Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The extracellular matrix protein known as elastin has resilience and tensile properties similar to those of rubber. As such, elastin plays an important role in the structure and the functional properties of organs that require flexibility and elasticity, such as blood vessels and lungs. Elastin is an unusual polymer originating as a secreted monomer, tropoelastin, that is post-translationally modified (1). A key modification involves the oxidative deamination of a majority of the -amino groups of the peptidyl lysyl residues by the enzyme lysyl oxidase (2). The resulting aldehyde groups then condense with other oxidized or unmodified lysyl residues to form crosslinks. Little is known about the normal assembly of tropoelastin monomers into crosslinked polymeric elastin. The elastin contains characteristic crosslinks including desmosine (DES) and isodesmosine (IDES) involving four lysyl residues, three of which are post-translationally processed by the enzyme lysyl oxidase. The similarity of crosslink content in elastin from different tissues in a given species suggests the regulation of crosslink formation (3).

In the present study tropoelastin was purified from Escherichia coli expressing high levels of recombinant human tropoelastin (rhTE) lacking the hydrophilic sequence encoded by the exon 26A that is not expressed in the rat (4, 5). This recombinant tropoelastin was added to neonatal rat aorta smooth-muscle cell (NNRSMC) cultures. The rhTE was incorporated into the NNRSMC-derived insoluble elastin with the formation of elastin crosslinks. In separate experiments, acellular NNRSMC matrices were replated with Rat-1 fibroblasts, cells that were found to express lysyl oxidase but not tropoelastin. This system was used to study the assembly in the extracellular matrix of added recombinant tropoelastin molecules without competition from endogenously synthesized tropoelastin molecules.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Production and Expression of Tropoelastin Lacking Exon 26A

The rhTE corresponded to SHEL26A of Wu and colleagues (6). Its expression and purification were based on work by Martin and associates (5). A 50-ml LB medium containing 100 µg/ml ampicillin (Amp) was inoculated with BL21(DE3) pSHEL26A and grown for 16 h at 37°C. The cell pellet was washed three times with Complete Supplement Mix (CSM) media (0.67% yeast nitrogen base without amino acids, 2% glucose, 0.79 g CSM/liter) and used to inoculate 500 ml CSM containing 100 µg/ml Amp, then grown at 37°C to optical density of ~ 0.7 at 600 nm. To prepare rhTE, cells were then centrifuged and washed three times with CSM-lys medium (0.67% yeast nitrogen base without amino acids, 2% glucose, 0.74 g/liter CSM-lys) and used to inoculate 500 ml CSM-lys containing 100 µg/ml Amp. After 30 min at 37°C, expression was induced with 0.4 mM isopropyl--D-thiogalactopyranoside. After a further hour, 100 µL [4,5-³H]lysine monohydrochloride (specific activity ~ 76 Ci/mmol) was added to the culture. After another hour, cells were collected by centrifugation. rhTE was purified from a cleared lysate by adding 1.5 vol cold n-propanol at 4°C. Cold butanol (2.5 vol) was then added and the solution was stirred overnight at 4°C. The precipitate was removed at 10,000 × g and the supernatant filtered. After rotary evaporation at 20°C, the crude protein was dissolved in and extensively dialyzed against cold sterile 50 mM ammonium acetate, pH 5.0. The specific activity of the rhTE was determined for each batch and ranged from 6,300 disintegrations/min (dpm)/µg to 17,600 dpm/µg tropoelastin. The protein was filter-sterilized and lyophilized.

Cell Cultures

NNRSMCs were isolated from the aortas of 1- to 3-d-old rats and grown in primary culture with Dulbecco's modified Eagle's medium containing 3.7 g/liter sodium bicarbonate, 1% sodium pyruvate, 1% penicillin and streptomycin (DV3.7), and 20% fetal bovine serum (FBS) as described previously (7). After subcultivation into first passage in 25-cm² tissue culture flasks (20,000 cells/cm²), the cells were maintained with 5 ml of DV3.7 containing 10% FBS. The medium was changed twice weekly. Cell cultures were routinely monitored by phase contrast microscopy.

