The human surfactant protein A (SP-A) locus consists of two functional genes (SP-A1, SP-A2) with gene-specific products exhibiting qualitative and quantitative differences. The aim here was twofold: 1) generate SP-A1 gene-specific antibody, and 2) use this to assess gene-specific SP-A content in the bronchoalveolar lavage fluid (BALF). An SP-A1-specific polyclonal antibody (hSP-A1_Ab68-88_Col) was raised in chicken, and its specificity was determined by immunoblot and ELISA using mammalian Chinese hamster ovary (CHO) cell-expressed SP-A1 and SP-A2 variants and by immunofluorescence with stably transfected CHO cell lines expressing SP-A1 or SP-A2 variants. SP-A1 content was evaluated according to age and lung status. A gradual decrease (P < 0.05) in SP-A1/SP-A ratio was observed in healthy subjects (HS) with increased age, although no significant change was observed in total SP-A content among age groups. Total SP-A and SP-A1 content differed significantly between alveolar proteinosis (AP) patients and HS, with no significant difference observed in SP-A1/SP-A ratio between AP and HS. The cystic fibrosis (CF) ratio was significantly higher compared with AP, HS, and noncystic fibrosis (NCF), even though SP-A1 and total SP-A were decreased in CF compared with most of the other groups. The ratio was higher in culture-positive vs. culture-negative samples from CF and NCF (P = 0.031). A trend of an increased ratio was observed in culture-positive CF (0.590 ± 0.10) compared with culture-positive NCF (0.368 ± 0.085). In summary, we developed and characterized an SP-A1 gene-specific antibody and used it to identify gene-specific SP-A content in BALFs as a function of age and lung health.
- alveolar proteinosis
- bronchoalveolar lavage fluid
- cystic fibrosis
- peptide antibody
- surfactant protein A
human surfactant protein a (SP-A), a member of the C-type lectin or collectin family, is encoded by two functional genes, SP-A1 and SP-A2, with a high percentage of sequence similarity detected at the nucleotide (94%) and the amino acid (aa) (96%) levels. More than 30 alleles have been characterized for these two genes, with four SP-A1 alleles (6A, 6A2, 6A3, and 6A4) and six SP-A2 alleles (1A, 1A0, 1A1, 1A2, 1A3, and 1A5) most frequently observed in the general population (10, 14, 16, 40, 67). Although a high level of sequence similarity is shared among the SP-A genes and their corresponding alleles, in vitro studies have identified both qualitative (22, 36, 57, 60, 77, 80, 81) and quantitative (24, 35, 38, 46, 56, 69, 70, 78, 79) differences between the two genes.
Qualitative differences between the SP-A1 and SP-A2 genes include functional, structural, and biochemical differences. Functionally, SP-A2 exhibits a higher level of activity than SP-A1 in its ability to enhance TNF-α and IL-8 production by the macrophage-like THP-1 cell line (80, 81), phagocytosis of Pseudomonas aeruginosa by rat alveolar macrophages (57), and secretion of phosphatidylcholine by type II alveolar cells (77). Structurally, circular dichroic spectroscopy studies have shown SP-A2 to be more stable than SP-A1 (22). Biochemical differences in response to ozone exposure (36, 77, 81), oligomerization patterns (36, 77), aggregation (71), and the ability to bind carbohydrates (60) were also observed for SP-A1 and SP-A2 variants. Quantitative differences between SP-A1 and SP-A2 genes were observed in H441 cells, a lung adenocarcinoma cell line (46, 70), and in fetal lung explant cultures under basal conditions and/or in response to dexamethasone (24, 38, 46, 51, 56). Quantitative differences in the total SP-A mRNA (18) or protein levels (25, 27) among individuals were also observed. Data obtained from constructs containing 5′-UTR (untranslated region) splice variants indicated that SP-A2 is more efficiently translated than SP-A1 (78). Thus sequence and splice variability differences (20, 40, 45, 67) within the 3′- and 5′-UTR observed between the SP-A1 and SP-A2 genes and/or alleles may contribute to the observed regulatory differences.
