INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung transplantation extends survival of patients with end-stage lung diseases. Indications for lung allotransplantation include emphysema, idiopathic pulmonary fibrosis (IPF), primary pulmonary hypertension, bronchoalveolar cell carcinoma, and others. Postoperative survival is limited in more than half the cases by the development of obliterative bronchiolitis (OB) (1). Acute rejection, which often precedes OB, is typified by lymphocytic inflammation around small vessels and airways (2).

Nitric oxide (·NO) and the product of its reaction with superoxide (O₂), peroxynitrite anion (ONOO), have been implicated in lung allograft rejection. ·NO production, assessed by serum nitrite measurements, increased during acute rejection of allografts in a rat lung transplant model. Aminoguanidine, an inhibitor of inducible nitric oxide synthase (iNOS), decreased rat lung allograft rejection (3). In contrast, epithelial iNOS decreased in a heterotopic rat tracheal allograft model of chronic OB (4). Aminoguanidine appeared to worsen experimental OB, whereas L-arginine, a precursor of ·NO, significantly reduced obliteration of tracheal lumens.

Nitrotyrosine (NT), a marker of protein nitration by peroxynitrite, has been found in inflammatory cells, airway epithelium, and vascular endothelium of human lung allotransplants (5). Nitrotyrosine staining has been noted in other lung injuries, including acute respiratory distress syndrome (ARDS) (6), idiopathic pulmonary fibrosis (IPF) (7), oxygen toxicity (6), and asbestos inhalation (8).

Airway epithelial lining fluid (ELF) is adjacent to a rich source of ·NO, the respiratory epithelium, and may contain inflammatory cells that produce O₂ and myeloperoxidase (MPO). ·NO can react rapidly (> 10⁹ M¹ s¹) with O₂ to form ONOO, a strong oxidant capable of nitrating proteins (9). MPO also can catalyze the formation of a nitrating and chlorinating intermediate using nitrite and hydrogen peroxide or hypochlorite (10).

Nitration of ELF proteins could have functional implications. Exposure of surfactant to reactive nitrogen species decreases its ability to lower surface tension (11). Exposure of surfactant protein A (SP-A) to ONOO decreased its ability to aggregate lipids and interact with SP-B and SP-C to lower surface tension (12). Surfactant recovered from sheep that breathed high concentrations of ·NO (80 to 200 ppm) also failed to reach a low surface tension (13). Nitration can interfere with signaling pathways that involve tyrosine kinase phosphorylation. Kong and colleagues (14) show that ONOO nitrates two tyrosines in a synthetic peptide corresponding to phosphorylation sites of a lymphocyte-specific tyrosine kinase. Nitration completely inhibits phosphorylation and could impair signal transduction.

On the basis of these considerations, we designed this study to measure NT, nitrite, and nitrate in airway lining fluid and total nitrite in blood of lung transplant patients. Our goals were to find whether lung allotransplants harbored metabolites of increased reactive nitrogen species (i.e., NT, NO₂, or NO₃) compared with normal control subjects and to test whether such markers correlated with lung inflammation.

Both patients who had undergone lung allotransplantation and normal control subjects have detectable NT (by HPLC and ELISA) in ELF. Chlorotyrosine (CT) is present in the ELF of patients with lung allotransplants, but not in control subjects. Total nitrite is present in significantly greater concentrations in the ELF and blood of allotransplant recipients than in those of normal control subjects. Immunohistochemistry showed NT and iNOS to be variably distributed and colocalized in some open lung biopsies.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Study Population

We sampled all lung transplant patients at UAB who underwent surveillance bronchoscopy for 18 mo. That population is heterogenous with respect to underlying lung disease, time since transplantation, and type of transplant (single or bilateral). Patients were not stratified or selected in any way. These protocols were approved by the UAB institutional review board (IRB) for human studies.

