascorbate.
Type IV collagen detected in the cell layer In the presence of ascorbate detectable amounts of type IV collagen began to accumulate in the cell layer only after the cells reached confluence on day 4 (Fig. 3). Surprisingly, in the absence of ascorbate type IV collagen was also incorporated into the cell layer such that there was little difference between the untreated and ascorbate-treated cell cultures by the end of the culture period. This occurred both
Soluble elastin detected in the media The presence of ascorbate and whether the ascorbate exposure was short-term or long-term had a dramatic effect on the accumulation of soluble elastin and/or its fragments in the media of these cell cultures (Fig. 2C). Soluble elastin was not detected in the media in the absence of ascorbate or in cultures which were maintained in the presence of ascor-
A. C. Hospelhorn et al.
358 6
« z
C
A 5
C)
:::J
c
4
(I) C)
,!g
'0
3
0
~ (I)
2
Co
>-
I-
C)
c
O-+---III----1.._-. .-
o
2
4
6
__:..----,r---'T"""'"--r--.,.---r---1
8
10
12
14
16
18
20
Days In Culture
6
« z C
B 5
C)
:::J
c
4
(I) C)
,!g
'0
3
0
~
2
(I)
Co
>-
I-
C)
c
o
2
4
6
8
10
12
14
16
18
20
Days In Culture
bate. However, when these celle were exposed to ascorbate for 24 h, a peak of soluble elastin was readily detected in the media from day 6 to day 10. This was the case whether or not serum was present during the incubation period. In contrast to the effect of serum on both type IV collagen and laminin, the amount of soluble elastin detected in the media after an additional 24-h exposure to serum was considerably less than in the absence of serum.
Fig. 3. Effect of no ascorbate (_), shortterm ascorbate (0) and continuous ascorbate (0) on the accumulation of type IV collagen in the cell layer of cultures derived from two different primary isolates, A and B. See text for details.
Discussion The expression, accumulation (synthesis and degradation), organization, and remodeling of the extracellular matrix are extremely complex processes which can be influenced by cell culture conditions, known modulators (ascorbic acid, lathrogens), growth factors, cell-matrix interactions, and feedback mechanisms (Blumenfeld et al., 1983; Mecham, 1986; Streuli and Bissell, 1990). We have examined the long-term production of type IV collagen in
Type IV Collagen 40
359
A
30
c
CI>
0)
cu
'0
-
0
20
c
CI>
() ~
CI>
Q.
10
0+---r---,r----r--"""'T"---,r----r--"""'T"-_,.--~-_1
o
2
4
6
8
10
12
14
16
18
20
Days In Culture
50 " T " " - - - - - - - - - - - - - - - - - - - - - - . ,
B ~
.~
c
40
E
E It) CO)
c
'cu o
30
20
~
Q.
0)
:s
Fig. 4. Effect of no ascorbate (.) and shortterm ascorbate (0) on A) percent collagen of total protein detected in the media (collagenase assay), B) amount of protein detected in the media (amino acid analysis). See text for details.
10
O+---r---,r----r--"""'T"---,--~-_._-_,.--~-~
o
2
the media and cell layer compartments of neonatal rat aortic smooth muscle cells using ascorbic acid as a modulator of extracellular matrix. As long as ascorbate is present during the test period, type IV collagen in the media peaks within the first four days and quickly falls off to lower levels. This early rise coincides with the rapid proliferation of these cells in culture. In the absence of ascorbate, no type IV collagen is immunologically detected in the media for most of the culture period. Accumulation of type IV collagen in the cell
4
6
8
10
12
14
16
18
20
Days In Culture
layer is detected after the peak seen in the media and after the cells become confluent. It is remarkable that in the absence of ascorbate type IV collagen accumulation in the cell layer is only slightly less by the end of the culture period than in those cultures treated with ascorbate. The data suggest that some event occurs soon after confluence that greatly enhances type IV collagen deposition. Integrin receptors on smooth muscle cells have recently been described for type IV collagen (Clyman et aI., 1990a; Clyman et aI., 1990b; Syfrig et aI., 1991). These receptors
360
A. C. Hospelhorn et al.
may be induced by cell-cell interactions and aid in matrix assembly. Okada et al. (1990) suggest that cell-cell interactions may be responsible for the observed modulation of smooth muscle cells in long-term cultures from a synthetic state back to a contractile state with increased synthesis of type IV collagen and the ultrastructural appearance of basement membrane. It is also possible that deposition of other components, such as proteoglycans (Hamati et al., 1989), is required prior to type IV collagen deposition. There is little information available on the deposition of type IV collagen in the absence of ascorbate. One report (Katsuda et al., 1987) suggests that type IV collagen can be deposited in the absence of ascorbate. On the other hand, in two other cell systems, no deposition of type IV collagen was seen in the absence of ascorbate (Eldridge et al., 1987; Frenette et al., 1989). It is interesting to note that in a Schwann cell system, ascorbic acid was required to detect extracellular type IV collagen as well as to detect organization of laminin and heparan sulfate proteoglycan (Eldridge et al., 1987). It remains to be determined in the neonatal rat system whether the organization of type IV collagen, as well as other components, is altered in the absence of ascorbate and whether type IV collagen associated with the cell layer in the absence of ascorbate is underhydroxylated. The presence or absence of ascorbic acid has a dramatic influence on the levels of insoluble elastin in these cultures. In its absence elastin is a major component of the extracellular matrix; whereas, in the presence of ascorbic acid, insoluble elastin decreases substantially (Barone et al., 1985). In the present studies, the amount of type IV collagen associated with the cell layer at the end of the culture period was the same in the presence or absence of ascorbate suggesting that elastin deposition and type IV collagen deposition are independent. It is of interest to note the differences in soluble elastin detection in the media during the short-term and continuous ascorbate conditions. The insoluble matrix under the short-term ascorbate conditions would essentially be the same as that in the absence of ascorbate since cells have been maintained in the absence of ascorbate and would only see ascorbate for 24 h. That is, insoluble elastin would be a major component of the extracellular matrix. In the absence of ascorbate soluble elastin is efficiently incorporated into an insoluble matrix, and no soluble elastin is detected in the media. However, when cells are exposed to ascorbate for 24 h, soluble elastin is detected in the media, probably representing overhydroxylated elastin (Barone et al., 1985), and a window of peak elastin production is observed. Under continuous ascorbate conditions, no soluble elastin is detected in the media, possibly representing down-regulation of elastin production. In this regard, the lathrogen beta-aminopropionitrile (BAPN) inhibits crosslinking of elastin in the cell layer and results in increased levels of soluble elastin in the media of these cultures (Chipman et al., 1985). Recently, it has been shown in our
laboratory that incubation of these cells with BAPN down regulates elastin mRNA Uackson et al., 1991), most likely because of soluble elastin fragments feeding back on the system. The generation of soluble elastin by ascorbate may cause a similar effect. Because cell-matrix interactions are important in determining cell behavior (Kleinman et al., 1982), it is important to determine what matrix components are present under various conditions and ultimately to define differences in the cell's response to its environment. The accumulation of type IV collagen in long-term culture in both the presence and absence of ascorbate may be indicative of both normal and abnormal matrix development. This system should be useful in probing smooth muscle cell function and how it is affected by the extracellular matrix.
