Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Rapid replacement of lung elastin following elastic fiber destruction brought about by instilling elastase into the lungs of hamsters is well known (1). It is believed that the repair occurs as a result of de novo synthesis of elastin by mesenchymal cells, i.e., synthetic repair. In support of this concept, an increase in elastin gene expression and extractable soluble elastin has been reported in pulmonary fibroblast cultures following elastolytic injury (2). Recent studies show that elastase digestion of pulmonary fibroblast matrices results not only in proteolysis of elastin but also in the release of a potent regulator of elastin gene transcription whose activity can influence repair mechanisms (3). We have also described a second repair mechanism, termed "salvage repair," in which we believe damaged elastic fibers are not debrided but are restored both biochemically and ultrastructurally (4). The evidence for this mechanism is that in the presence of neonatal rat aortic smooth muscle cells (NRSM) in culture, hot-alkali resistance, lost as a result of elastase treatment, is restored to previously radiolabeled elastic fibers. The restoration does not occur in the absence of NRSM cells. Possible molecular mechanisms include the crosslinking of newly synthesized tropoelastin into damaged elastic fibers and formation by lysyl oxidase of additional crosslinks between damaged domains in the elastin molecule.

Our observations with NRSM cells may have relevance for proteolytic injury of elastic fibers in blood vessel walls, as during the genesis of atherosclerosis. We wished to determine whether the salvage repair mechanism might also have relevance for repair of elastolytically damaged elastic fibers in the lungs, a process believed to occur during the development of emphysema (5, 6). Accordingly, we undertook a study of the ability of neonatal rat lung lipid interstitial fibroblasts (LIF) to carry out salvage repair in culture. We now report that salvage repair does occur in an elastase-damaged, elastin-containing matrix laid down by LIF. Salvage repair also occurs in an elastase-damaged matrix rendered acellular by azide treatment and then repopulated with LIF or other elastogenic cells, as well as a cell type that expresses lysyl oxidase but not elastin.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Lipid Interstitial Fibroblasts

LIF cells were isolated from neonatal rat lungs by enzyme dispersion and centrifugal flotation (7). Primary cultures were seeded at a density of 0.5 × 10⁶ cells/flask (25 cm²) and maintained for 1 wk in Dulbecco's modified Eagle's medium (DMEM) with 10% fetal bovine serum (FBS) and 1% pen-strep, 1% sodium pyruvate. After 1 wk, cells were passed by trypsinization and replated at a density of 0.5 × 10⁶ cells per cm². Cultures were maintained in 5% serum until confluent (around 7 days). After the cells reached confluence, the cultures were maintained in a serum concentration of 0.4% for the duration of the experiment.

General Protocol

The experimental design is summarized in Table 1. Cultures in T-25 flasks were pulse-labeled for 24 h with 1.7 µCi ¹⁴C-U-lysine on day 10 of the first passage in lysine-free media containing 5% serum. After a total of 5 to 7 wk in first passage, the cell layer was washed twice with Puck's saline to remove serum which contains elastase inhibitors. Dulbecco's balanced salt solution (DBSS) by itself (control) or containing porcine pancreatic elastase (PPE) (25 µg per flask) was added and the flasks were incubated for 45 min at 37°C as described previously (8). The elastase had been purified and characterized as described previously and exhibited greater than 90% activity by assay with ³H-elastin (8). The enzyme incubation media were poured off and analyzed as described below. Some cell layers were harvested immediately (CON-initial or PPE-initial), in order to assess the extent of injury. Fresh medium with 5% serum (to inactivate residual elastase) was added to the remaining flasks for 1 day. Cultures were then maintained in the original serum concentration for the duration of the experiment.

Treatment with sodium azide and replating with fresh cells. After elastase treatment as described above, fresh media were added to flasks for 1 h to allow flattening of the cells. Sodium azide (AZ) and -aminoproprionitrile (to inhibit lysyl oxidase bound to the matrix) were then added to final concentrations of 5% and 0.05 M, respectively. Two days later the medium was removed and the flasks washed twice with Puck's saline to remove residual azide and -aminoproprionitrile, and fresh medium was added. Five days later the medium was replaced with medium containing 5% serum and pulmonary fibroblast cells trypsinized from primary culture. The cells were plated on the acellular matrix at the same density as used originally to start the cultures (0.5 × 10⁶ cells per flask). This group was designated as PPE+AZ+LIF. Other matrices were replated with: (1) neonatal rat aorta smooth muscle cells (PPE+AZ+NRSM), isolated as previously described and maintained in DMEM with 10% FBS (4). (2) Calf pulmonary artery (calf-PA) smooth muscle cells (PPE+AZ+ calf-PA), isolated as previously described (9) and maintained in DMEM with 10% FBS. (3) The transformed mouse macrophage cell line J774A.1 (PPE+AZ+J774A.1), from the American Type Culture Collection (Rockville, MD) maintained in DMEM with 10% FBS. (4) NIH3T3 cells (PPE+AZ+NIH3T3) maintained in DMEM with 10% calf serum and high glucose (0.5 × 10⁶ cells per flask in all cases).