The following clinical data were obtained at the time patients underwent bronchoscopy: FEV₁, % change from baseline FEV₁, whether single or double transplantation, days after transplantation, presence of infection (BALF cultures and cytology), A and B histology scores (transbronchial biopsies), and whether treatment for rejection occurred around the time of the bronchoscopy. The patients' mean FEV₁ at the time of sample collection was 2.26 ± 0.88 L (median, 2.09 L, 25th and 75th percentiles, 1.55 and 2.79 L, respectively).

Clinical Examinations

Bronchoscopies were done routinely as part of a post-transplant protocol. Patients underwent fiberoptic bronchoscopy using standard clinical techniques (15). Bronchoalveolar lavage was done at the time of bronchoscopy with the patient conscious, but sedated. Four 30-ml aliquots of normal saline were instilled into a subsegmental airway. The lavage fluid dwell time was minimized (less than 1 min per aliquot) to prevent artifacts caused by diffusion of molecules from serum into BALF. Fluid was aspirated into a syringe and kept on ice. Lavage fluids were pooled. An aliquot (~ 10 to 15 ml) was removed for study and the remainder sent for clinical testing. The fluid was centrifuged at low speed to remove cells and cellular debris. Fluids were stored at 80° C until analysis.

Patients had transbronchial lung biopsies at bronchoscopy, which were processed routinely for pathologic examination. The degree of vessel (A0-Ax) and airway (B0-Bx) inflammation was categorized according to the scheme proposed by the Lung Rejection Study Group (2). In this system the A score represents the amount of vascular inflammation, and the B score represents the amount of airway inflammation. The scores depend on the number and distribution of lymphocytes around the structure. Such tissue samples were too small to provide material for NT staining. However, patients occasionally had open lung biopsies or video-assisted thoracoscopic biopsies for clinical indications. We obtained blocks from some of these biopsies. Sections were immunostained with antibodies to NT and iNOS.

Pulmonary function studies were done in the pulmonary function laboratory at the UAB Kirklin Clinic. Standard clinical techniques were used. FEV₁ was the primary indicator of airflow obstruction.

Nitrotyrosine and Chlorotyrosine HPLC Assay

We used HPLC with electrochemical (EC) detection to assay percent nitration (%NT = nitrated tyrosines/total tyrosines × 100) and chlorination (%CT = chlorinated tyrosines/total tyrosines × 100) of tyrosine in BALF proteins (16). Proteins and supernatants of biologic fluid clarified by high speed centrifugation were precipitated using 10 volumes of ethanol or acetone. Solutions were placed at 40° C for 10 min and precipitates gently pelleted by centrifugation (3,000 × g for 5 min). BALF proteins were resuspended and hydrolyzed with 6 M hydrochloric acid (HCl). Hydrolysis mixtures were heated to 110° C for 16 h. Hydrolyzed mixtures were ultrafiltered, diluted 1:2 in HPLC buffer (50 mM sodium acetate at pH 4.7), and injected.

Chromatographic separation was achieved using a high resolution 4.6 × 250 mm TOSO Haas C-18 HPLC column. A characteristic voltamagram (current as a function of applied potential) was obtained for each compound (NT, CT, or tyrosine) by ramping the potential in small increments near its maximum response potential. Tyrosine was measured by comparing the HPLC peak area with a standard containing 500 pmol of tyrosine, 250 pmol of CT,and 250 pmol of NT. NT and CT retention times and channel ratios were based on injections of authentic commercial standards at concentrations similar to those measured in BALF samples. Using either peak height or area we observed the linear range extending down to an absolute concentration of 0.5 pmol for NT and 0.1 pmol for tyrosine. NT or CT concentrations were normalized to total tyrosines (Tyr) in the hydrolysates.

