Size Distribution of Human Lung Elastin-derived Peptide Antigens Generated In Vitro and In Vivo 1- 3

U. KUCICH, J. ROSENBLOOM, R KIMBELt, G. WEINBAUM, and W. R. ABRAMS

Introduction

Chronic obstructive pulmonary disease (COPO), characterized mainly radiologically and by a loss of pulmonary function, has been difficult to diagnose accurately in the early stages of the disease when clinical symptoms are minimal. Radiologic methods have difficulty detecting the destruction when the loss of lung function is less than 25 to 50070 (1). Pulmonary function tests, which are more sensitive, are not routinely performed unless the patient complains of dyspnea or presents clinical signs of COPO (2-5). The mechanism proposed by the protease-antiprotease balance hypothesis (6) of emphysema development to explain the destructive damage to lungs observed in emphysema patients predicts that an excess elastolytic burden in the alveolar interstitium causes the hydrolysis of the matrix of elastic fibers and produces elastin-derived peptides (EOP), which are then cleared into the systemic circulation. Extensive circumstantial support for the hypothesis that the human disease is an elastase-induced emphysema comes from experiments using animal models in which elastolytic enzymes wereintroduced directly into the lung (7, 8) and from the ultrastructural immunolocalization of neutrophil elastase in the alveolar interstitium of human smokers (9). Immunologic identification methods have been used to monitor the antigenic products of elastolytic damage (EOP) that appear in plasma. In early studies EOP were identified and their immunologic levelsmeasured by a passive hemagglutination assay in dogs treated with different amounts of pancreatic elastase (10). The data suggested that the breakdown of lung tissue was proportional to the amount of elastase aerosolized into the lung and that there was a relationship between the lung damage and the amount of circulating elastin-derived antigen.

SUMMARY The prote.....ntlprote_ hypothesis of emphy8erna development suggests that degradation of elastin In the lung Interstitium m-v give rise to abnormal quantltle. of circulating elastinderived peptldes (EDP)during periods of Inflammation. Recent .tudle. hllY8 .hown a relatlon.hlp between emphysema and high level. of EDPIn human pla.ma ( _ reference 11). Thl. report characterize. ela.tln digests on the basi. of antigenicity, .Ize, and method of preparation, a. well .. the size dl.trlbutlon of EDP found In the pla.mas of non.mo....., .mo....., and emphyaema patients. Gel filtration of elastin dlge.ts prepared by hydrolysl. of human lung elastin u.lng a low (1:500) ratio of neutrophil elasta.. to ela.tln generated a broad protein peak of approximately 70,000 daltons.lncontrast, a high (1:25) ratio of neutrophil ela.ta.. to human lung ela.tln gave a broad protein peak, with a .Ize dl.trlbutlon In the 10,000 to 30,000 dalton range. Thl. dlgeat .howed distinct 1mmunochemlcal properties. A polyclonal antibody directed agaln.t the low-ratio digest .howed a minimum detection of 2 nglml for the homologou. antigen but required 1,000 nglml of the high-ratio digest for detectable Inhibition In an Indirect ELISA a888Y. Gel filtration of plasmas from normal nonsmokers and the majority of normal .mokers nwealed a .Ingle Immunoreactive EDP fraction of approximately 70,000 daltons. Pla.ma. from selected normal .mo"rs and emphyaema patients with high levels of circulating EDP (> 90 nglml) fractionated Into a complex pattern of peptide. In which the 70,000 dalton component represented 50~ of the Immunoreactive material and aeveral lower molecular weight peptldes represented the remaining circulating ela.tln antigen•• These data suggest that unique epltopes are exposed by different concentrations of neutrophUelaataae and that specific Immunoreactive peptide fragments m-v appear In pla.ma of susceptible IndiVidual., po88lbly reflecting variations In the extent or type of lung ela.tln proce..lng. AM REV RESPIR DIS 1991; 143:279-283