Rat-1 fibroblast cells were trypsinized from primary culture (gift of Dr. Linda Taylor, Department of Biochemistry, Boston University School of Medicine) and subcultivated at 20,000 cells/ cm². The medium, DV3.7 and 5% FBS, was changed twice weekly. After 4 wk, the cultures were harvested for determination of insoluble elastin as described later. RNA was isolated and analyzed as also described later.

Replating Acellular Matrices with Rat-1 Fibroblast Cells

Sodium azide renders cultures acellular, and -aminoproprionitrile (BAPN) is an irreversible inhibitor of lysyl oxidase. Both were dissolved in Puck's saline. The sterile, filtered azide-BAPN solution was added to 3-wk-old NNRSMC cultures with final concentrations of 5% sodium azide and 0.05 M BAPN. The medium was removed 2 d later and the flasks were washed twice with Puck's saline to remove residual azide and BAPN. Fresh medium containing 10% FBS, but not azide or BAPN, was added and the incubation repeated. The medium was replaced 4 d later and Rat-1 fibroblast cells trypsinized from primary culture were added. Aliquots of the trypsinized cells were also subcultivated in tissue culture flasks with no pre-existing matrix. At 5 d later RNA was harvested from the cultures.

Incubation with rhTE

After 3 wk in first passage, NNRSMC were washed twice with Puck's saline to remove serum and 5 ml Dulbecco's balanced salt solution was added. As a control, BAPN was sterile-filtered and added to selected cultures to a concentration of 100 µM. Puck's saline containing rhTE was added to all cultures. Serum was omitted when rhTE was added, to avoid proteolytic enzymes present in serum (8). After 3 d, the medium was removed and assessed for [³H] radioactivity, and 5 ml DV3.7 containing 10% FBS was added back to the cultures. Each time, medium was replenished; fresh BAPN was added to the cultures previously receiving BAPN. Spent medium was assessed for radioactivity. When acellular matrices were replated with Rat-1 fibroblasts, cultures were incubated with rhTE for only 1 d before removal of the spent medium and addition of DV3.7 containing 5% FBS. Longer periods of time in the absence of serum resulted in lifting of the cell layer.

Harvest of Cell Layers for Lactate Dehydrogenase and Elastin Measurement

Cultures were harvested by scraping each flask in cold water containing a protease inhibitor cocktail (50 mM ethylenediaminetetraacetic acid, 0.1 mM phenylmethylsulfonyl fluoride, 0.04 mM leupeptin, and 0.03 mM pepstatin) and homogenizing with a glass-on-glass homogenizer. An aliquot was removed for assay of lactate dehydrogenase (LDH), a cytosolic enzyme proportional to the number of viable cells (9). The remainder was treated with hot alkali to purify intact elastin (10). Elastin is defined as the residual protein after treatment of the cell layer with 0.1 N NaOH at 98°C for 45 min (hot alkali method) (11). The supernatant material from the hot alkali treatment was assessed for [³H] radioactivity. The hot alkali residue (elastin) was hydrolyzed in 6 N HCl for 24 h at 110°C. Amino acid analysis (Beckman Model 6300 with System Gold software; Beckman Instruments, Palo Alto, CA) with an 80-min cycle was used to confirm the characteristic amino acid composition of the elastin, typically consisting of more than 80% nonpolar amino acids (3). Before loading samples of hydrolyzed elastin on the amino acid analyzer, aliquots were assessed for radioactivity.

Measurement of Radioactivity in Modified Lysyl Residues

Characterization of elastin-associated [³H] radioactivity, originally present in rhTE as [4,5-³H]lysine, was performed by collecting the ninhydrin-reacted material eluting from the amino acid analyzer in 1-min fractions. Radioactivity in each fraction was measured as disintegrations per min using a liquid scintillation spectrometer with external quench correction (Packard Model 1900 TR). Elution times for radioactive standards were also determined. [¹⁴C]DES and [¹⁴C]IDES were prepared as previously described (12). [³H]Lysinonorleucine was a generous gift of Dr. Lila Graham (Department of Biochemistry, Boston University School of Medicine) (13). Acid-hydrolyzed samples analyzed included elastin, hot alkali supernatant material, and spent medium.