Acknowledgements The authors gratefully acknowledge the expert technical assistance of George Crombie, Rosemarie Moscaritolo and Valerie Verbitzki. We would also like to acknowledge the valuable suggestions of Christina Hospelhorn (Ben May Institute, University of Chicago) concerning the dot immunobinding assay. This study was supported by NIH Grants HL 07429, HL 13262, HL 19717 and HL 31453.
References Barnes, M.J.: Collagens in atherosclerosis. Collagen ReI. Res. 5: 65-97,1985. Barone, L.M., Faris, B., Chipman, S.D., Toselli, P., Oakes, B. W. and Franzblau, c.: Alteration of the extracellular matrix of smooth muscle cells by ascorbate treatment. Biochim. Biophys. Acta 840: 245-254, 1985. Barone, L.M., Wolfe, L., Faris, B. and Franzblau, C.: Elastin mRNA levels and insoluble elastin accumulation in neonatal rat smooth muscle cell cultures. Biochemistry 27: 3175-3182, 1988. Blumenfeld, 0.0., Bienkowski, R.S., Schwartz, E. and Seifter, S.: Extracellular matrix of smooth muscle cells. In: Biochemistry of Smooth Muscle, ed. by Stephens, N.L., CRC Press, Inc., Boca Raton,FL,1983,pp.137-187. Burke, J. M. and Ross, R.: Synthesis of connective tissue macromolecules by smooth muscle. Int. Rev. Connective Tissue Res. 8: 119-157,1979. Burton, K.: A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem.]. 62: 315 - 323, 1956. Chipman, S. D., Faris, B., Barone, L. M., Pratt, C. A. and Franzblau, c.: Processing of soluble elastin in cultured neonatal rat smooth muscle cells.]. BioI. Chem. 260: 12780-12785,1985. Chobanian, A. V.: Adaptive and maladaptive responses of the arterial wall to hypertension. Hypertension 15: 666-674, 1990. Clyman, R. I., McDonald, K. A. and Kramer, R. H.: Integrin receptors on aortic smooth muscle cells mediate adhesion to fibronectin, laminin, and collagen. Circulation Research 67: 175-186,1990a. Clyman, R.I., Turner, D.C. and Kramer, R.H.: An al/~l-like
Type IV Collagen integrin receptor on rat aortic smooth muscle cells mediates adhesion to laminin and collagen types I and IV. Arteriosclerosis 10: 402-409, 1990b. Eldridge, CF., Bartlett Bunge, M., Bunge, R.P. and Wood, P.M.: Differentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. ]. Cell BioI. 105: 1023-1034,1987. Freiberger, H., Grove, D., Sivarajah, A. and Pinnell, S. R.: Procollagen I synthesis in human skin fibroblasts. Effect of culture conditions on biosynthesis.]. Invest. Dermatol. 75: 425 -430, 1980. Frenette, G.P., Ruddon, R.W., Krzesicki, R.F., Naser, J.A. and Peters, B. P.: Biosynthesis and deposition of a noncovalent laminin-heparan sulfate proteoglycan complex and other basal lamina components by a human malignant cell line. ]. BioI. Chem. 264: 3078-3088, 1989. Furthmayr, H.: Immunization procedures, isolation by affinity chromatography, and serological and immunochemical characterization of collagen specific antibodies. In: Immunochemistry of the Extracellular Matrix, Vol. 1, ed. by Furthmayr, H., CRC Press, Inc., Boca Raton, FL, 1982, pp. 143-178. Goding, ]. W.: Conjugation of antibodies with fluorochromes. Modifications to the standard methods. ]. Immunol. Methods 13: 215-226, 1976. Hamati, H. F., Britton, E. L. and Carey, D.J.: Inhibition of proteoglycan synthesis alters extracellular matrix deposition, proliferation, and cytoskeletal organization of rat aortic smooth muscle cells in culture.]. Cell BioI. 108: 2495-2505, 1989. Hawkes, R., Niday, E. and Gordan, ].: A dot-immunobinding assay for monoclonal and other antibodies. Anal. Biochem. 119: 142-147,1982. Herbst, T.J., McCarthy, ]. B., Tsilibary, E. C and Furcht, L. T.: Differential effects of laminin, intact type IV collagen, and specific domains of type IV collagen on endothelial cell adhesion and migration.]' Cell Bioi. 106: 1365 -1373,1988. Jackson, L. E., Faris, B., Martin, B. M., Jones, H. V., Rich, C B., Foster,]. A. and Franzblau, C: The effect of ~-aminopropionit rile on elastin gene expression in smooth muscle cell cultures. Biochem. Biophys. Res. Commun. 179: 939-944, 1991. Jahn, R., Schiebler, W. and Greengard, P.: A quantitative dotimmunobinding assay for proteins using nitrocellulose membrane filters. Proc. Natl. Acad. Sci. USA 81: 1684-1687,1984. Katsuda, S., Okada, Y., Minamoto, T. and Nakanishi, I.: Enhanced synthesis of type IV collagen in cultured arterial smooth muscle cells associated with phenotypic modulation by dimethyl sulfoxide. Cell Bioi. Int. Rep. 11: 861-870, 1987. Kleinman, H. K., Rohrbach, D. H., Terranova, V. P., Varner, H. H., Hewitt, A. T., Grotendorst, G. R., Wilkes, C M., Martin, G. R., Seppa, H. and Schiffmann, E.: Collagenous matrices as determinants of cell function. In: Immunochemistry of the Extracellular Matrix, Vol. II, ed. by Furthmayr, H., CRC Press, Boca Raton, FL, 1982, pp. 151-174. Kresina, T.F. and Miller, E.].: Isolation and characterization of basement membrane collagen from human placental tissue. Evidence for the presence of two genetically distinct collagen chains. Biochemistry 18: 3089-3097, 1979. Laemmli, U. K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685, 1970. Mayne, R.: Vascular connective tissue. Normal biology and derangement in human diseases. In: Connective Tissue Disease.