Acellular cultures that had not been replated (PPE+ AZ) also received fresh medium twice a week. Final harvest occurred at wk 9 or 11. Control cultures rendered acellular and replated 1 wk later (CON+AZ+LIF) were also included in experiment #1.

Co-culture of LIF with elastase-treated acellular matrices. LIF were plated in 100-mm² dishes (20,000 cells/cm²) for 5 wk. They were then treated with DBSS alone or containing 50 µg elastase, as described above. Some dishes were fixed for electron microscopy immediately after incubation; some were re-fed and allowed to repair as described above. Other dishes were rendered acellular by treatment with sodium azide and -aminoproprionitrile and, 1 wk later, LIF (20,000 cell/cm²) were replated either directly onto the acellular matrices or onto 75-mm² microporous cell culture inserts (0.4-µm pore size) (Corning Costar, Cambridge, MA). After an additional 5 wk, both the inserts and the dishes were processed for ultrastructural analysis as described below.

Biochemical analyses. Aliquots of the enzyme incubation media were lyophilized, hydrolyzed in 6 N HCl, and analyzed for protein by amino acid analysis on an amino acid analyzer (Beckman Model 6300 with System Gold software; Beckman, Palo Alto, CA). Another aliquot of the acid hydrolyzate was assessed for radioactivity (dpm) in a liquid scintillation spectrometer with external quench correction (Packard Model 1900 TR, Packard Instruments, Meriden, CT). The elastin-specific crosslink amino acids, desmosine (DES) and isodesmosine (IDES), eluting at 47 and 49 min, were used to quantify enzyme-solubilized elastin in the incubation media. The amount of rat elastin, in micrograms, that had been solubilized can be calculated by multiplying the nanomoles of DES+IDES by 43, based on the content of IDES+DES in the elastin (2 residues/1,000) and the average residue weight (85) in elastin (8).

Other aliquots of the enzyme incubation media were assessed for lactate dehydrogenase activity, a cytosolic enzyme whose presence in the medium would indicate cell lysis (10). Elastolytic activity was assessed with ³H-elastin substrate (11).

Cell layers were harvested by scraping each flask in cold water and homogenizing with a glass on glass homogenizer. Aliquots were removed for lactate dehydrogenase assay and for treatment with hot alkali in order to purify intact elastin (12). Proteolytically damaged elastin was extracted by treatment with hot alkali (8). Elastin was defined as the residual protein after treatment of the cell layer with 0.1 N NaOH at 98°C for 45 min (12). However, amino acid analysis was used to confirm the purity of the elastin, which exhibits a characteristic composition high in nonpolar amino acids (13). For selected amino acid analyses, the specific activity of DES, IDES, and lysyl residues were determined; the ninhydrin-reacted material was collected in 1-min fractions and assessed for radioactivity in a liquid scintillation spectrometer. Total protein in the cell layer of representative cultures was calculated as the sum of the protein in the hot-alkali residue plus the protein in the supernatant after hot-alkali treatment.

Calculation of salvage repair and de novo synthesis. The percent salvage repair (Equation 1) was calculated as the percent restoration of elastin-associated radioactivity (EAR) in cultures after recovery from elastase treatment (PPE_final), as compared with the EAR in control cultures (CON_final). This value was then corrected for the percent hot alkali- resistant radioactivity in cell layers immediately after elastase treatment (initial).

(1)

The amount of elastin protein repaired by the salvage repair mechanism (Equation 2) was the fractional salvage repair times the amount of elastin protein per flask present at the time of treatment with elastase.

(2)

The amount of elastin resulting from de novo synthesis (Equation 3) in cultures that had been treated with elastase was calculated as the elastin protein per flask minus the amount of elastin protein immediately after elastase treatment and minus the elastin protein repaired by the salvage repair mechanism.

(3)

For control cultures, de novo synthesis of elastin was calculated as the increase in elastin protein between the time of treatment and the final harvest.

Isolation and Analysis of RNA

Total RNA was isolated and analyzed by Northern blotting. LIF, NRSMC, calf-PA, J774A.1, and NIH3T3 cells were grown to confluence. RNA was extracted and purified using the method of Chirgwin and coworkers (14). Briefly, cell layers were scraped from two 75-cm² flasks, pooled, and homogenized in 4 M guanidinium thiocyanate solution containing 0.5% sodium N-lauryl sarcosine, 25 mM sodium citrate, and 0.1 M -mercaptoethanol. The homogenates were centrifuged for 10 min at 8,000 × g at 10°C. The supernatants were decanted and mixed with 0.025 volume of 1 M acetic acid and 0.75 volume of absolute ethanol, incubated at 20°C for 2 days, centrifuged, re-precipitated, washed with ethanol, and dried with a stream of nitrogen gas. RNA was dissolved in sterile water and A_260/280 found to be at least 2.0.