It has been reported that both SP-A1 and SP-A2 mRNAs are expressed in normal human epithelial alveolar type II cells with SP-A2 also being expressed in the submucosal glands of the conducting airways (43, 44, 69). Human SP-A is an octadecamer comprised of six trimers. Initial studies indicated that each trimer may consist of two SP-A1 molecules and one SP-A2 molecule suggesting a 2:1 ratio (75). However, at the mRNA level, the SP-A1-to-SP-A2 ratio has been shown to vary considerably (39). Although the SP-A1 and SP-A2 variants differ collectively at 10 aa, the gene specificity between SP-A1 and SP-A2 variants is determined by differences in 4 aa (66, 73, 81, 85) within the collagen region. Of the 10 aa differences between SP-A1 and SP-A2 variants (10, 16, 40), 2 aa are located in the NH2-terminal signal peptide, 6 in the collagen-like region, and 2 in the carbohydrate recognition domain. A cysteine at position 85 that is present in SP-A1 but not in SP-A2 (16, 20) is thought to be involved in SP-A oligomerization, forming intermolecular or intertrimeric disulfide bonds (11). Native SP-A undergoes several co- and posttranslational alterations, including cleavage of the signal peptide (63), proline hydroxylation (62), N-linked glycosylation (19), sialylation (19), and acetylation (20), giving rise to multiple isoforms with different isoelectric points.
Alterations in SP-A levels have been noted for several pulmonary diseases, including neonatal respiratory distress syndrome (RDS) (82) and acute RDS (ARDS) (1, 4), alveolar and airway infections (37), and respiratory syncytial virus (RSV) infection (41). Significantly higher levels of SP-A have been reported in alveolar proteinosis (AP) (34, 47, 72), sarcoidosis, and hypersensitivity pneumonitis patients (32). Decreased SP-A levels were observed in inflammatory lung diseases (68), patients with bacterial pneumonia (5), bronchopulmonary dysplasia infection in baboons (3), and interstitial pneumonia with collagen vascular disease (47). For idiopathic pulmonary fibrosis (IPF), both decreased (55) and increased levels (33, 48, 64) of SP-A have been reported. Reports for cystic fibrosis (CF) have shown decreased levels of SP-A in children (26, 65), increased levels associated with bacterial infection (59), and normal-to-increased levels with airway infection (37). In a recent study, a decrease in SP-A content was observed with increasing age in bronchial lavage, but not in alveolar lavage (7). Moreover, SP-A genetic variant associations with various lung diseases, including RDS (15, 66), RSV (52), IPF (71), and tuberculosis (17), have also been reported.
In the published literature, it is unclear whether the differences observed in SP-A levels between healthy individuals and subjects with pulmonary diseases reflect an alteration in the levels of one or both of the SP-A genes. This is an important point to address because qualitative and quantitative differences between in vitro-expressed SP-A1 and SP-A2 variants have been observed, as noted above. Therefore, it is likely that differences in the relative levels of each SP-A gene product, rather than the total SP-A protein content, are better indicators of the overall functional activity of SP-A and perhaps better predictors of disease specificity. To address this issue, we have produced and characterized an SP-A1 gene-specific polyclonal peptide antibody and used it to study the relative levels of SP-A1 and/or the ratio of SP-A1 to total SP-A as a function of age and health lung status.
MATERIALS AND METHODS
Production of SP-A1 Gene-Specific Antibody (IgY)
An SP-A1 gene-specific antibody was raised in chickens (Aves Lab, Tregard, OR). A 21-aa peptide was synthesized that included the collagen region of SP-A1 Cys68-Lys88, as shown in Fig. 1. The 21-aa peptide contained three SP-A1 gene-specific aa: Asp73, Ile81, and Cys85 at aa positions 73, 81, and 85 of SP-A1. The proline residues at positions 76, 79, and 82 are either hydroxylated or unhydroxylated. A mixture of hydroxylated and unhydroxylated peptide molecules was synthesized, and their purity was established by micro-HPLC analysis and MALDI mass spectrometric analysis. The SP-A1 gene-specific peptide was conjugated using m-maleimidobenzoyl-N-hydroxysuccinimide ester conjugation through thiol groups on cysteine residues. The Cys was then conjugated to keyhole limpet hemocyanin, and two chickens were immunized.