361
Molecular Pathology of the Extracellular Matrix, ed. by Uitto, ]. and Perejda, A.J., Marcel Dekker, Inc., New York, 1987, pp.163-183. Mecham, R.: Regulation of Matrix Accumulation. Academic Press, Inc., 1986. Miller, E.]. and Rhodes, R.K.: Preparation and characterization of the different types of collagen. In: Methods In Enzymology, Vol. 82, ed. by Cunningham, L. W. and Frederiksen, D. W., Academic Press, New York, 1982, pp. 33 -64. Murray,]. C, Stingl, G., Kleinman, H., Martin, G. and Katz, S.: Epidermal cells adhere preferentially to type IV (basement membrane) collagen.]. Cell Bioi. 80: 197- 202,1979. Oakes, B. W., Batty, A. C, Handley, C]. and Sandberg, L. B.: The synthesis of elastin, collagen and glycosaminoglycans by high density primary cultures of neonatal rat aortic smooth muscle. An ultrastructural and biochemical study. Eur.]. Cell Bioi. 27: 34-46,1982. Okada, Y., Katsuda, S., Matsui, Y., Watanabe, H. and Nakanishi, I.: Collagen synthesis by cultured arterial smooth muscle cells during spontaneous phenotypic modulation. Acta Pathol. Jpn. 40: 157-164,1990. Paralkar, V.M., Vukicevic, S. and Reddi, A.H.: Transforming growth factor beta type 1 binds to collagen IV of basement membrane matrix: implications for development. Dev. BioI. 143: 303-308, 1991. Patterson, M. K., Jr.: Measurement of growth and viability of cells in culture. In: Methods In Enzymology, Vol. 58, ed. by Jakoby, W.B. and Pastan, I.H., Academic Press, New York, 1979, pp.141-152. Perejda, A.].: Vascular and cutaneous complications of diabetes mellitus. The role of nonenzymatic glycosylatoin of collagens. In: Connective Tissue Disease. Molecular Pathology of The Extracellular Matrix, ed. by Uitto,]. and Perejda, A.]., Marcel Dekker, Inc., New York, 1987, pp. 475-490. Peterkofsky, B., Chojkier, M. and Bateman, J.: Determination of collagen synthesis in tissue and cell culture systems. In: Immunochemistry of the Extracellular Matrix, Vol. 1, ed. by Furthmayr, H., CRC Press, Inc., Boca Raton, FL, 1982, pp.19-47. Ross, R.: The pathogenesis of atherosclerosis-an update. New Engl.]. Med. 314: 488-499,1986. Sage, H., Woodbury, R. G. and Bornstein, P.: Structural studies on human type IV collagen. ]. Bioi. Chem. 254: 9893-9900, 1979. Sankey, E.A. and Barnes, M.J.: Comparison of the collagenous products synthesized in culture by pig aortic endothelial and smooth-muscle cells. Biochem.]. 218: 11-18,1984. Streuli, C H. and Bissell, M.J.: Expression of extracellular matrix components is regulated by substratum. ]. Cell. Bioi. 110: 1405-1415,1990. Syfrig, J., Mann, K. and Paulsson, M.: An abundant chick gizzard integrin is the avian al~l integrin heterodimer and functions as a divalent cation-dependent collagen IV receptor. Exp. Cell Res. 194: 165-173, 1991. Towbin, H. and Gordon, J.: Immunoblotting and dot immunobinding-current status and outlook. ]. Immunol. Methods 72: 313-340,1984.
Dr. Ann C. Hospelhorn, Surgery Research Unit - 5 West, Massachusetts General Hospital, 149 13th Street, Charlestown, MA 02129, USA.
Type IV Collagen Synthesis and Accumulation in Neonatal Rat Aortic Smooth Muscle Cell Cultures ANN C. HOSPELHORN 1 , BERNICE M. MARTIN and CARL FRANZBLAU Boston University School of Medicine, Department of Biochemistry, 80 East Concord Street, Boston, MA 02118, USA.
Abstract The production of type IV collagen by cultured neonatal rat aortic smooth muscle cells was monitored over a three-week period to further characterize the extracellular matrix of this unique culture system. Type IV collagen was quantified using a dot immunobinding assay and was found to represent 1 % or less of the total collagen produced by these cells in culture. Total collagen represented up to 33 % of the total protein. The pattern of type IV collagen production in the media and the cell layer suggests that although these cells synthesize and secrete type IV collagen from the onset of culture, type IV collagen deposition only occurs after the cells have reached confluence. In the presence of ascorbate the amount of type IV collagen peaked in the media in preconfluent cultures. In the absence of ascorbate, little type IV collagen was detected in the media. On the other hand, the presence or absence of ascorbate made little differehce in the amount of the total collagen detected in the media, although hydroxylation was affected. Remarkably, in the absence of ascorbate type IV collagen accumulation in the cell layer was similar by the end of the culture period to that in cultures treated with ascorbate. Laminin was not affected by the presence or absence of ascorbate. When these cells were exposed to ascorbate for 24 hours, a peak of soluble elastin was detected in the media. However, soluble elastin was not detected in the media in the absence of ascorbate or in cultures which were maintained in the presence of ascorbate. Modulation of the extracellular matrix with ascorbic acid indicated that type IV collagen deposition did not depend on the presence of ascorbic acid and that there was no discernable interaction between type IV collagen, laminin, and elastin. Key words: ascorbic acid, dot immunobinding, extracellular matrix, smooth muscle cells, type IV collagen.
Introduction Production of extracellular matrix components is a key function of vascular smooth muscle cells, and one that may contribute to the pathology of diseases such as atherosclerosis, diabetes, and hypertension (Barnes, 1985; Chobanian, 1990; Mayne, 1987; Perejda, 1987; Ross, 1986). An important model for the study of extracellular matrix production has been the cultured smooth muscle cell 1 Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy, Boston University, 1990.
(Blumenfeld et aI., 1983; Burke and Ross, 1979). Not all such cells, however, are able to produce large amounts of insoluble elastin, a major component of the extracellular matrix of aortic smooth muscle cells in vivo. The ability of neonatal rat aortic smooth muscle cell cultures to produce an extensive insoluble elastin matrix makes this system an invaluable model for both normal and abnormal extracellular matrix production and assembly. Although elastin production has been fairly well characterized in this system (Barone et aI., 1985; Barone et aI., 1988; Chipman et aI., 1985; Oakes et aI., 1982), the production of other extracellular matrix components such as the collagens and glyco-
Type IV Collagen proteins remain to be examined. This study focuses on the production of type IV collagen by neonatal rat aortic smooth muscle cells in culture. In addition some comparisons are made with laminin and soluble elastin, with the aim of defining mechanisms of extracellular matrix development and possible relationships between the various matrix components. Type IV collagen is a major component of basement membranes, which surround smooth muscle cells in vivo, and may playa critical role in smooth muscle cell function (Herbst et al., 1988; Kleinman et al., 1982; Murray et al., 1979; Paralkar et al., 1991). Because type IV collagen is a quantitatively minor component of the total extracellular matrix of aortic smooth muscle cells (Sankey and Barnes, 1984), a simple and sensitive quantitative assay was required in order to examine the production of this protein over extended periods of time. A dot immunobinding assay was developed for this purpose. By monitoring the appearance and accumulation of type IV collagen from the time of plating to post-confluence and using ascorbic acid as a modulator of extracellular matrix production, we have further characterized the development of the extracellular matrix in this model system.