Nitrite and Nitrate Assays

We measured nitrite (NO₂) alone (before reduction of the samples) and total nitrite (i.e., nitrite plus enzymatically reduced nitrate) (NO₂ + NO₃) using the Griess reaction (17). Samples and standards were treated with nitrate reductase at 37° C for 3 h, a time sufficient to convert all nitrate (NO₃) to measurable nitrite (NO₂). One percent sulfanilamide and 0.02% (1-Naphthyl)ethylenediamine (NEDA) were added. Absorbance at 550 nm was read and sample concentrations calculated from standards., Nitrate concentration was then obtained by subtracting the value before reduction (nitrite) from the value after reduction (nitrite plus nitrate).

Protein

Protein concentration of BALF was assayed by the Bio-Rad (Coomassie blue) assay (Bio-Rad Laboratories, Hercules, CA). Bovine serum albumin (BSA) was the standard.

Urea

Concentration of urea in BALF was used to estimate dilution of ELF by the lavage fluid (18). The calculated dilution factor was used to estimate concentrations of molecules of interest (e.g., NT) in ELF. A blood sample was obtained on the day of bronchoscopy for measurement of serum urea. Urea concentration in BALF and serum was measured using a UV assay kit from Sigma Chemical Co. (St. Louis, MO). Urea concentration was determined by using urease to convert urea to ammonia and CO₂. Ammonia reacts with -ketoglutarate to form glutamic acid while consuming NADH. The rate of disappearance of NADH was followed at 340 nm.

Immunohistochemistry

Lung tissues were fixed in 4% paraformaldehyde in 0.1 M potassium phosphate buffer at pH 7.4. Blocks were ethanol-dehydrated and embedded in paraffin. Lung sections were immunostained with a polyclonal IgG antibody to NT or iNOS using the avidin-biotin-peroxidase method (19). Endogenous peroxidase activity was inactivated effectively by 0.3% hydrogen peroxide. To block nonspecific binding, sections were incubated with 10% nonimmune serum for 1 h. They were incubated with the primary antibody overnight. After three washes in phosphate buffer, sections were incubated with respective second antibody (biotin-conjugated goat antirabbit IgG). This was followed by further incubation with avidin-biotin-peroxidase complex (Elite Kit; Vector Laboratories, Burlingame, CA). Slides were developed with 3,3'-diaminobenzidine hydrochloride (DAB) and lightly counterstained with hematoxylin. Control experiments included absorption of the primary antisera with a nitrated tripeptide, 100 µM glycyl-nitrotyrosyl-alanine (NT), or incubation with nonimmune serum instead of primary antisera (iNOS).

Measurement of Minimum Surface Tension

Surfactant (i.e., lipids and lipid-soluble apoproteins) were isolated from BALF of four patients and four normal control subjects using the method of Bligh and Dyer (20). Fluid was centrifuged (150 × g for 15 min) to remove cellular debris. Phospholipids were sedimented by centrifugation at 27,000 × g for 30 min and then resuspended in saline. The suspension was layered over 0.8 M sucrose/0.9% NaCl and centrifuged at 8,500 × g for 30 min. The pellicle was removed and centrifuged at 27,000 × g for 30 min. Total phosphorus was measured in the pellet as described by Ames and Dubin (21). The surfactant mixture was resuspended at 1 mg/ml.

The surface tension was measured during dynamic compression using a pulsating bubble surfactometer (Electronetics Corp., Buffalo, NY). Pressure drop across the interface of a small air bubble formed in the surfactant was determined. The bubble was pulsed at 20 cycles/ min between minimum and maximum radii of 0.40 and 0.55 mm (50% surface area compression). Surface tension was calculated from the law of Laplace. Oscillation continued for 15 min or until the surface tension was < 3 mN/m (usually < 5 min) (11).

NT ELISA

BALF samples were concentrated using CentriPrep centrifugal concentrators (Amicon, Inc., Beverly, MA). Nitrated bovine serum albumin (NT-BSA) standard was made by incubating fatty-acid-free BSA in 50 mM potassium phosphate at pH 8.0 with 500 µM tetranitromethane (TNM). TNM-treated BSA was dialyzed to remove excess TNM. Absorbance at 430 nm (1 mg/ml solution) was measured. The concentration of NT (per mg protein) was calculated from the extinction coefficient for NT,  ₄₃₀ = 4,400 M/cm (22).