Assay of Lysine Oxidation

Lysyl oxidase is the enzyme that catalyzes the oxidation of endopeptidyl lysine residues to peptidyl -aminoadipic--semialdehyde in collagen and elastin, initiating the formation of covalent crosslinks (2). Oxidative deamination of the -amino group of [4,5-³H]lysyl residues, metabolically incorporated into rhTE, releases tritium ions from the carbon-5 position by solvent exchange with water (14). Tritiated water formed by the cultures during and after the treatment with the [³H]rhTE was assessed by vacuum distillation of 1-ml aliquots of spent medium and measured by liquid scintillation spectrometry (4). The results, in disintegrations per min, were expressed per flask.

RNA Isolation and Analysis

RNA was extracted from cells present in two 75-cm² flasks (10, 15). After electrophoresis of the RNA and transfer to a Nytran filter, the portion of the filter containing the 28S ribosomal RNA was hybridized with ³²P-labeled rat tropoelastin and exposed at 80°C to Kodak X-Omat AR film in cassettes equipped with an intensifying screen. After exposure, the filter was stripped by heating for 30 min at 80°C in 0.1 × saline sodium citrate and 0.1% sodium dodecyl sulfate (SDS). It was then re-probed with ³²P-labeled lysyl oxidase complementary DNA (16). The bottom portion of the filter was hybridized with ³²P-labeled glyceraldehyde-3-phosphate dehydrogenase.

Ultrastructure of Acellular Matrices Replated with Rat-1 Cells

Cultures were fixed for 2 h at 4°C with 1% glutaraldehyde buffered with 0.1 M sodium phosphate, pH 7.1. They were then rinsed in phosphate buffer, postfixed for 1 h in 1% osmium tetroxide, dehydrated through a graded series of ethyl alcohols, and embedded in Polybed (Polysciences; Warrington, PA). Thick sections (1 micron) were cut with glass knives and stained with toluidine blue. Thin sections were cut with a diamond knife on an LKB ultramicrotome and were mounted on collodion-covered nickel grids. The sections were stained with 1% palladium chloride (17), 1% aqueous uranyl acetate, and lead citrate. Sections were examined with a Philips 300 electron microscope.

Statistics

The mean value and standard deviation (SD) of the mean for each treatment group were calculated using Statview 4.01 (Abacus Concepts, Berkeley, CA). Statview 4.01 was employed to make post hoc comparisons among three or more groups using Fisher's protected least significant difference test (analysis of variance); comparisons between two groups were made using Student's t test. Probability values < 0.05 were considered significant.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Analysis of Tropoelastin Lacking Exon 26A

Amino acid analysis of rhTE yielded an amino acid composition (61% of the residues were glycine, alanine, or valine) consistent with that reported for human tropoelastin (4). Collection of 1-min fractions eluting from the amino acid analyzer and analysis of each fraction by liquid scintillation spectrometry indicated that more than 99% (3,058 out of 3,078 dpm) of the recovered radioactivity eluted with the lysine peak. SDS-polyacrylamide gel electrophoresis of rhTE revealed a major band at 60 kD (Figure 1).

View larger version (36K): [in this window] [in a new window]   Figure 1.   Coomassie blue-stained SDS-polyacrylamide gel of rhTE (lane 2). Standards were run in lane 1.

Insolubilization of rhTE by Smooth-Muscle Cells

The first experiment was designed to assess the dose- response relationship between the amount of rhTE added over a 10-fold range and rhTE incorporated by NNRSMCs into insoluble elastin. On the basis of radioactivity, the addition of 1.5, 5.8, or 14.5 µg of rhTE resulted in recovery 3 d later of 8 ± 3% (mean ± 1 SD, n = 3) of the initial radioactivity in the hot alkali-insoluble elastin (Table 1). Radioactivity recovered in the spent medium, hot alkali-insoluble elastin, and hot alkali supernatant combined represented 66 ± 15% (n = 3) of the added rhTE. In the presence of BAPN, the amount of rhTE associated with the elastin was decreased. These values likely underestimate the amount of rhTE associated with the elastin, due to the loss of tritium from the 5-carbon position of lysyl residues during catalysis by lysyl oxidase (see DISCUSSION).