An enzyme-linked immunosorbent assay (ELISA) was developed for EDP that was quantitative, reproducible, and sensitive and was able to detect differences in peptide levels in the circulation of emphysema patients (127 ± 47 ng/ml) when compared to normal individuals (58 ± 17 ng/ml) (11). Darnule and coworkers (12) also measured circulating EDP in humans and reported high levels in selected patient populations. Other investigators have used alternative assays to measure elastin destruction as indicators of lung disease. Goldstein and Starcher (13) developed a radioimmunoassay for urinary desmosine (an amino acid derived from cross-linked lysine, which is unique to elastin). Harel and colleagues (14) quantitated desmosine levels in 24-h urine collections from a limited number of patients with COPO or other respiratory diseases. These studies showed that COPD patients excreted larger amounts of desmosine than unaffected control subjects. However, recent reports by Davies and coworkers (15)

and Pelham and coworkers (16)have indicated that urinary desmosine was not elevated in emphysema patients compared to normal nonsmokers when a larger number of patients was examined. Although this assay has high sensitivity and is potentially useful, it requires that the sample be acid hydrolyzed, since desmosine antibodies used in this study do not (Received in original form December 8, 1989 and in revised form August 7, 1990) 1 From the Graduate Hospital, Department of Medicine, Research Division, and the University of Pennsylvania School of Dental Medicine, Philadelphia, Pennsylvania. 2 Supported by Grant HL-29734 from the National Heart, Lung, and Blood Institute; Grant 1604 from the Council for Tobacco Research-USA Inc.; Grant 87-1138from American Heart Association, and a Research Award from the American Lung Association. 3 Correspondence and requests for reprints should be addressed to U. Kucich, The Graduate Hospital, Department of Medicine, Research Division, Philadelphia, PA 19146. t Deceased.

279

280

KUCICH, ROSENBLOOM, KIMBEL, WEINBAUM, AND ABRAMS

react with this amino acid when it is present in peptide linkage. In addition, 24-h urine collection is impractical for studies of large population screenings, and attempts to quantitate plasma or serum desmosine may fail because large amounts of other amino acids would interfere in this radioimmunoassay (17). The immunology and chemical nature of the EDP generated and detected in vivo remain largely unknown. The goals of this study were to begin to characterize elastin antigens generated in vitro and in vivo, relate the change in antigen size to the conditions of elastin digestion in vitro, and separate EDP antigens present in plasma samples from nonsmokers, smokers, and emphysema patients to establish the approximate sizes of the antigens generated in vivo. To accomplish these aims, peptides were generated in vitro using different ratios of enzyme to substrate and the digestion products were fractionated by gel filtration and each fraction analyzed for their immunologic reactivity using a polyclonal antibody prepared for an amplified indirect ELISA assay (11). The sizes of elastin hydrolysis products generated during normal in vivo elastin turnover were compared to those found in patients with emphysema. The results of these studies suggest that there is extensiveproteolytic processing of lung elastin in COPD patients and selected normal smokers. Methods Goat antirabbit (GAR) IgG (heavy, H, and light, L, chains) was obtained from Cappell Laboratories (Cochranville, PA); rabbit per.. oxidase antiperoxidase complex (PAP) was obtained from Sternberger-Meyer Immunocytochemicals Inc. (Jarrettsville, MD). Micro.. titer plates (Immulon 112) were obtained from Dynatech Laboratories Inc. (Alexandria, VA). All other chemicals were reagent grade or of the highest commercially available grade.

Preparation of Elastin-derived Peptides The amorphous component of human lung elastin was prepared according to the procedure of Lansing and colleagues (18) as modified by Sloan and colleagues (8). Elastin peptides were prepared using a ratio of either 1:500 or 1:25 (wt/wt) of purified human neutrophil elastase (HNE) to purified human lung elastin. The typical 1:500 reaction mixture contained 0.05 M TRIS, pH 8.0, 0.14 M NaCI, 0.025010 Triton X-loo, 80 J.lg HNE, and 40 mg elastin in a final volume of 4.0 ml. Before addition of protease, the elastin was suspended by a combination of gentle homogenization followed by sonication with a microprobe (Kontes, Vineland, NJ). The reaction mixture was incubated for 20 hat 37 0 C with shaking and terminated by the addition of a serine

protease inhibitor [5 mM phenylmethylsulfonylfluoride (PMSF»). Unreacted insoluble elastin was removed by centrifugation (10,000 x g for 10 min).