Preparation of Affinity Column
Two milligrams of SP-A1 peptide were coupled to the sulfo-linked coupling gel according to the manufacturer's instructions (Pierce, Rockford, IL). The availability of cysteine groups in the peptide was determined by Ellman's reagent (Pierce). Two milliliters of coupling gel were aliquoted into 5-ml columns and equilibrated with 8 ml of coupling buffer containing 50 mM Tris and 5 mM EDTA-Na (pH 8.5). Two milligrams of the reduced 21-aa peptide were dissolved in 2–3 ml of coupling buffer, mixed with the coupling gel, rotated end-over-end for 15 min, and then incubated for 30 min without mixing. Following incubation, binding efficiency was determined by comparing the unbound peptide to the total amount of starting peptide using Ellman's reagent and cysteine (Pierce) as a standard and reading absorption at 412 nm. The coupling gel was then washed with 6 ml of coupling buffer, blocked with 15.8 mg of l-cysteine-HCl for 15 min with end-over-end rotation, and incubated for 30 min. Columns were washed with 12 ml of 1 M NaCl and 0.05% NaN3, and a porous disc was inserted above the coupling gel to prevent drying.
Affinity Purification With Peptide Columns
Chicken egg yolk immunoglobulin (IgY) fractions (Aves Lab) for the peptide SP-A1 antibody were affinity purified using sulfo-linked columns. Columns were equilibrated with 8 ml of sodium phosphate buffer (pH 6.8), and the IgY fraction was prepared in sodium phosphate buffer (pH 6.8), passed over the column, and incubated for 1 h at room temperature (RT). Columns were then washed with 12 ml of phosphate buffer and eluted with 8 ml of 100 mM glycine (pH 2.5–3) buffer, and 1-ml fractions were collected and neutralized with 50 μl of 1 M Tris (pH 9.0). Protein concentrations were determined by UV spectrophotometer at 280 nm, and the high-protein concentration fractions were pooled and dialyzed against phosphate buffer (pH 7.2). Protein concentration was verified again and reconstituted to a final concentration of 1 μg/μl.
Quantification of Sulfhydryls Using Ellman's Reagent
Column binding efficiency and the presence of cysteine groups within the peptide were determined using Ellman's reagent (Pierce). Four milligrams of Ellman's reagent were dissolved in 1 ml of reaction buffer (0.1 M sodium phosphate solution, pH 8.0, containing 1 mM EDTA). Cysteine hydrochloride monohydrate (Pierce) diluted in the reaction buffer was used as the standard (0.1–1.5 mM). Tubes containing 10 μl of Ellman's reagent solution and 500 μl of reaction buffer were prepared, and 50 μl of each standard or unknown was added. Samples were vortexed and incubated for 15 min at RT, and the absorbance was measured at 412 nm.
Cell Lines and Cell Culture Conditions
The mammalian CHO-K1 cell line (American Type Culture Collection, Manassas, VA) was used for expression of the mammalian SP-A variants. Cells were cultured in Glasgow's modified Eagle's medium (GMEM; Invitrogen, Carlsbad, CA) with 10% FBS (dialyzed and heat inactivated at 56°C for 1 h) at 37°C, with 5% CO2, as previously described (81). In brief, the stably transfected cell lines were grown in glutamine-free GMEM plus 25 μM methionine sulfoximine (MSX). SP-A variants were expressed under tissue culture conditions in medium containing 0.5 mM ascorbic acid without serum and harvested after 5 days.
Purification of the In Vitro-Expressed SP-A Variants and Human SP-A From Bronchoalveolar Lavage Fluid
In vitro-expressed SP-A variants.