ELISA plates were coated with standards (80 to zero ng BSA) and 1:2 serial dilutions of BALF proteins (1,000 ng maximum), using 50 mM Na₂CO₃-NaHCO₃ buffer at pH 9.6. Nitrated standards (NT-BSA) and BALF proteins were reacted with polyclonal IgG rabbit antibody to NT (1:5,000 in diluting buffer). After incubation, wells were washed with PBS. The detection antibody (HRP-conjugated goat antirabbit IgG) in diluting buffer (1:2,500) was added and incubated for 1 h. Color was developed by adding o-phenylediaminedihydrochloride (OPD) and H₂O₂ to each well. Absorbance was read at 490 nm in a microplate reader (Bio-Rad) (22).

The change in molar concentration of NT per absorbance unit (pmol NT/Abs) was calculated as slope of the standard curve. Change in absorbance per mass of BALF protein (Abs/mg pro) was calculated from each sample's slope. Multiplication of the two slopes yielded concentration of NT (pmol) per mass of BALF protein (mg protein).

Data Analysis

Data that could identify patients were kept confidential. Data reported in the text are presented as medians (25th-75th % confidence interval) or arithmetic means ± SD.

When two independent groups were compared, the Mann-Whitney rank sum or unpaired t test was used depending on distribution of the data. Two-way ANOVA and t tests corrected for multiple comparisons were used to compare the surface tension measurements (23).

Spearman's rank correlation was used to test association among prospectively chosen variables, including NT, BALF protein, total nitrite plus nitrate nitrite, nitrate, histology scores, and days after transplantation (23, 24).

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient and Control Group Descriptions

Controls. The control group consisted of four healthy, nonsmoking volunteers. Mean age of the volunteers was 32 ± 6 SD yr (range, 28 to 41 yr).

HPLC data group. Patients (n = 29) from whom samples for HPLC were obtained were 52 ± 12 SD yr of age. The study population consisted of 14 men and 15 women and included patients who had either single (n = 19) or double (n = 9) lung transplants and one who had heart-lung transplant. Samples were obtained at a range of times after lung transplantation: median, 293 (83-820) d. None of the patients smoked at the time of the study.

ELISA data group. Mean age of the patients was 52 ± 10 SD yr (range, 23 to 66 yr). Samples were obtained 345 (88- 732) d after transplantation. Forty-five patients had single lung transplantation, 31 patients had double lung transplantation, and six patients had heart-lung transplantation. None of the patients smoked at the time of the study.

Indications for transplantation (in order of frequency) included chronic obstructive pulmonary disease, idiopathic pulmonary fibrosis, -antitrypsin deficiency, Eisenmenger's syndrome, bronchoalveolar cell carcinoma, cystic fibrosis, sarcoid, primary pulmonary hypertension, bronchiectasis, lymphangioleiomyomatosis, and histiocytosis X.

HPLC Assay of NT and CT in BALF

A highly specific EC-HPLC technique was used to measure NT, CT, and total tyrosines in BALF from patients and volunteers. Results are presented as percentages of total tyrosines nitrated or chlorinated. The data shown here are calculated as %NT = (nitrotyrosines/total tyrosines) × 100 and % CT = (chlorotyrosines/total tyrosines) × 100. The HPLC assay showed that NT was present in BALF from both normal control subjects: 0.01 (0.01-0.04)% and patients: 0.01 (0.00- 0.02)%. Although group medians were the same, high NT values in the patients (range, 0.00 to 1.08%) clearly exceeded the control range (0.01-0.06).