Time Course of Insolubilization of rhTE and Formation of DES Crosslinks

A time-course experiment with NNRSMC cultures was carried out using the highest dose of rhTE (14.5 µg) employed in the previous dose-response experiment. The amount of rhTE associated with insoluble elastin increased from 0.31 ± 0.05 (n = 2) µg after a 1-d incubation to 0.65 ± 0.08 (n = 2) µg after 3 d of incubation, representing 5% of the added rhTE. This value did not increase after rhTE was removed from the medium and replaced with fresh medium containing serum (Figure 2). The presence of BAPN decreased the amount of radioactivity associated with the insoluble elastin by 4-fold or more. The radioactivity in the cell layer not associated with elastin and found in the hot alkali supernatant decreased from 11% of recovered radioactivity on Day 1 to 4.5% at Day 14 (not shown).

View larger version (22K): [in this window] [in a new window]   Figure 2.   The time course of insolubilization of rhTE. NNRSMC cultures were incubated for 1 or 3 d with 14.5 µg rhTE, the medium was removed, medium with serum was added back, and the incubation continued without rhTE for a total of 7 or 14 d with refeeding. BAPN was present as a control in some of the flasks. NaOH treatment of the cell layer was used to isolate the insoluble elastin fraction, which was assessed for radioactivity. The incorporation is expressed as micrograms of rhTE (mean ± SD) per flask. Two flasks were analyzed per data point. Total mean elastin protein per flask was 0.4, 0.5, 0.8, and 1.6 mg in samples taken on Days 1, 3, 7, and 14, respectively.

The radioactivity found in DES and IDES crosslinks increased with time. After 1 d of incubation with rhTE, 56% of the recovered elastin-associated radioactivity eluted with lysyl residues, 10% with DES and IDES crosslink amino acids, and 10% with lysinonorleucine (Figure 3, top panel). Most of the remaining radioactivity (18% of recovered radioactivity) eluted between 30 and 45 min (see DISCUSSION). At 14 d after treatment with rhTE additional crosslinking had occurred, lowering the relative amount of radiolabeled lysine: 32% of the recovered radioactivity coeluted with lysine, 24% with DES and IDES, and 6% with lysinonorleucine (Figure 3, top panel). Expressed as the ratio of DES and IDES radioactivity to lysine radioactivity, the ratio was 0.18 for Day 1, 0.26 for Day 3, 0.53 for Day 7, and 0.76 for Day 14. The presence of BAPN for 14 d resulted in the inhibition of lysyl oxidase-mediated conversion of peptidyl lysyl residues and a decrease of insoluble rhTE. Amino acid analysis of BAPN-treated cultures showed a single radioactive peak that eluted with lysine (Figure 3, bottom panel).

View larger version (30K): [in this window] [in a new window]   Figure 3.   Elution time of radioactivity from an amino acid analyzer in 1-min fractions after loading elastin hydrolyzates from incubation of rhTE with NNRSMC cultures. Top panel: Samples were obtained 1 d after rhTE was added to cultures (428 dpm loaded, representing 22% of the material from one flask; striped line) or 14 d after rhTE (700 dpm loaded, representing 16% of the material from one flask; solid line). For the 14-d material, 151 dpm were recovered associated with DES and IDES (horizontal bar), 199 dpm with lysine, and 169 dpm in the 30- to 45-min region. Bottom panel: At 14 d after rhTE in the presence of BAPN (390 dpm loaded, representing 53% of the material from two pooled flasks; solid line). Standards: DES and IDES (250 dpm of each loaded; striped line), lysinonorleucine (195,000 dpm loaded, data points shown at 500-fold reduction; circles). Note that the sample from cultures incubated with rhTE in the presence of BAPN has a single peak of radioactivity eluting with lysine.

Spent medium was also collected. The spent medium was assessed for cumulative, BAPN-inhibitable release of tritiated water. The cumulative amount of BAPN-inhibitable tritiated water was roughly parallel to the amount of elastin-associated tritium radioactivity (not shown). Cumulative tritiated water did not increase significantly after the rhTE was removed at Day 3 and, for the four time points, represented 91 ± 21% (n = 8) of the insoluble elastin-associated tritium radioactivity corrected for that found with BAPN controls.