Ten-microgram samples of RNA in duplicate were fractionated on 1.1% agarose, 6% formaldehyde gels, and transferred electrophoretically to a Nytran filter (Schleicher and Shuell, Keene, NH) (15). Prior to electrophoretic transfer, the gel was cut in half so that one of each duplicate sample could be stained with acridine orange to monitor sample loading and RNA integrity and to establish electrophoretic size markers. Hybridization was performed with ³²P-labeled rat lysyl oxidase cDNA (16). Filters were exposed at 80°C to Kodak X-Omat AR film (Eastman Kodak, Rochester, NY) in cassettes equipped with an intensifying screen; the resulting autoradiograms were analyzed with a Molecular Dynamics Phosphoimager (Molecular Dynamics, Sunnyvale, CA).

Ultrastructure

For ultrastructure, cultures were fixed for 2 h at 4°C in 1% glutaraldehyde buffered with 0.1 M sodium phosphate, pH 7.1. The samples were rinsed in phosphate buffer, post-fixed in 1% osmium tetroxide, dehydrated through a graded series of ethyl alcohols, and embedded in Polybed (Polysciences, Warrington, PA). Thin sections were cut with a diamond knife on an ultramicrotome, mounted on collodion-covered nickel grids, and stained with 0.1% palladium chloride, 1% aqueous uranyl acetate, and lead citrate (17).

For experiments in which cells were plated onto microporous cell culture inserts, the cultures both on the dishes and on the inserts were fixed for 2 h in 1% glutaraldehyde in 0.1 M sodium cacodylate buffer, pH 7.4. After rinsing in cacodylate buffer, the polycarbonate membrane of the insert was excised from the insert and cut into 1-mm strips. All material was then post-fixed in cacodylate-buffered osmium tetroxide, dehydrated, and embedded as described above.

Statistics

The mean value for each treatment group was calculated. Using Statview 4.01 (Abacus Concepts, Berkeley, CA), comparisons among three or more groups were made using the Scheffe test; comparisons between two groups were made using student's t test. Probability values < 0.05 were considered significant.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Injury and Long-term Repair in Elastase-treated Cultures

The media removed after elastase treatment contained 16% (119 µg) of the total protein present in the control cell layer, 32% (42,894 dpm) of the radioactivity, and 12% (0.4 nmol) of the DES+IDES crosslinks as compared with only 2%, 2%, and 0%, respectively, for control cultures treated with media lacking elastase.

The effect of elastase treatment on the elastin component of the extracellular matrix is indicated in Figure 1. Control cultures at the initial harvest (CON-initial) contained 157 µg elastin representing 26% of the total protein. Immediately after treatment with elastase (PPE-initial), only 19% (30 µg) of the elastin protein and 20% (12,225 dpm) of the associated radioactivity were resistant to extraction by hot alkali. Considering that 12% of the elastin in the cell layer had been solubilized by the elastase, another 69% must have suffered severe enough proteolytic damage to be susceptible to subsequent treatment with hot alkali.

View larger version (24K): [in this window] [in a new window]   Figure 1.   Effect of replating on repair of proteolytically damaged elastin protein (A) and elastin-associated radioactivity (B) in acellular pulmonary fibroblast cultures (long-term repair). Cultures were pulsed with 14C-lysine on day 10, treated with 25 µg elastase on wk 5, immediately rendered acellular with sodium azide, replated with naive pulmonary fibroblast cells, and harvested on wk 11 (PPE+AZ+LIF). Controls included PPE-treated flasks that were rendered acellular and harvested on wk 11 without replating (PPE+AZ); cultures treated with medium not containing elastase (CON-initial and CON-final); and cultures treated with medium not containing elastase, immediately rendered acellular, replated, and harvested on wk 11 (CON+AZ+LIF). Note that immediately after treatment with elastase (PPE-initial), little elastin (hot alkali-resistant protein or associated radioactivity) remained, whereas 6 wk later (PPE-final), elastin protein and radioactivity were partially restored. Similar restoration is seen in cultures that were rendered acellular after elastase, replated 1 wk later, and harvested on wk 11 with the other groups. The asterisks denote that the values for PPE-initial and PPE+AZ were significantly different from PPE-final and PPE+AZ+LIF. Values for elastin protein in CON-initial, CON-final, CON+AZ+LIF, PPE-initial, PPE-final, PPE+AZ, and PPE+AZ+LIF groups were 157 ± 5 (n = 6), 353 ± 41 (5), 401 ± 36 (5), 30 ± 3 (6), 312 ± 36 (6), 18 ± 2 (6), and 269 ± 20 (7) µg elastin per flask, respectively. Values for EAR were 61,131 ± 3,184 (6), 84,740 ± 6,431 (5), 95,367 ± 3,918 (5), 12,225 ± 1,663 (6), 52,496 ± 2,176 (6), 7,404 ± 979 (6), and 51,295 ± 4,116 (7) dpm per flask, respectively.