No CT was detected in the BALF from normal control subjects. A wide range of CT concentrations as found in the patients: 0.01 (0.00-0.02)%. However, many patients had no detectable CT. These data are summarized in Figure 1.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Percent nitration and chlorination of proteins in BALF. Data are displayed as box-whisker plots (left), including median, 5th, 25th, 75th, and 95th percentiles as well as mean ± SD (right) for the patient (n = 29) and the control (n = 4) groups analyzed by HPLC. HPLC assay confirmed that NT was present in BALF from both normal control subjects (con) (0.02 ± 0.03%) and patients (pts) (0.14 ± 0.28%). Several high NT values in the patients clearly exceeded these in the control subjects, but the distribution of the data precluded finding a significant difference. No CT was found in control BALF. A wide range of CT concentrations was found in the patients (0.11 ± 0.48%), although some patients had no detectable CT.

Urea Assay and Calculated Dilution of ELF

We used the method of Rennard and colleagues (18) to measure ELF dilution by simultaneous measurement of serum and BALF concentrations. We found a range of ELF dilutions in both control subjects and patients. The calculated average dilutions were: control subjects, 188 ± 180-fold; patients, 138 ± 104-fold (p > 0.05). Average dilution factors were used to estimate ELF total nitrite and NT concentrations.

Nitrite and Nitrate in BALF and Serum

We used the Greiss reaction to assay total nitrite (nitrite plus nitrate after enzymatic reduction) and nitrite alone (before enzymatic reduction) concentrations in BALF and serum. This assay measures nitrite only (19), so this approach permits calculation of nitrate and nitrite concentrations separately. The detection limit of the assay is < 1 µM.

Significantly higher total nitrite plus nitrate concentrations were present in BALF of the transplant patients. The concentration of total nitrite in BALF was 3.3 (1.9-5.1) µM compared with 1.3 (0.8-1.3) µM in the control subjects (p < 0.05). The serum total nitrite was 37 (26-55) µM compared with 19 (17) µM in the control subjects (p < 0.05). Total nitrite data are summarized in Figures 2 and 3. Nitrate and nitrite concentrations are shown individually in Table 1. Based on calculation of the average BALF dilutions in each group, we estimate the ELF total nitrite concentration to be ~ 188 µM in the control subjects and ~ 531 µM in the patients.

View larger version (15K): [in this window] [in a new window]   Figure 2.   BALF total nitrite concentrations. Data are displayed as box-whisker plots (left), including median, 5th, 25th, 75th, and 95th percentiles as well as the mean ± SD (right) for the patient (n = 82) and the control (n = 4) groups analyzed by ELISA. Data shown are BALF concentrations of total nitrite (the sum of nitrite plus enzymatically reduced nitrate). The detection limit of the assay is around 1 µM. The normal control subjects had low concentrations of total nitrite in BALF. Total nitrite was present in significantly higher concentrations in BALF from allotransplant patients (*p = 0.0133 compared with control subjects by rank sum test).

View larger version (13K): [in this window] [in a new window]   Figure 3.   Serum total nitrite concentrations. The format is identical to that in Figure 2. Data are displayed as box-whisker plots (left), including median, 5th, 25th, 75th, and 95th percentiles as well as the mean ± SD (right) for the patient (n = 82) and the control (n = 4) groups analyzed by ELISA. Data shown are serum concentrations of total nitrite (the sum of nitrite plus enzymatically reduced nitrate). The normal control subjects had detectable total nitrite in serum, most of which was present as nitrate (see Table 1). A significantly higher concentration of total nitrite was present in serum from allotransplant patients (*p = 0.0037 compared with control subjects by signed ranks test).

Immunohistochemistry

We stained available tissue blocks from five patients for NT and iNOS using a polyclonal IgG antibody to NT developed by Beckman and colleagues (19) or a commercial antibody to inducible nitric oxide synthase (iNOS). Immunostaining showed variable, patchy distribution of NT in cells and structures. In some sections airway or alveolar epithelial cells showed NT and iNOS costaining. The pleura was nitrated in many sections. Alveolar macrophages occasionally stained positive for NT. A representative sample is shown in Figure 4.