The following results indicate that nearly all radioactive lysyl residues that had been modified by lysyl oxidase in NNRSMC cultures were present in the insoluble elastin fraction. The spent medium from the 3-d incubation was assessed by amino acid analysis of acid hydrolyzates. This resulted in 2,114 dpm or 100% of the recovered radioactivity coincident with the lysine peak. The absence of radioactivity in other fractions eluting from the amino acid analyzer suggest that the radioactive lysyl residues in the medium had not been modified by lysyl oxidase. Amino acid analysis of the hot alkali supernatant material after a 1-d incubation with rhTE revealed that 93% of the recovered radioactivity (568 dpm) was coincident with the lysine peak.

Selection of a Cell that Expresses Lysyl Oxidase but Not Elastin

A culture system was devised to eliminate the competitive effect of endogenously synthesized tropoelastin on the association of rhTE with elastin in the extracellular matrix. NNRSMC cultures were rendered acellular as described in MATERIALS AND METHODS. Selected NNRSMC matrices that were prepared for electron microscopy 2 wk after this treatment showed that a substantial matrix containing collagen and elastic fibers, but no intact smooth-muscle cells, was present (Figure 4A). The absence of intact cells was confirmed by LDH assay showing no enzyme activity in the matrices. When the acellular matrices were replated with Rat-1 fibroblast cells and examined 5 d later, the Rat-1 cells had migrated throughout the matrix (Figure 4B) and were often in close proximity to many of the elastic fibers (Figure 4C).

View larger version (70K): [in this window] [in a new window]   Figure 4.   (A) Typical ultrastructure of NNRSMC matrices after smooth-muscle cells have been killed with sodium azide. Bar, 1 µm. (B) At 5 d after replating with Rat-1 cells, the Rat-1 cells are integrated throughout the matrix. Bar, 2 µm. (C) At higher magnification many of the Rat-1 cells are seen in close proximity to elastic fibers. Bar, 1 µm. (Arrow points to elastic fibers.)

RNA from cultures of Rat-1 cells grown on plastic or plated on acellular matrices of NNRSMCs was extracted for Northern blot analysis. The results revealed that plating Rat-1 cells on plastic or on a matrix containing elastin did not induce detectable elastin expression (not shown). Lysyl oxidase message levels were not increased by plating on preformed matrices containing elastin as compared with plating on plastic. Rat-1 cells were also grown on plastic and analyzed by amino acid analysis, verifying that the cells did not synthesize hot alkali-insoluble elastin (data not shown).

Replated Rat-1 Fibroblasts Insolubilize rhTE and Form Crosslinks

Having characterized the Rat-1 fibroblast's lack of elastogenic capabilities, we assessed the ability of these cells to assemble exogenous tropoelastin. To increase the amount of radioactivity recovered in the peaks corresponding to crosslink and lysyl residues, flasks were incubated for 1 d with 87 µg rhTE. At 14 d later 0.83 ± 0.02 (n = 3) µg/flask was found associated with the endogenous elastin. The ratio of DES and IDES (188 dpm) to lysine-associated radioactivity (490 dpm) was 0.38 (Figure 5).

View larger version (25K): [in this window] [in a new window]   Figure 5.   Crosslink profile of rhTE incorporated into insoluble elastin of acellular NNRSMC cultures replated with Rat-1 cells. Shown is the pattern of [3H] radioactivity in elastin hydrolyzates eluting from an amino acid analyzer. Samples were obtained 14 d after a 1-d incubation with rhTE (1,510 dpm loaded, representing 10% of the material from one flask).

Spent medium was assessed for cumulative BAPN-inhibitable release of tritiated water as described earlier. Tritiated water corresponded to 120 ± 6 (n = 3)% of the elastin-associated tritium radioactivity. Amino acid analysis of the spent medium revealed no modified radioactive lysyl residues; 963 dpm eluted under the lysine peak (not shown).