Preparation of Antibody Elastin digests (200tJ,g/injection), obtained by 1:500 E/S ratio, were injected into different groups of New Zealand White rabbits at 2-wk intervals over a period of 2 to 3 months (19). The rabbits were bled by ear venipuncture 7 to 10 days following the last injection, the serum collected by centrifugation, and the IgG fraction partially purified by 35070 (NH4)2S04 precipitation followed byexhaustive dialysis against phosphate-buffered saline (PBS) at 4 0 C. The resulting antiserum was titered by a direct ELISA, followed by selection of a limiting amount of primary antibody to be used for the inhibition curve of an indirect ELISA assay for EDP (11). The indirect ELISA was used to quantify the immunologic reactivity of the different elastinderived peptide preparations and EDP concentrations in patient plasma, as well as to identify the location of the elastin antigens after patient plasma samples were fractionated by gel filtration (see fractionation method described subsequently). No cross-reactivity was observed with purified human neutrophil elastase, even when microgram amounts of this antigen were added in this indirect ELISA assay.

Plasma EDP Fractionation A chromatography column containing Sephacryl S-2OO (Pharmacia Fine Chemicals, Piscataway, NJ; 1.6 x 90 em) was equilibrated at 4 0 C with phosphate-buffered saline containing 0.05070 1\veen~-20 (PBS-T20). Samples (5 ml total) for chromatography were prepared by diluting plasma samples or elastin digests fivefold with PBS-T20, centrifuging the samples at 9,000 x g for 5 min at 4 0 C, and applying the supernatant to the column. Fractions (2.7 ml) were collected at a flow rate of 10 ml/h. Each fraction was analyzed by the indirect ELISA to determine the EDP concentration. The recovery of antigen from the column, calculated by summing the amount of antigen measured in each peak and compared to the amount present in the un fractionated sample, was routinely greater than 90070. The molecular size distribution of elastin antigens was established by noting the elution positions of standard proteins of known molecular weights run on the identical column as the EDP sample.

Indirect ELISA Assay Quantitation of elastin peptide digests generated in vitro, EDP levels in human plasma, and fractions from the S-2OO columns was carried out by an indirect ELISA assay as described previously (11).

Results

Patient Population Selection The plasmas to be analyzed for circulating EDP werecollected, after obtaining informed consent, from individuals in one of three groups: Group I, normal nonsmokers with normal pulmonary function measurements: Group II, smokers with normal pulmonary function studies; and Group III, patients with clinically diagnosed COPD. Spirometry and lung volumes were determined for each subject (M-800 autobox system, SRL Inc.). The results were interpreted using the Intermountain Thoracic Society criteria (20) and included a detailed respiratory history questionnaire completed by each subject after the physical examination. All members of Group II had smoked at Ieasta pack of cigarettes a day for periods exceeding 1 yr. All subjects were free of confounding pulmonary diseases such as asthma. Smokers and nonsmokerswere identified as normal on the basis of spirometry, history, and radiologic examinations.

Plasma Collection Plasma samples (10 ml) from each patient were collected directly into EDTA tubes containing 100 ul s-aminocaproic acid (1 M) and ethylenediaminetetraacetic acid (EDfA, 0.25 M). After low-speed centrifugation to remove cells, the supernatant (5 ml) was transferred to another set of tubes containing 100 ul N-ethylrnaleimide (NEM, 0.35 M) and PMSF (0.25 M). The collected plasma samples were storedat --70 0 C.

Immunologic and Biochemical Characterization of Lung Elastin-derived Peptides Human lung elastin digests wereprepared at two different ratios of neutrophil elastase to human lung elastin and fractionated by gel filtration after inhibition of elastase. A typical chromatogram (figure 1) showing the protein profile of the digests illustrates that changes in the elastase-elastin ratio give rise to the different sizes of elastin peptides. An elastase-elastin ratio of 1:500 gave a digest with a single diffuse protein peak in the 50,000 to 70,000 dalton range (figure IA); a high elastase-elastin ratio (1:25) resulted in a digest with a major protein peak in the 10,000 to 30,000dalton range (figure IB). The different sizes of EDP were also immunologically distinct. Figure 2 demonstrates that the low-molecularweight elastin-derived peptides were recognized minimally using a polyclonal antibody preparation directed against the high-molecular-weight peptides and that the small elastin antigens required ap.proximately 300-fold more peptide for detection with this ELISA assay. This small immunologic cross-reactivity was due primarily to the high-molecularweight protein seen in fractions 75

281

ELASTIN-DERIVEO PEPTlDES

J

0.03


0

o

I

I

I

~

'.~"

0

co 0.02 C\l