View larger version (113K): [in this window] [in a new window]   Figure 4.   Immunohistochemical stains showing nitrotyrosine and iNOS staining in an open lung biopsy. The figure shows, as a representative example, a section taken from a transplanted lung open biopsy stained with hematoxylin-eosin (panel a), or incubated with a polyclonal IgG antibody to NT (panel b) or iNOS (panel c). In this example, positive staining (brown color development), indicating the presence of NT or iNOS, is detected in epithelial cells lining a remodeled alveolar region (arrows). The NT antibody reaction was fully blocked by preincubation with a nitrated tripeptide (100 µM glycyl-nitrotyrosyl-alanine) confirming specificity of the NT reaction (panel d ).

Surface Tension Measurements

We did surface tension measurements on fluids from four patients, who had high BALF NT by ELISA. The mean concentration of NT in BALF from these four patients was 72 ± 30 pmol/mg (p < 0.05 compared with the control mean).

Surfactant was isolated by ultracentrifugation from BALF of the lung allotransplant patients and the four control subjects. Lipids were resuspended at 1 mg/ml and surface-tension-lowering assay by bubble surfactometry. The surfactant from the patients did not reach a low minimum surface tension (as occurred with the normal surfactant), suggesting functional impairment of the surfactant system. Data are shown in Table 2.

BALF NT Concentration by ELISA

A sensitive, previously validated ELISA (24) was used to assay immunoreactive NT in lavage fluid. ELISA detected NT in lavage fluids from all control subjects and from some patients. The concentration of immunoreactive NT did not differ significantly in BALF from either patients or normal control subjects. The concentration of NT in BALF from the volunteers was 28 (26-33) pmol/mg pro. The concentration of NT in BALF from the transplant patients was 9 (0-41) pmol/mg pro (p > 0.05). These data are summarized in Figure 5. Based on calculations of the average BALF dilution in each group, we estimate the ELF NT concentration to be ~ 11 nmol/mg protein in the control subjects and ~ 6 nmol/mg protein in the patients.

View larger version (12K): [in this window] [in a new window]   Figure 5.   Nitrotyrosine (NT) in BALF assayed by ELISA. The format is identical to that in Figures 2 and 3. Data are displayed as box-whisker plots (left), including median, 5th, 25th, 75th, and 95th percentiles as well as the mean ± SD (right) for the patient (n = 82) and control (n = 4) groups analyzed by ELISA. The figure shows the BALF concentration (pmol/mg protein) of immunoreactive nitrotyrosine measured by ELISA. NT concentrations in the patients were not significantly greater than those in the control subjects, although some transplant samples contained extremely high NT concentrations.

BALF NT concentrations in the patient samples correlated significantly with protein concentrations, a marker of airway inflammation (p = 0.000, r = 0.368 by Spearman's rank order correlation). BALF NT concentrations in the patient samples also correlated significantly with airway inflammation assessed by transbronchial biopsy histologic scoring (p = 0.0396, r = 0.209). The data are illustrated graphically in Figure 6. Although FEV₁ appeared to decrease with increasing BALF NT concentration (p = 0.0577, r = 0.157 by Spearman's rank order correlation), the association did not reach statistical significance.