Materials and Methods Collagen isolation
Types I, III, IV and V collagens were pepsin-extracted from human placenta and separated by salt precipitation from acidic and neutral salt solutions (Kresina and Miller, 1979; Miller and Rhodes, 1982). Type IV collagen was purified further by incubation with DEAE cellulose (Kresina and Miller, 1979). Collagen fractions were monitored by amino acid analysis and SDS polyacrylamide gel electrophoresis (Laemmli, 1970; Miller and Rhodes, 1982). Antibody production and purification
Rabbit anti-human type IV collagen antibodies were generated and purified based on the methods of Furthmayr (1982). Immune sera were assayed for the presence of antibody employing a passive hemagglutination assay, and high titer (> 1: 80 000) antisera were affinity-purified. The IgG fraction of whole serum was eluted from Protein Asepharose CL-4B (Pharmacia, Inc., Uppsala, Sweden) with .58% (v:v) glacial acetic acid containing 0.15 M NaCl (Goding, 1976) and dialyzed agains PBS. This IgG fraction was sequentially passed over a series of extracellular matrix components coupled to CNBr-activated Sepharose 4B (Pharmacia): fibronectin (human plasma), a mixture of types I and III collagens (human placenta), type V collagen (human placenta), and laminin (murine EHS, Collaborative Research, Inc., Bedford, MA). The IgG specific for type IV collagen was then eluted from a type IV collagen affinity
353
column with 1 M acetic acid containing 0.15 M NaCl, neutralized, dialyzed against PBS, and concentrated. Rabbit anti-mouse laminin antibodies were purchased from Collaborative Research. The preparation of rabbit anti-bovine alpha elastin antibodies was described previously (Chipman et al., 1985). These antibodies did not cross-react with types I, III, IV and V collagens or fibronectin; nor did they cross-react with each other. Passive hemagglutination
Fixed human red blood cells were incubated with tannic acid and coated with type IV collagen (5 f.lg!ml). Titer and inhibition assays were carried out as described by Furthmayr (1982). A constant amount of antibody was preincubated with various concentrations of antigen in the inhibition assay. Dot immunobinding assay
The method used in these studies was a composite of several procedures (Hawkes et al., 1982; ]ahn et al., 1984; Towbin and Gordon, 1984). Grids (1 cm 2 ) were drawn on nitrocellulose with a pencil. The nitrocellulose was soaked for 5 -10 min in distilled water and dried completely before protein application. Protein was applied to nitrocellulose in 1-3 f.ll aliquots. In some cases successive applications of protein were made to the same area, allowing the nitrocellulose to dry completely between each application. If samples contained little protein, bovine serum albumin (1 mg! ml, IgG-free) was added as a carrier. Blots were incubated for 15 min in 10% (voVvol) acetic acidl25% (voVvol) isopropyl alcohol to fix proteins to the nitrocellulose, rinsed several times with distilled water, and equilibrated in TrisTween buffer for 15 min (0.15 M NaCl, 0.05 M Tris, 0.2% Tween, pH 7.2). This was followed by incubation for 1 h with blocking buffer (Tris-Tween containing 3% nonfat dry milk). The appropriate antibodies were added to incubation buffer (Tris-Tween containing 1% nonfat dry milk): 1 f.lg!ml anti-type IV collagen IgG, 5 f.lg!ml anti-laminin IgG, or 1/200 dilution of anti-elastin serum. Blots were incubated in the appropriate antibody solution for 1 h followed by three 10-min washes in Tris-Tween buffer. [125 1]_ protein A was added to incubation buffer (500,000 cpm! ml, 70-100 f.lCilf.lg), and blots were incubated in this solution for 1 h. This was followed by a quick rinse in TrisTween buffer, four quick rinses in distilled water, and four 20-min washes in distilled water. Blots were dried at room temperature. Blocking, antibody incubation, and 25 1]_ protein A incubations were done at 37°C. All other steps were done at room temperature. Kodak XAR film was preflashed and exposed to the blots for 4 to 48 h at - 80°C in cassettes containing enhancing screens. Quantification was based on densitometry of the exposed film. Dots were scanned using a Shimadzu Dual-
e
354
A. C. Hospelhorn et ai.
Wavelength Thin-Layer Chromato Scanner Model CS-930 densitometer. Two-fold dilutions (1600-6 pg) of standard proteins were included on each blot and were used to generate a standard curve. Two or three concentrations of each sample, and duplicate or triplicate samples were blotted. The amount of a specific protein in each sample was determined from the standard curve generated under the same conditions, i. e. standard antigen diluted in media or acetic acid (cell layer), and all values for each time point were averaged. When necessary, multiple exposures of the same blot were scanned to obtain values on the linear portion of the standard curve.
Smooth muscle cell culture Neonatal rat smooth muscle cells were isolated from aortae of 3-day-old rats (Oakes et aI., 1982). First passage cells were used for all experiments. Cells were seeded in 35mm dishes (Falcon) at 1.5 x 10 4 cells/cm 2 and maintained in 2 ml of media containing 10% fetal bovine serum. Media were changed twice a week. Cells were cultured up to 20 days, with groups of cells harvested every other day starting with day 2. Twenty-four hours prior to each harvest, cells were washed (2 x) with Puck's saline and changed to 1 ml of the appropriate incubation media with or without serum and with or without ascorbate (50 f.A.g/ml). Cells grown continuously in the presence of ascorbate received ascorbate (20 f.A.g/ml) the day after plating, with each media change (twice a week), and 24 h prior to harvest (designated continuous ascorbate). When ascorbate was added only during the 24-h test period, this was denoted as short-term ascorbate treatment. DNA was analyzed using diphenylamine (Burton, 1956; Patterson, 1979). Cell counts were obtained using a Coulter counter after cells were released from culture dishes with a trypsin-collagenase solution (3 mg trypsin (Type III, Sigma Chemical Co., St. Louis, MO)-10mg collagenase (Type lA, Sigma)/12ml Dulbeccos's modified Eagle's medium) and diluted with Isoton (Coulter Electronics, Inc., Hialeah, FL).