At the final harvest, elastin protein and associated radioactivity were 312 µg and 52,496 dpm, respectively, for elastin in the cultures that had been treated with elastase (PPE-final) as compared with 353 µg and 84,740 dpm for control cultures (CON-final). That is, 6 wk after elastase treatment, elastin protein and EAR in the cell layer had been restored by the resident fibroblasts to 88 and 62%, respectively, of the levels in the corresponding control cultures. These results suggest that there was salvage repair of 42% (62%  20%) of the radioactively labeled elastin or 66 µg (0.42 × 157 µg) (Table 2). Since 19 µg of the elastin protein had been initially solubilized by the elastase treatment, salvage repair of proteolytically damaged elastin appeared to fall short of the possible 108 µg (157  19  30 µg). The remaining elastin found in cultures treated with elastase and allowed to recover for 6 wk represents de novo synthesis consisting of 216 µg (312 µg  66 µg  30 µg), where 66 µg represents salvage repair and 30 µg represents correction for the hot alkali-resistant protein present immediately after elastase treatment. By comparison, control cultures exhibited de novo synthesis of 196 µg elastin (353  157 µg).

Ultrastructural examination of control cultures at wk 5 revealed elastic fibers consisting of solid amorphous masses that were highly electron-dense after staining with palladium chloride and uranyl acetate (Figure 2). After treatment with elastase, elastic fibers exhibited a frayed appearance (Figure 3) and a granular texture that was evident at high magnification. In the more severely damaged fibers, microfibrils were exposed. By wk 11, elastic fibers in elastase-treated cultures were morphologically indistinguishable from those in control cultures (Figure 4).

View larger version (80K): [in this window] [in a new window]   Figure 2.   Control at wk 5 (CON-initial). No exposure to porcine pancreatic elastase. Elastic fibers (arrows) are solid amorphous masses stained black with palladium chloride and uranyl acetate. Bar = 1 µm.

View larger version (94K): [in this window] [in a new window]   Figure 3.   Elastase treatment at wk 5 (PPE-initial). Immediately after exposure to elastase, elastic fibers exhibit a frayed, porous appearance (arrow). Bar = 1 µm.

View larger version (64K): [in this window] [in a new window]   Figure 4.   Six weeks after elastase treatment (PPE-final). Cultures that were repaired by resident fibroblasts exhibit solid amorphous elastic fibers (arrow) identical to those seen in control cultures. Bar = 1 µm.

Effect of Replating Acellular Matrices

Five weeks after replating acellular matrices with LIF (PPE+AZ+LIF), elastic fibers were ultrastructurally similar to elastic fibers in control cultures (Figure 5). Furthermore, elastin protein and EAR were comparable to elastase-treated cultures in which the resident fibroblasts (PPE-final) were allowed to carry out the repair (Figure 1). In replated cultures there were significant increases exceeding 900 and 500% in elastin protein and EAR, respectively, as compared with cultures immediately after treatment with elastase (PPE-initial) or acellular cultures that were not replated (PPE+AZ) (Figure 1). Salvage repair in replated cultures was comparable to that seen in cultures in which repair was effected by the resident fibroblasts (42%) and accounted for 41% (61%  20%) or 64 µg of the elastin originally present before elastase. An additional 175 µg (269  64  30 µg) resulted from de novo synthesis (Table 2). Control cultures, rendered acellular and replated (CON+AZ+LIF), exhibited de novo synthesis of 244 µg of elastin (Figure 1, Table 2).

View larger version (76K): [in this window] [in a new window]   Figure 5.   Replated matrix (PPE+AZ+LIF). Five weeks after replating, elastic fibers resembling those in control cultures and in cultures repaired by resident fibroblasts (PPE-final) are present (arrows). Bar = 1 µm.

Ultrastructure of the PPE+AZ cultures that were not replated (Figure 6) and the absence of lactate dehydrogenase activity indicated the absence of living cells. At the final harvest, no significant change was found in the amount of hot alkali-resistant elastin protein (18 µg) or associated radioactivity (7,405 dpm or 9% of the CON-final value) as compared with immediately after elastase treatment (Figure 1).

View larger version (164K): [in this window] [in a new window]   Figure 6.   Non-replated matrix (PPE+AZ). When PPE+AZ cultures were not replated, elastic fibers (arrow) remained badly damaged 6 wk after elastase injury and there was no evidence of intact cells. Bar = 1 µm.