Rat-1 Fibroblasts Plated on Plastic Do Not Insolubilize rhTE

Confluent Rat-1 fibroblast cultures on plastic were incubated for 1 d with rhTE (14 µg) followed by 3 d with fresh medium. The hot alkali-insoluble fraction contained only 0.03 (n = 2) µg rhTE as compared with 0.02 (n = 2) µg for the BAPN control (P = 0.31). The amino acid composition of the insoluble fraction was not consistent with elastin (not shown). Further, all of the radioactivity eluted with lysyl residues. This indicates the absence of modified lysyl residues.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Experiments presented here used nonelastogenic Rat-1 fibroblasts, plated on the extracellular matrix laid down by the previous elastogenic cells. Recombinant human tropoelastin that was added became associated with the extracellular matrix elastin. This association was accompanied by the formation of DES and IDES crosslinks characteristic of mature elastin and was blocked by the presence of BAPN, an irreversible inhibitor of lysyl oxidase. These findings also apply to experiments using living NNRSMC cultures. In the presence of BAPN, there was decreased association of rhTE with the elastin component of the extracellular matrix and decreased formation of crosslinks, accompanied by a corresponding reduction in the release of tritiated water from the rhTE substrate. The association of rhTE with extracellular matrix elastin in the presence of BAPN may have involved hydrophobic interactions, inasmuch as it was difficult to find evidence of new crosslink formation.

The recovery of tritiated water from the catalytic effect of lysyl oxidase on peptidyl lysine indicates that the specific radioactivity of the rhTE that was incorporated into insoluble elastin was reduced approximately 50% by solvent exchange of the carbon-5-position tritium (14). Recombinant tropoelastin contains 47 residues lysine per 1,000 residues (5). Mature human and rat elastins contain approximately 5 residues of lysine per 1,000 residues, indicating that approximately 89% of the lysyl residues have been chemically modified. This compares with our finding that cumulative tritiated water represented around 91% of the insoluble elastin-associated radioactivity when rhTE was incubated with NNRSMCs and 120% when incubated with replated Rat-1 fibroblasts. Thus, our assumption of constant specific radioactivity of rhTE may have resulted in our underestimating the incorporation of rhTE into insoluble elastin by nearly 2-fold.

The analysis of spent medium and hot alkali supernatant material indicated that once acted upon by lysyl oxidase, most of the rhTE is incorporated into insoluble elastin. The low efficiency of rhTE incorporation in our in vitro system was probably the result of several factors. One factor may be proteolysis of rhTE that likely occurred in the medium. In situations in which endogenously produced tropoelastin is secreted onto elastic fibers by elastogenic cells, the distance between the cell and the site of incorporation in the matrix is likely quite small. The relative efficiency of incorporation of tropoelastin in vivo has been estimated and varies by at least a factor of 3 depending on the tissue studied (18).

In parallel with the first dose-response experiment, we estimated the maximal effect of catabolic reutilization by NNRSMCs of the [³H]lysine in rhTE by incubating cultures with similar amounts of [³H] radioactivity in the form of [4,5-³H]lysine, rather than rhTE (not presented in RESULTS). For those two experimental systems the overall specific radioactivity of lysine in the media, which contained 1 mM lysine, was similar. Three-fold lower levels of elastin-associated radioactivity were found in the cultures incubated with [³H]lysine as compared with rhTE (2 versus 6%, respectively). These results suggest that reutilization of lysyl residues from catabolized rhTE for production of endogenous tropoelastin was likely not a major factor in the experiments with living NNRSMC cultures. However, one cannot exclude the possibility of selective reutilization of lysine derived from the catabolism of rhTE.

Although the amount of rhTE incorporated into insoluble elastin did not increase after a 3-d incubation and removal of rhTE, there was evidence of ongoing crosslink maturation in living NNRSMC cultures. DES and IDES radioactivity represented 10% of the radioactivity eluting from the amino acid analyzer for the 1-d incubation time, 20% for 3 d of incubation, 25% at 7 d, and 24% at 14 d. Most of the remaining radioactivity not in lysyl residues eluted between 30 and 45 min and is thought to be associated with crosslinks involving two modified lysyl residues, but these crosslinks are not well defined for elastin that has not been subjected to sodium borohydride reduction and alkaline hydrolysis (19). A difficulty with the use of this procedure is the chemical reduction of DES and IDES (19).