Analysis ofcell cultures At each harvest the media were transferred to microfuge tubes and centrifuged at 16000 x g for 30 sec to remove cellular debris. Some media samples were dialyzed against distilled water, and total protein and hydroxyproline were determined by amino acid analysis. Other aliquots of media were applied directly to nitrocellulose after the addition of carrier BSA and analyzed for specific matrix proteins by dot immunobinding assays. Cell layers were washed twice with Puck's saline G, and the washes were discarded. Some cell layers were analyzed for DNA and total protein. Total protein was determined by amino acid analysis of the residue remaining after DNA extraction (Patterson, 1979). Other cell layers were incu-
bated with 1 ml of cold 0.5 M acetic acid containing 100 f.A.g of pepsin (Sigma 1:60000) for 20 hat 4°C on a shaker to release matrix proteins. The supernatant was transferred to a microfuge tube and mixed with 50 f.A.I of a pepstatin A (Sigma) solution (100 f.A.g/ml 0.5 M acetic acid). Aliquots were applied directly to nitrocellulose after the addition of carrier BSA and assayed by the dot immunobinding procedure.
Collagenase assay To assess collagenase susceptible protein, radiolabeled proline (10 f.A.Ci/ml, L-[2,3,4,5- 3H] proline, 111 Ci/mmol) was added to the cell culture media for 24 h. Radiolabeled proteins from cell cultures were then incubated with bacterial collagenase (Form III, Advanced Biofactures) (Peterkofsky et aI., 1982; Freiberger et aI., 1980). Media and sonicated cell layer samples were dialyzed extensively against 1 mM proline followed by collagenase buffer [0.15 M NaCl, 0.05 M Tris, and 0.005 M CaCh, pH 7.5 (Sage et aI., 1979)] containing 2.5 mM n-ethylmaleimide. Duplicate 200 f.A.I control and digest samples were aliquoted. Bovine serum albumin was added as carrier (200 f.A.I, 0.5 mg/200 f.A.I), and 5 BTC units of collagenase were added to digest samples. Samples were brought to 500 f.A.I with buffer and agitated at 37°C for 20 h. Proteins were precipitated with an equal volume of cold 10% trichloroacetic acid (TCA), 0.5% tannic acid and centrifuged at 16000 x g for 10 min in a microfuge. The supernatant was transferred to another tube, and the pellet was washed with 5% TCA, 0.25% tannic acid and centrifuged. The wash was combined with the supernatant, and an aliquot was added to scintillation fluid and counted. The pellets of the digested samples were hydrolyzed in 6 N HCI, and an aliquot of the hydrolysate was added to scintillation fluid and counted. Percent collagen was calculated (Peterkofsky et aI., 1982): _ _ _ _ _ _ _d~p~m_in_c_o_ll~ag~e_n (dpm in noncollagen protein x 5.4) + dpm in collagen
x 100
Isolation oftype IV collagen from cell culture medium and determination ofhydroxylation Cells were seeded in 100-mm dishes at 1.5 x 10 4 cells/ cm 2 and maintained in 15 ml of media containing 10% fetal bovine serum. On day 3 the media were aspirated. The cell layers were washed twice with serum-, proline- and lysinefree media and then were incubated with 5 ml of media containing [3H]-proline (20 f.A.Ci/ml, L-[2,3,4,5-3H] proline, 111 Ci/mmol) and [3H]-lysine (20 f.A.Ci/ml, L-[4,5- 3H]lysine, 76 Ci/mmol) with and without 50 f.A.g/ml ascorbate for 24 h. Media from three dishes were pooled, and proteolytic inhibitors were added to a final concentration of 2.5 mM EDTA, 2 mM phenylmethylsulfonyl fluoride, and 10 mM n-ethylmaleimide. The media were stored at - 20°C. After
Type IV Collagen thawing, 2 ml of media were incubated with 20 Ilg of affinity-purified type IV collagen antibody for 2 h at room temperature with gentle mixing. This solution was then loaded on a protein A-sepharose CL-4B column which had been washed with PBS containing 1 mM proline and 1 mM lysine. After loading, the column was eluted with 0.15 M NaCl, 0.58% (v:v) acetic acid. After dialysis against distilled water, the eluted material was lyophilized and hydrolyzed for amino acid analysis.
355
equally well with native and heat-denatured type IV collagen. Types I, III and V collagens, as well as fibronectin, showed no reaction. Although there was a consistent slight reaction with EHS laminin, incubation of this preparation with bacterial collagenase eliminated the reaction. Since antibodies to EHS laminin reacted equally well with control and collagenase-treated laminin samples, the reactivity of laminin samples with anti-type IV collagen IgG appeared to be caused by collagen contamination of the laminin preparation. These antibodies specifically labeled basement membranes in human lung tissue (not shown). Reaction of pepsin-solubilized human type IV collagen with affinity-purified rabbit IgG against pepsin-solubilized human type IV collagen in the dot immunobinding assay showed a linear dose response (Fig. 1).
Results Antibody specificity and dot immunobinding assay
The specificity of type IV collagen antibodies was determined by passive hemagglutination inhibition and dot immunobinding. Type IV collagen antibodies reacted
A
IN IV 6
13
25
50
100
200 400 800 1600 (pg)
B lil 045
«l
i'-
a:
t
:Ii :II
•
25
+'
a
CI 0=
2D 16
Fig. 1. Dot immunobinding assay. A) Autoradiogram of two-fold dilutions of laminin and type IV collagen incubated with antitype IV collagen IgG followed by [125 1]_ protein A. B) Plot shows linear portion of the type IV collagen standard curve. Relative areas of dots were obtained by scanning densitometry.
10 6 0 0
( 25
50
76
100
T~
126
160
176 200
22fi
IV Collogl!ln (plcograne)
260
276
:Ill
356
« z
c
A. C. Hospelhorn et al. 30
Cl
25
c
20
:::s
-y----------------------, A
CIl
Cl
~
o
15
>
10
()
CIl C.
>-
~
Cl
C
o
10
« z
2
4
6
2
4
6
8
10
12
14
16
18
20
14
16
18
20
Days In Culture
B
8
C
Cl
:::s
6
c c E ('Q
4
Cl
2
..J
c
0 0
8
10
12
Days In Culture
10
« z c
C
8
Cl
:::s
w
6
c
Cfj
~
Cl
c
4
2
0 0
2
4
6
8
10
12
Days In Culture
14
16
18
20
Fig. 2. Effect of no ascorbate (_), shortterm ascorbate (0), continuous ascorbate (0) and serum + short-term ascorbate (~) on A) type IV collagen, B) laminin and C) soluble elastin detected in the media. See text for details. The amount of laminin detected in the absence of ascorbate was essentially the same as that for each of the ascorbate conditions and is not shown.
TypeIV Collagen
Proliferation ofneonatal rat aortic smooth muscle cells in culture The proliferation of first passage neonatal rat aortic smooth muscle cells was monitored by cell counts, DNA content, and phase-contrast microscopy. There was little difference in cell number or DNA content between control and treated cultures. The most rapid proliferation occurred between days 2 and 4, and phase-contrast microscopy demonstrated that cells were confluent by day 4.