There were 0.4 IU of lactate dehydrogenase for azide-treated cultures that were replated with cells (PPE+AZ+ LIF), 0.3 for control cultures, and 0 for azide-treated cultures that were not replated (PPE+AZ).

Effect of a Short-term Repair Interval

An experiment was designed to determine how soon repair was initiated. The results are summarized in Figure 7. Exposure to elastase (PPE-initial) solubilized 27% (551 µg) of the total protein present in the control cell layer, 23% (38,418 dpm) of the radioactivity, and 3% (0.3 nmol) of the DES+IDES crosslinks. After subsequent treatment with hot alkali, 69 µg elastin protein and 9,328 dpm of the EAR remained, or 19 and 16%, respectively, of the control values of 372 µg and 59,143 dpm. After 2 wk of recovery, elastin protein and EAR were restored to 266 µg and 27,185 dpm, or 51 and 39% of the wk 9 control values. Thus, salvage repair consisted of 23% (39%  16%) of the radioactivity in the elastin and 86 µg of the elastin protein. De novo synthesis consisted of an additional 111 µg of elastin (266  86  69 µg) as compared with 146 µg for controls. The cultures treated with elastase also exhibited a significant decrease in elastin-specific radioactivity during the 2-wk recovery period as new elastin was laid down (134 ± 6 to 102 ± 2 dpm/µg).

View larger version (21K): [in this window] [in a new window]   Figure 7.   Effect of a short interval of repair on proteolytically damaged elastin protein (A) and EAR (B) in pulmonary fibroblast cultures (short-term repair). Cultures were pulsed with 14C-lysine on day 10, treated with 25 µg elastase on wk 7, immediately rendered acellular with sodium azide, replated with naive pulmonary fibroblast cells on wk 8, and harvested on wk 9 (PPE+AZ+ LIF). Controls included PPE-treated flasks that were rendered acellular and harvested on wk 9 without replating (PPE+AZ); cultures treated with medium not containing elastase (CON-initial and CON-final); and cultures treated with medium not containing elastase, immediately rendered acellular, replated, and harvested on wk 11 (CON+AZ+LIF). Note that immediately after treatment with elastase (PPE-initial), little elastin (hot alkali-resistant protein or associated radioactivity) remained. Two weeks later, elastin protein and radioactivity were partially restored. Less restoration was seen in cultures that were rendered acellular after elastase, replated 1 wk later, and harvested on wk 9 with the other groups. The single asterisk denotes that the values for PPE-initial were significantly different from PPE-final. The double asterisks denote that the values for PPE+AZ were significantly different from the values for PPE+AZ+LIF. Values for elastin protein in CON-initial, CON-final, CON+AZ+LIF, PPE-initial, PPE-final, PPE+AZ, and PPE+AZ+LIF groups were 372 ± 18 (n = 6), 518 ± 32 (3), 415 ± 25 (3), 69 ± 10 (5), 266 ± 16 (8), 81± 9 (7), and 117 ± 10 (12) µg elastin per flask, respectively. Values for elastin-associated radioactivity were 59,143 ± 4,641 (6), 68,893 ± 5,840 (3), 67,215 ± 8,770 (3), 9,328 ± 1,373 (5), 27,185 ± 1,967 (9), 10,006 ± 1,260 (7), and 15,288 ± 1,103 (12) dpm per flask, respectively.

If elastase-treated cultures were rendered acellular with sodium azide, no significant change was found in the amount of hot alkali-resistant elastin protein or associated radioactivity at the end of 2 wk as compared with immediately after elastase treatment (Figure 7).

Some of the elastase-treated cultures rendered acellular were replated with pulmonary fibroblast cells. Two weeks after treatment with elastase and 1 wk after replating, significant increases of 44 and 53% were seen in elastin protein and EAR, respectively, as compared with cultures immediately after treatment with elastase (117 versus 81 µg elastin, respectively, and 15,288 versus 9,328 dpm). Salvage repair consisted of 10% (26%  16%) of the radioactivity in the elastin and 37 µg of elastin protein, while de novo elastin synthesis was only 11 µg (117  69  37 µg). Little de novo synthesis was also indicated by the failure of the specific radioactivity of the elastin to significantly decrease during the 2-wk period after elastase treatment (137 ± 10 [n = 12] dpm/µg elastin in the replated cultures that had been treated with elastase as compared with 122 ± 3 [n = 7] dpm/µg elastin for acellular cultures that had been treated with elastase, but not replated).