View larger version (12K): [in this window] [in a new window]   Figure 6.   Correlation of BALF NT concentration and airway inflammation assessed histologically. The figure shows degree of airway inflammation (categorized as a score from 0 to 4 [B score]) as a function of BALF NT concentration. The calculated linear regression line with 25th and 75th percentile confidence limits is shown to illustrate the relationship. BALF NT concentrations in the patient samples tended to increase significantly with the degree of airway inflammation assessed by the histologic scoring of the transbronchial lung biopsies (p = 0.037, r = 0.209 by Spearman's rank order correlation).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Significance of RNS Metabolites in ELF

We sampled ELF from normal control subjects and lung transplant patients. Fluid was analyzed for metabolites of RNS and protein nitration because nitrogen oxides (NO_x) are important mediators of inflammation. NO_x react with other molecules to form bioreactive compounds, including peroxynitrite, nitrosamines, and S-nitrosothiols (25). Of note, we measured significantly higher concentrations of total nitrite in BALF from the patients than in BALF from the control subjects. Serum total nitrite levels were also significantly higher in the patients. Nitration of tyrosine, the most typical protein modification attributed to RNS, occurred in the ELF of both patients and control subjects. Chlorination of tyrosine, possibly resulting from reactions with hypochlorite (HOCl) or other MPO- derived product, was detected only in ELF from the patients.

By design this study collected data from all patients who underwent surveillance bronchoscopy after transplantation. The patients were not stratified prospectively, so descriptive data herein represent the entire population rather than a subset. However, because data were collected in the outpatient clinic, most of the patients were clinically stable and not acutely ill when samples were collected.

These findings are noteworthy for several reasons. First, they definitely confirm the existence of protein nitration (NT) and chlorination (CT) reactions in the alveolar region and airways. Second, immunohistochemistry data show that protein nitration occurs in some allografts selected for open lung biopsy. Third, total nitrite is quite concentrated in ELF of lung allografts, where it may reflect macrophage and epithelial cell ·NO production by nitric oxide synthase (NOS). Nitrite-mediated reactions might also participate in protein nitration and in ELF (10).

The total nitrite concentration we found in BALF from patients who had undergone lung allotransplantation (i.e., several hundred micromolar) is similar to that reported by O'Donnell and colleagues (26) in BALF from sarcoid patients (6.7 nmol/ml). Like them, we estimate (based on dilution of urea) ELF total nitrite concentration to be quite high. The total nitrite concentration we found in the transplant BALF probably was not due to diffusion of nitrate from the capillary space because anions do not readily cross intact endothelial and epithelial barriers.

Breakdown products of RNS, measured as nitrite plus nitrate, were significantly higher in the serum and BALF from the patients than from those of the control subjects. The increase in RNS implies that nitric oxide production is greater in the lung transplant population. Finding increased RNS products has potential clinical significance because inhibition of nitric oxide synthase modulates inflammation in models of lung transplantation (3).

Nitrotyrosine and Chlorotyrosine in ELF

NT in the ELF suggests that either epithelial cells or macrophages produce ·NO and that a source of superoxide is present. The role of ·NO production in the ELF is not fully defined. Host defense clearly is important because the ELF is in contact with the external environment. ·NO also may have an important role as a signal that regulates airway muscle tone. S-nitrosothiols in the airways have bacteriostatic and bronchodilator properties (25, 27). Neutrophils in the inflamed airways may participate in protein nitration through MPO-mediated mechanisms, especially in nitrite-rich compartments (10).

NT within the ELF of the normal subjects also suggests that some peroxynitrite is formed in the absence of inflammation. These findings are consistent with previous observations that ·NO is present in expired gas of normal subjects (25) and that nitrosothiols exist in normal BALF (27).

The ELISA developed by Zhu and colleagues (22) is sensitive in a picomole NT/milligram protein range. This ELISA is specific for NT, as confirmed by our ability to block the reaction completely by reduction of BALF proteins with 10 mM dithionite. To unequivocally quantify protein-bound NT and CT, we assayed %NT, %CT, and total tyrosines in protein hydrolysates by HPLC. We employed conditions that give complete hydrolysis (18). Conditions are optimized to eliminate contaminating nitrite and chloride, which could give rise to artifactual production of NT and CT. Hydrolysis conditions used do not cause chlorination of proteins, as shown by the lack of any chlorinated proteins in the BALF from the control subjects.