Type IV collagen detected in the media Neonatal rat aortic smooth muscle cell cultures produced more type IV collagen in the presence of ascorbate than in its absence. All cell cultures incubated in the presence of ascorbate, regardless of the time of exposure to ascorbate (short-term vs continuous) or the presence or absence of serum during the 24-h incubation period, exhibited a similar pattern of type IV collagen accumulation in the media (Fig.2A). Type IV collagen was detected in the media as early as day 2. The greatest amounts were secreted into the media prior to detection of type IV collagen in the cell layer (Fig. 3). Although the pattern of type IV collagen production was similar under the conditions examined, the data suggest that type IV collagen production was influenced by the culture conditions during the early phase ~f culture. The presence of serum in conjunction with ascorbate during the 24-h incubation period appeared to stimulate type IV collagen production. Continuous exposure to ascorbate, on the other hand, appeared to attenuate type IV collagen production. Although cultures from different primary isolates were used for the "continuous ascorbate" and "short-term ascorbate" experiments, the difference in the amount of type IV collagen found during the early phase of culture versus later time points was greatly diminished in cultures continuously exposed to ascorbate. In the absence of ascorbate, only a small amount of type IV collagen could be detected in the media after the first week in culture. As expected, in the absence of ascorbate, type IV collagen isolated from the media was found to be underhydroxylated compared with that isolated from media of cultures exposed to ascorbate (25% of the total proline was hydroxylated in the absence of ascorbate, 57% in the presence of ascorbate).
357
when the cells were incubated with ascorbate for only 24 h (Fig. 3A) and when cells were maintained in the presence of ascorbate (Fig. 3B).
Total collagen detected in the media Fig. 4 compares percent collagenase-susceptible protein (4A) to total protein (4B) detected in the media in the presence and absence of ascorbate. In each 24-h period, approximately 90% of the counts in collagen were associated with the media fraction. Based on these data the amount of type IV collagen represented less than 1% of the total collagen. Overall, the presence or absence of ascorbate made little difference in the amount of total collagen detected in the media, although preconfluent cell cultures produced slightly more collagen in the presence of ascorbate than in its absence. Hydroxyproline could be detected in the media in the presence of ascorbate but not in the absence of ascorbate, indicating that the collagen produced in the absence of ascorbate was underhydroxylated (data not shown). Compared with total collagen, type IV collagen accumulation in the media appeared to be more affected by the presence or absence of ascorbate (Fig. 2A).
Total collagen detected in the cell layer Again, somewhat surprisingly, the presence or absence of ascorbate did not appear to affect the amount of collagen detected in the cell layer during the 24-h period examined (data not shown). The level of hydroxylation in the cell layer remains to be determined. Both type IV collagen and total collagen were detected in the cell layer only after the cells had reached confluence.
Laminin detected in the media and cell layer Laminin was detected in the media by day 4 (Fig. 2B). In contrast to type IV collagen production, laminin production was not affected by the presence or absence of ascorbate. Laminin levels remained relatively constant throughout the culture period, although additional exposure to serum 24 h prior to harvest slightly enhanced laminin production on days 4-8. Small but increasing amounts of laminin were detected in the cell layer from day 12 (0.44 ng! !!g DNA) to day 20 (5.09 ng!!!g DNA) in both the presence and absence of ascorbate.
Type IV collagen detected in the cell layer In the presence of ascorbate detectable amounts of type IV collagen began to accumulate in the cell layer only after the cells reached confluence on day 4 (Fig. 3). Surprisingly, in the absence of ascorbate type IV collagen was also incorporated into the cell layer such that there was little difference between the untreated and ascorbate-treated cell cultures by the end of the culture period. This occurred both
Soluble elastin detected in the media The presence of ascorbate and whether the ascorbate exposure was short-term or long-term had a dramatic effect on the accumulation of soluble elastin and/or its fragments in the media of these cell cultures (Fig. 2C). Soluble elastin was not detected in the media in the absence of ascorbate or in cultures which were maintained in the presence of ascor-
A. C. Hospelhorn et al.
358 6
« z
C
A 5
C)
:::J
c
4
(I) C)
,!g
'0
3
0
~ (I)
2
Co
>-
I-
C)
c
O-+---III----1.._-. .-
o
2
4
6
__:..----,r---'T"""'"--r--.,.---r---1
8
10
12
14
16
18
20
Days In Culture
6
« z C
B 5
C)
:::J
c
4
(I) C)
,!g
'0
3
0
~
2
(I)
Co
>-
I-
C)
c
o
2
4
6
8
10
12
14
16
18
20
Days In Culture
bate. However, when these celle were exposed to ascorbate for 24 h, a peak of soluble elastin was readily detected in the media from day 6 to day 10. This was the case whether or not serum was present during the incubation period. In contrast to the effect of serum on both type IV collagen and laminin, the amount of soluble elastin detected in the media after an additional 24-h exposure to serum was considerably less than in the absence of serum.
Fig. 3. Effect of no ascorbate (_), shortterm ascorbate (0) and continuous ascorbate (0) on the accumulation of type IV collagen in the cell layer of cultures derived from two different primary isolates, A and B. See text for details.
Discussion The expression, accumulation (synthesis and degradation), organization, and remodeling of the extracellular matrix are extremely complex processes which can be influenced by cell culture conditions, known modulators (ascorbic acid, lathrogens), growth factors, cell-matrix interactions, and feedback mechanisms (Blumenfeld et al., 1983; Mecham, 1986; Streuli and Bissell, 1990). We have examined the long-term production of type IV collagen in
Type IV Collagen 40
359
A
30
c
CI>
0)
cu
'0
-
0
20
c
CI>
() ~
CI>
Q.
10
0+---r---,r----r--"""'T"---,r----r--"""'T"-_,.--~-_1
o
2
4
6
8
10
12
14
16
18
20
Days In Culture
50 " T " " - - - - - - - - - - - - - - - - - - - - - - . ,
B ~
.~
c
40
E
E It) CO)
c
'cu o
30
20
~
Q.
0)
:s
Fig. 4. Effect of no ascorbate (.) and shortterm ascorbate (0) on A) percent collagen of total protein detected in the media (collagenase assay), B) amount of protein detected in the media (amino acid analysis). See text for details.