The pulmonary fibroblast cells that were used to replate the acellular matrices were also followed by electron microscopy during the 7-day repair interval. Twenty-four hours after addition of cells to the matrix, the fibroblasts had migrated throughout the matrix (Figure 8) and were frequently seen in close proximity to damaged elastic fibers (Figure 9). Swollen endoplasmic reticulum, free ribosomes, and a small number of lipid inclusions were characteristic of the cells at this time. There was no evidence of cells adhering to and spreading across the surface of the matrix, and all of the matrix was exposed to the culture medium at the 24-h time point. Four days after replating, large areas of the matrix were covered by fibroblasts that were stretched to create a thin layer of cells separating the damaged matrix from the culture medium. In several areas a partially formed cover layer was observed; it appeared that the layer was produced by cells that had migrated a short distance into the matrix and then stretched to meet other cells at approximately the same level (Figure 10). By 7 days, all of the culture was covered by a continuous layer of cells (Figure 11). Cells below the cover layer often partially surrounded elastic fibers that still showed considerable evidence of elastase damage.

View larger version (93K): [in this window] [in a new window]   Figure 8.   Matrix 1 day after replating. Fibroblasts (F) have migrated throughout the matrix. Bar = 2 µm.

View larger version (117K): [in this window] [in a new window]   Figure 9.   Matrix 1 day after replating. Cells exhibit swollen endoplasmic reticulum, numerous free ribosomes, and lipid inclusions. Many cells are closely associated with damaged elastic fibers (e). Bar = 0.5 µm.

View larger version (40K): [in this window] [in a new window]   Figure 10.   Matrix 7 days after replating. Fibroblasts near the surface of the culture formed a thin continuous layer of cells to cover the matrix. The remains of an azide-killed cell (arrow) are seen above this cover. Bar = 1 µm.

View larger version (50K): [in this window] [in a new window]   Figure 11.   Matrix 7 days after replating. One of the replated fibroblasts partially surrounds still-damaged elastic fibers (e). Bar = 1 µm.

Co-culture of LIF with Elastase-treated Acellular Matrices

LIF plated on polycarbonate membranes synthesized an extracellular matrix that contained well-formed elastic fibers (Figure 12). However, elastic fibers in the elastase-treated acellular matrices on the surface of the dish below remained fragmented and resembled those in azide-treated cultures that had not been replated with LIF (Figure 13). There was no evidence of repair of the elastic fibers, even though LIF had been growing above on the membrane for 5 wk. LIF replated directly onto acellular matrices in other dishes repaired elastic fibers, as in other experiments described above.

View larger version (163K): [in this window] [in a new window]   Figure 12.   LIF plated on the polycarbonate membrane used in the co-culture experiment. Well-formed elastic fibers (arrows) are seen among the cells. Bar = 1 µm.

View larger version (100K): [in this window] [in a new window]   Figure 13.   Acellular matrix (PPE+AZ) co-cultured for 5 wk on the bottom of a Petri dish with LIF plated on a polycarbonate membrane insert. Elastic fibers showed no evidence of repair. Bar = 1 µm.

Effect of Replating with Elastogenic and Non-elastogenic Cells

An experiment was carried out to determine the relationship between percent salvage repair and formation of insoluble elastin (Figure 14). A mouse macrophage cell line, J774A.1, and NIH3T3 cells, which do not express elastin (see DISCUSSION) were used as controls. The presence of lactate dehydrogenase activity at the time of harvest confirmed the presence of viable cells. As expected, the J774A.1 cells did not exhibit salvage repair or de novo synthesis when plated onto acellular matrices of LIF cultures that had been treated with elastase. Of the other types of cells used to replate acellular matrices, NRSMC expressed the highest amount of insoluble elastin, LIF less, and CALF-PA still less. NRSMC exhibited the highest percent salvage repair and de novo synthesis of elastin. Although LIF exhibited significantly higher de novo synthesis than CALF-PA, and NIH3T3 no de novo synthesis, all three exhibited comparable salvage repair.

View larger version (20K): [in this window] [in a new window]   View larger version (20K): [in this window] [in a new window]   Figure 14.   Effect on salvage repair (A) and de novo synthesis (B) of elastin by replating cells with different elastogenic phenotypes. Acellular matrices replated with LIF, RSMC, calf-PA, or NIH3T3 cells exhibited salvage repair that was significantly different from that exhibited by acellular matrices that were not replated or were replated with the transformed mouse macrophage cell line J774A.1. Values for repair for PPE+AZ+J774A.1, PPE+AZ, PPE+AZ+ LIF, PPE+AZ+RSMC, PPE+AZ+calf-PA, and PPE+AZ+ NIH3T3 were 0 (n = 3), 0 (7), 28 ± 2 (6), 53 ± 3 (5), 25 ± 4 (4), and 27 ± 4 (4)%, respectively. De novo synthesis of elastin was 0 (3), 0 (7), 124 ± 3 (6), 714 ± 55 (5), 20 ± 3 (4), and 0 (4) µg elastin per flask, respectively.