In our previous study, which used a different preparation of rhTE, a similar percentage of recovered radioactivity (7 to 12%) was found in the hot alkali insoluble elastin (20). That study also employed NNRSMC cultures, but the formation of DES or other crosslinks was not assessed. Using partially purified [³H]lysine radiolabeled tropoelastin, Faris and coworkers found a ratio of radioactivity in DES and IDES to that in lysine of 0.92 after a 16-d chase period (21). In the present study a ratio of 0.76 was measured 14 d after the addition of rhTE.

Because NNRSMCs synthesize considerable amounts of tropoelastin, we devised an experimental approach to eliminate the possible competitive effects of endogenously produced tropoelastin on the incorporation of added rhTE. Northern analysis indicated that Rat-1 fibroblast cells plated on plastic or on acellular NNRSMC cultures express lysyl oxidase, but not elastin.

Although we verified that Rat-1 cells grown on plastic did not synthesize insoluble elastin, an experiment was carried out to confirm that Rat-1 fibroblasts replated on acellular NNRSMC matrices do not synthesize and incorporate tropoelastin. The replated Rat-1 fibroblasts were pulsed with [¹⁴C]lysine. The [¹⁴C]lysine-labeled, hot alkali- insoluble material contained only trace amounts of [¹⁴C] radioactivity and those eluted with lysine, rather than with crosslinks (not shown). Thus, another advantage to replating with Rat-1 fibroblasts was that radiolabeled lysine from catabolized rhTE could not be reutilized for incorporation into insoluble elastin.

The absence of endogenously produced tropoelastin did not improve the efficiency of tropoelastin incorporation into insoluble elastin. The efficiency of incorporation of rhTE in the presence of Rat-1 fibroblasts was 4% of an 8-µg rhTE dose used in an initial experiment (not presented in RESULTS). This compares with the values for smooth-muscle cells of 6% using the 5.8- and 14.5-µg doses of rhTE and 5% in the time-course experiment with 14.5 µg rhTE. Factors that might be expected to decrease the efficiency of incorporation of rhTE in the presence of Rat-1 cells include the following. Because cell layers replated with Rat-1 fibroblasts tended to lift when exposed for longer than 1 d to the serum-free medium used for incubation with rhTE, Rat-1 cultures were incubated with rhTE for only 1 d as compared with 3 d for NNRSMCs. Another possibility is that an elastin binding protein that is thought to be involved in the targeting of tropoelastin to sites of fiber formation on elastogenic cell surfaces might be missing or present at lower levels in Rat-1 fibroblasts (22).

We considered that the observed crosslinks associated with added rhTE might be intermolecular aggregates of rhTE that formed in the medium and precipitated out, thus appearing to be associated with elastin in the extracellular matrix. Vrhovski and coworkers have shown that rhTE coacervates efficiently at 37°C (23). Bedell-Hogan and coworkers were able to generate a product that was insoluble in hot 0.1 N NaOH by incubating purified lysyl oxidase with recombinant human tropoelastin (4). In the present study we added rhTE to confluent Rat-1 fibroblasts that had been plated on plastic. Under these conditions, there would be no pre-existing elastin matrix. At 4 d later there were negligible amounts of radioactivity in the hot alkali-insoluble fraction and the amino acid composition was not consistent with that for elastin. In addition, there was no evidence of radioactivity eluting with crosslinks.

This is the first report demonstrating the crosslinking of tropoelastin into extracellular matrix elastin by a nonelastogenic cell type. BAPN-inhibitable release of tritiated water and formation of DES crosslinks accompanied the insolubilization of rhTE in the matrix. Additional studies will be required to determine whether the incorporated elastin is ultrastructurally and functionally appropriate and whether incorporation efficiency changes during two or more separate additions of rhTE. Those studies might also reveal the potential usefulness of the replated Rat-1 culture for studying incorporation of recombinant tropoelastin isoforms. Potential therapeutic uses for recombinant tropoelastin include disease states, such as aortic aneurysms and pulmonary emphysema, that might be ameliorated by improved elastin repair (24, 25). The delivery of exogenously produced human tropoelastin to loci containing damaged elastic fibers might represent a novel approach to accelerated elastin repair.