10
O+---r---,r----r--"""'T"---,--~-_._-_,.--~-~
o
2
the media and cell layer compartments of neonatal rat aortic smooth muscle cells using ascorbic acid as a modulator of extracellular matrix. As long as ascorbate is present during the test period, type IV collagen in the media peaks within the first four days and quickly falls off to lower levels. This early rise coincides with the rapid proliferation of these cells in culture. In the absence of ascorbate, no type IV collagen is immunologically detected in the media for most of the culture period. Accumulation of type IV collagen in the cell
4
6
8
10
12
14
16
18
20
Days In Culture
layer is detected after the peak seen in the media and after the cells become confluent. It is remarkable that in the absence of ascorbate type IV collagen accumulation in the cell layer is only slightly less by the end of the culture period than in those cultures treated with ascorbate. The data suggest that some event occurs soon after confluence that greatly enhances type IV collagen deposition. Integrin receptors on smooth muscle cells have recently been described for type IV collagen (Clyman et aI., 1990a; Clyman et aI., 1990b; Syfrig et aI., 1991). These receptors
360
A. C. Hospelhorn et al.
may be induced by cell-cell interactions and aid in matrix assembly. Okada et al. (1990) suggest that cell-cell interactions may be responsible for the observed modulation of smooth muscle cells in long-term cultures from a synthetic state back to a contractile state with increased synthesis of type IV collagen and the ultrastructural appearance of basement membrane. It is also possible that deposition of other components, such as proteoglycans (Hamati et al., 1989), is required prior to type IV collagen deposition. There is little information available on the deposition of type IV collagen in the absence of ascorbate. One report (Katsuda et al., 1987) suggests that type IV collagen can be deposited in the absence of ascorbate. On the other hand, in two other cell systems, no deposition of type IV collagen was seen in the absence of ascorbate (Eldridge et al., 1987; Frenette et al., 1989). It is interesting to note that in a Schwann cell system, ascorbic acid was required to detect extracellular type IV collagen as well as to detect organization of laminin and heparan sulfate proteoglycan (Eldridge et al., 1987). It remains to be determined in the neonatal rat system whether the organization of type IV collagen, as well as other components, is altered in the absence of ascorbate and whether type IV collagen associated with the cell layer in the absence of ascorbate is underhydroxylated. The presence or absence of ascorbic acid has a dramatic influence on the levels of insoluble elastin in these cultures. In its absence elastin is a major component of the extracellular matrix; whereas, in the presence of ascorbic acid, insoluble elastin decreases substantially (Barone et al., 1985). In the present studies, the amount of type IV collagen associated with the cell layer at the end of the culture period was the same in the presence or absence of ascorbate suggesting that elastin deposition and type IV collagen deposition are independent. It is of interest to note the differences in soluble elastin detection in the media during the short-term and continuous ascorbate conditions. The insoluble matrix under the short-term ascorbate conditions would essentially be the same as that in the absence of ascorbate since cells have been maintained in the absence of ascorbate and would only see ascorbate for 24 h. That is, insoluble elastin would be a major component of the extracellular matrix. In the absence of ascorbate soluble elastin is efficiently incorporated into an insoluble matrix, and no soluble elastin is detected in the media. However, when cells are exposed to ascorbate for 24 h, soluble elastin is detected in the media, probably representing overhydroxylated elastin (Barone et al., 1985), and a window of peak elastin production is observed. Under continuous ascorbate conditions, no soluble elastin is detected in the media, possibly representing down-regulation of elastin production. In this regard, the lathrogen beta-aminopropionitrile (BAPN) inhibits crosslinking of elastin in the cell layer and results in increased levels of soluble elastin in the media of these cultures (Chipman et al., 1985). Recently, it has been shown in our
laboratory that incubation of these cells with BAPN down regulates elastin mRNA Uackson et al., 1991), most likely because of soluble elastin fragments feeding back on the system. The generation of soluble elastin by ascorbate may cause a similar effect. Because cell-matrix interactions are important in determining cell behavior (Kleinman et al., 1982), it is important to determine what matrix components are present under various conditions and ultimately to define differences in the cell's response to its environment. The accumulation of type IV collagen in long-term culture in both the presence and absence of ascorbate may be indicative of both normal and abnormal matrix development. This system should be useful in probing smooth muscle cell function and how it is affected by the extracellular matrix.
Acknowledgements The authors gratefully acknowledge the expert technical assistance of George Crombie, Rosemarie Moscaritolo and Valerie Verbitzki. We would also like to acknowledge the valuable suggestions of Christina Hospelhorn (Ben May Institute, University of Chicago) concerning the dot immunobinding assay. This study was supported by NIH Grants HL 07429, HL 13262, HL 19717 and HL 31453.
References Barnes, M.J.: Collagens in atherosclerosis. Collagen ReI. Res. 5: 65-97,1985. Barone, L.M., Faris, B., Chipman, S.D., Toselli, P., Oakes, B. W. and Franzblau, c.: Alteration of the extracellular matrix of smooth muscle cells by ascorbate treatment. Biochim. Biophys. Acta 840: 245-254, 1985. Barone, L.M., Wolfe, L., Faris, B. and Franzblau, C.: Elastin mRNA levels and insoluble elastin accumulation in neonatal rat smooth muscle cell cultures. Biochemistry 27: 3175-3182, 1988. Blumenfeld, 0.0., Bienkowski, R.S., Schwartz, E. and Seifter, S.: Extracellular matrix of smooth muscle cells. In: Biochemistry of Smooth Muscle, ed. by Stephens, N.L., CRC Press, Inc., Boca Raton,FL,1983,pp.137-187. Burke, J. M. and Ross, R.: Synthesis of connective tissue macromolecules by smooth muscle. Int. Rev. Connective Tissue Res. 8: 119-157,1979. Burton, K.: A study of the conditions and mechanism of the diphenylamine reaction for the colorimetric estimation of deoxyribonucleic acid. Biochem.]. 62: 315 - 323, 1956. Chipman, S. D., Faris, B., Barone, L. M., Pratt, C. A. and Franzblau, c.: Processing of soluble elastin in cultured neonatal rat smooth muscle cells.]. BioI. Chem. 260: 12780-12785,1985. Chobanian, A. V.: Adaptive and maladaptive responses of the arterial wall to hypertension. Hypertension 15: 666-674, 1990. Clyman, R. I., McDonald, K. A. and Kramer, R. H.: Integrin receptors on aortic smooth muscle cells mediate adhesion to fibronectin, laminin, and collagen. Circulation Research 67: 175-186,1990a. Clyman, R.I., Turner, D.C. and Kramer, R.H.: An al/~l-like