Lysyl oxidase expression was investigated (Figure 15). NRSMC, LIF, CALF-PA, and NIH3T3, but not J774A.1, exhibited significant steady-state levels of lysyl oxidase mRNA. Acridine orange staining of duplicate samples confirmed equal loading of RNA and RNA integrity (not shown).

View larger version (56K): [in this window] [in a new window]   Figure 15.   Northern blot analysis of RNA from cells used to replate matrices using lysyl oxidase cDNA probe. Ten micrograms of RNA from NRSMC (a), LIF (b), calf-PA (c), J774A.1 (d), and NIH3T3 (e) were electrophoresed, transferred, and probed. The lysyl oxidase mRNA signal for calf appears at a higher molecular weight than that for rat or mouse (16).

Experiments to Assess Reutilization of ¹⁴C-Lysine

Following treatment at wk 5 in the long-term repair experiment, spent media was collected for 2 wk in separate pools from control and elastase-treated flasks. In the control cultures the recovered radioactivity per flask decreased from 18,000 dpm initially at wk 5 to 5,250 dpm in the final collection of spent media in wk 7. A total of 46,600 dpm and 55,390 dpm were recovered during the 2-wk period in the spent media of control and elastase-treated cultures, respectively. Within experimental error, the radioactivity in the spent media was similar to the decrease in radioactivity associated with non-elastin protein in control cultures of approximately 41,000 dpm during the 6-wk period from wk 5 to wk 11, a decrease from 72,037 to 30,883 dpm.

The proportion of EAR in lysyl residues from control cultures before treatment with elastase was determined. Only 7 ± 1% (n = 2 flasks) of the radioactivity in the elastin or 3,179 ± 187 dpm was present in lysyl residues. The specific radioactivity of lysyl residues in elastin protein was 239 ± 32 (n = 2) dpm/nmol or more than twice the specific radioactivity of 96 ± 14 (n = 2) dpm/nmol for the hot alkali-susceptible protein in the cell layer representing non-elastin proteins in the same flasks.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Repair by Endogenous LIF Cells

The experiments presented have used large amounts of elastase to produce an exaggerated injury to extracellular matrices that demonstrates a robust salvage repair mechanism in rat pulmonary LIF cultures. Although less than 12% of the elastin was solubilized from cell layers treated with elastase, 81% of the remaining elastin was initially susceptible to extraction by hot-alkali treatment. Control elastin is resistant to such hot-alkali treatment (8). Measurable salvage repair was present as early as 2 wk after treatment with elastase, as demonstrated by the partial restoration of hot-alkali resistance to the metabolically labeled elastin protein. There was also accumulation of newly synthesized (de novo) elastin. Neither elastin nor lysyl oxidase steady-state mRNA levels were elevated in elastase-treated as compared with control cultures (unpublished data). Ultrastructural studies were able to rule out a trivial mechanism for salvage repair; that is, coalescence of newly synthesized elastin around damaged elastic fibers. Damage in the interior of elastic fibers would still be visible when sections of the elastic fibers were cut and viewed under the electron microscope.

Repair by Replated Cells

When the cultures were rendered acellular shortly after treatment with elastase and not replated, measurable salvage repair was not detected. Repair could be induced by plating naive rat pulmonary LIF or other cells which express elastin and/or lysyl oxidase, but not a transformed mouse macrophage cell line, onto the acellular matrix. NRSM expressed approximately 6 times more insoluble elastin and approximately twice as much lysyl oxidase mRNA as compared with LIF cells and induced nearly twice as much salvage repair (53% versus 28%, respectively) (Figure 14). Bovine pulmonary artery smooth muscle cells produced small amounts of insoluble elastin and 25% salvage repair as compared with 28% repair for the LIF cells. NIH3T3 cells achieved 27% salvage repair, but have been shown not to exhibit detectable levels of elastin mRNA (18). On a plastic surface the bovine pulmonary artery smooth muscle cells produce insoluble elastin, but the NIH3T3 and J774A.1 cells do not (unpublished data). This bovine elastin, however, may not be effectively incorporated into a rat elastin matrix in the replating experiment. The J774A.1 cells did not repair the damaged elastin, nor did they appear to digest it, since the amount of radiolabeled material present in the acellular matrices replated with J774A.1 cells was not significantly lower than the non-replated group (PPE+AZ) (25,951 ± 4,289 versus 32,887 ± 1,529 dpm, respectively). In addition, less than 2 ng of elastolytic activity per 10⁶ J774A.1 cells was measured. We conclude from these studies that the synthesis of new tropoelastin is not required for salvage repair of damaged elastic fibers, but that repair has not been found in the absence of lysyl oxidase expression. De novo synthesis of elastin may be required for ultrastructural repair in which the "holes," produced by elastase exposure of elastic fibers, are filled in.