Percent nitration of tyrosines in the normal BALF was quite low (mean, 0.02%), but definitely detectable. On average, the percent tyrosine nitration was 7-fold higher (0.14%) in BALF from the patients. The failure to find significance is due to the distribution of data, as shown in Figure 1. (The power of the unpaired t test was 0.0500, indicating that a real difference not detected by this test may exist.) The highest percent nitration values clearly exceeded those of the normal subjects, showing that soluble BALF proteins were nitrated in some allograft patients.

Normalizing NT to total tyrosine accounts for both the variable amount of total protein in each sample and the efficiency of enzymatic hydrolysis or recovery after acid hydrolysis. Normalized values reported in the literature have ranged from 0.013 NT to 5.5 NT per 100 tyrosines (i.e., 0.013 to 5.5%) (reviewed in reference 16). Only a few specific proteins may be susceptible to nitration, and functional effects can result from nitration of only one tyrosine in a small fraction of protein molecules (28). Albumin is the most abundant protein in BALF, and therefore most likely nitrated and detected by ELISA. Both BALF albumin and NT reflect airway inflammation, and they tended to increase together. An equal amount of protein was present in each ELISA well, so the NT absorbance should reflect the percent nitration. Indeed, no correlation exists between the %NT found by HPLC and the BALF protein concentration (p = 0.436, r = 0.161 by Spearman's rank correlation).

CT in the BALF of lung allotransplant patients (but not in that of the control subjects) suggests neutrophilic inflammation. CT is a specific product of MPO-catalyzed oxidative damage (29). CT may be formed by reaction of nitrite and hypochlorous acid, but HOCl can chlorinate without nitrite. A hypothetical intermediate, nitryl chloride (NO₂Cl), formed in the presence of nitrite also may nitrate and chlorinate phenolic amino acids (10). CT is a specific marker of hypochlorous acid production in vivo and for the involvement of MPO in inflammation (30).

Correlation of Nitrotyrosine and Functional Surfactant Impairment

Pulmonary surfactant from the four lung allotransplant recipients was less efficient at lowering surface tension than surfactant from the control subjects. Peroxynitrite (1 mM in vitro) inhibited calf lung surfactant extract from reaching minimum surface tension (below 10 mN/m) on dynamic compression (11). Inhibition was due to oxidation of lipid-soluble surfactant proteins. The in vitro concentration used in that study exceeds the likely concentration of ONOO in the lung. Another study recently has confirmed abnormal function of surfactant recovered from lung allotransplants (31). Although the mechanism is not known (i.e., it may be unrelated to protein nitration), a decrease in surfactant function is associated with increased immunoreactive NT in BALF from transplant recipients. This association does not show that specific protein nitration (e.g., surfactant protein A) is responsible for the decreased surfactant function.

Immunohistochemistry and Airway Inflammation

Immunohistochemical studies of some available lung biopsies showed that some proteins in pleura, alveolar walls, and perivascular tissues were nitrated, although no consistent pattern was discerned. Immunostaining of NT and iNOS appeared colocalized to the respiratory epithelium in some biopsies. Colocalization may simply reflect inflammation in that region, but iNOS has been implicated in many other models of inflammation. In contrast, Mason and colleagues (32) have recently found that normal control (unused donor) lungs showed no NT staining.

Within the patient population, the concentration of NT in the airway correlated significantly with the degree of airway inflammation visible on transbronchial lung biopsies and with BALF protein concentration, another marker of inflammation. The association of reactive nitrogen species metabolites and nitrotyrosine with airway inflammation is important because airway inflammation, usually attributed to rejection, is a key problem after lung transplantation (2).

These data reveal several new insights. The epithelial lining fluid appears to contain nitrate in very high concentration, and the concentration of RNS metabolites (measured as total nitrite) in the transplanted airways significantly exceeds that of the control subjects. Reactive nitrogen species thus appear to participate in transplant airway inflammation and may be related to development of obliterative bronchiolitis.