Type IV Collagen integrin receptor on rat aortic smooth muscle cells mediates adhesion to laminin and collagen types I and IV. Arteriosclerosis 10: 402-409, 1990b. Eldridge, CF., Bartlett Bunge, M., Bunge, R.P. and Wood, P.M.: Differentiation of axon-related Schwann cells in vitro. I. Ascorbic acid regulates basal lamina assembly and myelin formation. ]. Cell BioI. 105: 1023-1034,1987. Freiberger, H., Grove, D., Sivarajah, A. and Pinnell, S. R.: Procollagen I synthesis in human skin fibroblasts. Effect of culture conditions on biosynthesis.]. Invest. Dermatol. 75: 425 -430, 1980. Frenette, G.P., Ruddon, R.W., Krzesicki, R.F., Naser, J.A. and Peters, B. P.: Biosynthesis and deposition of a noncovalent laminin-heparan sulfate proteoglycan complex and other basal lamina components by a human malignant cell line. ]. BioI. Chem. 264: 3078-3088, 1989. Furthmayr, H.: Immunization procedures, isolation by affinity chromatography, and serological and immunochemical characterization of collagen specific antibodies. In: Immunochemistry of the Extracellular Matrix, Vol. 1, ed. by Furthmayr, H., CRC Press, Inc., Boca Raton, FL, 1982, pp. 143-178. Goding, ]. W.: Conjugation of antibodies with fluorochromes. Modifications to the standard methods. ]. Immunol. Methods 13: 215-226, 1976. Hamati, H. F., Britton, E. L. and Carey, D.J.: Inhibition of proteoglycan synthesis alters extracellular matrix deposition, proliferation, and cytoskeletal organization of rat aortic smooth muscle cells in culture.]. Cell BioI. 108: 2495-2505, 1989. Hawkes, R., Niday, E. and Gordan, ].: A dot-immunobinding assay for monoclonal and other antibodies. Anal. Biochem. 119: 142-147,1982. Herbst, T.J., McCarthy, ]. B., Tsilibary, E. C and Furcht, L. T.: Differential effects of laminin, intact type IV collagen, and specific domains of type IV collagen on endothelial cell adhesion and migration.]' Cell Bioi. 106: 1365 -1373,1988. Jackson, L. E., Faris, B., Martin, B. M., Jones, H. V., Rich, C B., Foster,]. A. and Franzblau, C: The effect of ~-aminopropionit rile on elastin gene expression in smooth muscle cell cultures. Biochem. Biophys. Res. Commun. 179: 939-944, 1991. Jahn, R., Schiebler, W. and Greengard, P.: A quantitative dotimmunobinding assay for proteins using nitrocellulose membrane filters. Proc. Natl. Acad. Sci. USA 81: 1684-1687,1984. Katsuda, S., Okada, Y., Minamoto, T. and Nakanishi, I.: Enhanced synthesis of type IV collagen in cultured arterial smooth muscle cells associated with phenotypic modulation by dimethyl sulfoxide. Cell Bioi. Int. Rep. 11: 861-870, 1987. Kleinman, H. K., Rohrbach, D. H., Terranova, V. P., Varner, H. H., Hewitt, A. T., Grotendorst, G. R., Wilkes, C M., Martin, G. R., Seppa, H. and Schiffmann, E.: Collagenous matrices as determinants of cell function. In: Immunochemistry of the Extracellular Matrix, Vol. II, ed. by Furthmayr, H., CRC Press, Boca Raton, FL, 1982, pp. 151-174. Kresina, T.F. and Miller, E.].: Isolation and characterization of basement membrane collagen from human placental tissue. Evidence for the presence of two genetically distinct collagen chains. Biochemistry 18: 3089-3097, 1979. Laemmli, U. K.: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685, 1970. Mayne, R.: Vascular connective tissue. Normal biology and derangement in human diseases. In: Connective Tissue Disease.
361
Molecular Pathology of the Extracellular Matrix, ed. by Uitto, ]. and Perejda, A.J., Marcel Dekker, Inc., New York, 1987, pp.163-183. Mecham, R.: Regulation of Matrix Accumulation. Academic Press, Inc., 1986. Miller, E.]. and Rhodes, R.K.: Preparation and characterization of the different types of collagen. In: Methods In Enzymology, Vol. 82, ed. by Cunningham, L. W. and Frederiksen, D. W., Academic Press, New York, 1982, pp. 33 -64. Murray,]. C, Stingl, G., Kleinman, H., Martin, G. and Katz, S.: Epidermal cells adhere preferentially to type IV (basement membrane) collagen.]. Cell Bioi. 80: 197- 202,1979. Oakes, B. W., Batty, A. C, Handley, C]. and Sandberg, L. B.: The synthesis of elastin, collagen and glycosaminoglycans by high density primary cultures of neonatal rat aortic smooth muscle. An ultrastructural and biochemical study. Eur.]. Cell Bioi. 27: 34-46,1982. Okada, Y., Katsuda, S., Matsui, Y., Watanabe, H. and Nakanishi, I.: Collagen synthesis by cultured arterial smooth muscle cells during spontaneous phenotypic modulation. Acta Pathol. Jpn. 40: 157-164,1990. Paralkar, V.M., Vukicevic, S. and Reddi, A.H.: Transforming growth factor beta type 1 binds to collagen IV of basement membrane matrix: implications for development. Dev. BioI. 143: 303-308, 1991. Patterson, M. K., Jr.: Measurement of growth and viability of cells in culture. In: Methods In Enzymology, Vol. 58, ed. by Jakoby, W.B. and Pastan, I.H., Academic Press, New York, 1979, pp.141-152. Perejda, A.].: Vascular and cutaneous complications of diabetes mellitus. The role of nonenzymatic glycosylatoin of collagens. In: Connective Tissue Disease. Molecular Pathology of The Extracellular Matrix, ed. by Uitto,]. and Perejda, A.]., Marcel Dekker, Inc., New York, 1987, pp. 475-490. Peterkofsky, B., Chojkier, M. and Bateman, J.: Determination of collagen synthesis in tissue and cell culture systems. In: Immunochemistry of the Extracellular Matrix, Vol. 1, ed. by Furthmayr, H., CRC Press, Inc., Boca Raton, FL, 1982, pp.19-47. Ross, R.: The pathogenesis of atherosclerosis-an update. New Engl.]. Med. 314: 488-499,1986. Sage, H., Woodbury, R. G. and Bornstein, P.: Structural studies on human type IV collagen. ]. Bioi. Chem. 254: 9893-9900, 1979. Sankey, E.A. and Barnes, M.J.: Comparison of the collagenous products synthesized in culture by pig aortic endothelial and smooth-muscle cells. Biochem.]. 218: 11-18,1984. Streuli, C H. and Bissell, M.J.: Expression of extracellular matrix components is regulated by substratum. ]. Cell. Bioi. 110: 1405-1415,1990. Syfrig, J., Mann, K. and Paulsson, M.: An abundant chick gizzard integrin is the avian al~l integrin heterodimer and functions as a divalent cation-dependent collagen IV receptor. Exp. Cell Res. 194: 165-173, 1991. Towbin, H. and Gordon, J.: Immunoblotting and dot immunobinding-current status and outlook. ]. Immunol. Methods 72: 313-340,1984.
Dr. Ann C. Hospelhorn, Surgery Research Unit - 5 West, Massachusetts General Hospital, 149 13th Street, Charlestown, MA 02129, USA.