Little de novo insoluble elastin (11 µg) was seen 1 wk after replating, as compared with insoluble elastin derived from salvage repair (37 µg). Nevertheless, as soon as 1 day after replating, LIF cells were seen in proximity to damaged elastic fibers (Figure 9). Co-culture experiments indicated that repair did not occur if the LIF and damaged elastic fibers were separated by a membrane with 0.4 µm pores. Repair of damaged elastic fibers required close association of the damaged fiber and the elastogenic cell.

Replating control cultures that were rendered acellular resulted in levels of elastin protein and associated radioactivity similar to undisturbed control cultures. This suggests that replating naive cells on matrices not treated with elastase did not appear to depress or enhance the total accumulation of insoluble elastin as compared with control cultures over a 6-wk period (Figure 1).

With regard to the molecular mechanism of salvage repair, the following is known. The absence of salvage repair in cultures treated with elastase and then rendered acellular with sodium azide combined with -aminoproprionitrile, a potent inhibitor of lysyl oxidase, indicates that pre-existing aldehyde groups and secondary lysyl-derived crosslinks on the damaged elastin are insufficient by themselves to lead to repair. However, replating with cell types that express lysyl oxidase is sufficient to produce partial repair. This suggests that oxidation of existing peptidyl lysine residues in the damaged elastin can lead to partial salvage repair.

Reutilization of Radiolabel Is Not an Important Factor in Salvage Repair

We carried out experiments to determine whether reutilization of radiolabel from proteolytically damaged elastin and other proteins, rather than salvage repair, was an important source of radioactivity found in repaired elastin. At the time of treatment with elastase, approximately one-half of the radioactivity in the cell layer was associated with elastin. However, only about 7% of the EAR was available in lysyl residues. The remainder was in post-translationally modified forms of lysyl residues, such as DES and IDES, that were not available for metabolic reutilization. To study the fate of radioactivity released from control and elastase-treated cultures, we collected spent media. Most of the decrease in radioactivity in non-elastin protein could be found in the spent media, and in both control and elastase-treated cultures a more rapid turnover time, on average, was observed for non-elastin cell layer protein.

Another factor that would minimize reutilization of ¹⁴C-lysine was the low specific activity of available lysine, resulting from the high concentration of unlabeled lysine present in the media (1 mM) and the relatively low specific activity of cell-layer peptidyl lysine as compared with elastin-associated lysyl residues at the time of PPE treatment. The lysyl residues in the hot-alkali supernatant (non-elastin protein) had less than one-half the specific radioactivity of lysyl residues in elastin protein. Thus it is unlikely that significant amounts of radiolabeled lysine were released by enzyme-induced proteolysis of cell layer and reincorporated into de novo elastin during the recovery period. Finally, salvage repair was found with a non-elastogenic cell (NIH3T3) which should be unable to incorporate significant amounts of isotopically labeled lysine into elastin.

How relevant is this repair system to in vivo lung injury and repair? Patients with emphysema who are current smokers exhibit elevated levels of urinary DES and IDES consistent with a 4.9-fold increase in lung elastin degradation, as compared with never-smoking controls (6). There is evidence to suggest that the lung may be able to repair small amounts of damage to the connective tissue (19). This raises the possibility that emphysema develops only when the repair mechanisms of the lung are overwhelmed by a massive insult or by repeated insults over a long period of time. Another possibility is that repair mechanisms may become defective with age. In our in vitro experiments we have used fibroblasts derived from neonatal animals. Repair by fibroblasts from older animals has not been studied.

When elastase was given intratracheally to immature, rapidly growing hamsters, they appeared less susceptible than young adult hamsters to production of emphysema (20, 21). Explanations included the increased ability of rapidly growing lungs to repair lung injury (21) and less initial injury to the young lungs (20). We have cited animal model studies that contain observations consistent with salvage repair (22). We hypothesize that salvage repair, when effective, helps preserve the original architecture of the lung. Salvage repair may be prevented by debridement of damaged elastic fibers by the inflammatory milieu. Cytokines and growth factors likely modulate both salvage repair and de novo synthesis. For example, basic fibroblast growth factor is released from extracellular matrix by elastase treatment and can downregulate elastin expression in LIF cultures not treated with elastase (3, 23). On the other hand, cultures that are treated with elastase which is subsequently removed and inactivated exhibit upregulation of elastin expression when the digest is added back (2). Investigators have shown that intratracheal instillation of hamsters or rats with PPE followed by multiple dosing with an inhibitor of neutrophil elastase that does not inhibit pancreatic elastase, beginning 1-3 days after the elastase administration, can ameliorate elastase-induced emphysema (24, 25). The mechanism may involve prevention of debridement so that salvage repair of the damaged elastic fibers can occur.