SP-A was recovered from an average of 200 ml of serum-free culture medium (2 × 107 cells per plate/10 ml) after 5 days. CHO cells and debris were removed by centrifugation at 1,000 g for 10 min at 4°C, the supernatant was collected, and calcium concentration was adjusted to 2 mM Ca2+ with 1 M CaCl2 and passed through a mannose-sepharose 6B column (Sigma, St. Louis, MO). The column was washed with 200 ml of 5 mM Tris (pH 7.5) and 2 mM CaCl2, and the SP-A was eluted with a buffer containing 5 mM Tris and 2 mM EDTA (pH 7.5), collected in 2.5-ml fractions, and dialyzed against 0.5 M Tris (pH 7.5). All purification steps were carried out at 4°C.
Native Human SP-A From Bronchoalveolar Lavage Fluids of AP and Healthy Subjects
Bronchoalveolar lavage fluids (BALFs) were obtained from AP patients undergoing therapeutic lavage and from nondiseased lungs obtained through collaboration with The Gift of Life Donor Program (Philadelphia, PA). The protocol was approved by the Penn State College of Medicine Institutional Review Board. The lungs of AP patients were lavaged with several liters of 0.9% saline (10–15 liter/lung). The first lavage of ∼1 liter was frozen and shipped to Penn State for SP-A isolation from the Cleveland Clinic Foundation. Each donor lung (routinely the left lung) was lavaged with 1 liter of 0.9% saline, and the lavage fluid was collected for purposes of this study. BALFs were centrifuged at 150 g for 10 min to obtain cell-free BALF, and human SP-A was purified using the butanol extraction method, as previously described (31).
Quantitative ELISA Analysis With SP-A1-Specific Antibody (IgY)
Following informed consent, the CF and noncystic fibrosis (NCF) BALFs were collected from children undergoing diagnostic bronchoalveolar lavage. The protocol was approved by the Penn State College of Medicine Institutional Review Board. The CF patients had various types of positive cultures, with the majority being P. aeruginosa. However, other organisms, including Staphylococcus aureus, Candida species, and others were isolated. The NCF patients were very heterogeneous and had many types of infections. These patients were representative of a wide variety of pediatric patients who undergo bronchoscopy for clinical reasons. The BALF samples were collected and immediately centrifuged, aliquoted, and frozen at −20°C. Total cell and differential cell counts, as well as the assessment of bacteria, fungi, or viruses, was performed for each sample.
ELISA conditions were optimized for SP-A1 and SP-A antibodies so that the ratio of SP-A1/SP-A was equal to 1 when in vitro-expressed SP-A1 (6A2) proteins and total SP-A from BALF of known concentrations were used as standards. Plates coated with 100 μl/well of antigen, made from a mixture of 5 μl of BALF [healthy subjects (HS), AP, CF, and NCF] diluted in 995 μl of coating buffer, were incubated overnight at 4°C. Plates were then washed four times with PBS/0.5% Tween and blocked in the same solution for 1 h at RT. One-hundred microliters of antibody at a concentration of 10 μg/ml for the SP-A1 or at 33 ng/ml for the SP-A antibody were added per well and incubated at 37°C for 1 h. Following incubation, plates were washed four times with PBS/0.5% Tween. A secondary antibody of either goat anti-chicken horseradish peroxidase (HRP) conjugate for SP-A1 or goat anti-rabbit IgG HRP conjugate for SP-A at a 1:2,500 dilution was added to the respective plates and incubated at 37°C for 1 h. Following a final wash with PBS/0.5% Tween, plates were color developed with O-phenylene diamine dihydrochloride, the reaction was stopped with 2.5 M sulfuric acid, and optical density was measured at 490 nm.
Immunoblot of SP-A1 and Total SP-A
SDS-PAGE was performed as previously described (49). In vitro-expressed mammalian cell SP-A variants and human SP-A from BALF were subjected to electrophoresis under reducing conditions. The samples were dissolved in buffer containing 0.1 M DTT, 1.0 M 2-mercaptoethanol, 2% SDS, 0.1 M Tris·HCl (pH 6.8), and 10% glycerol, denatured at 95°C for 5 min and run on a 10% polyacrylamide gel at 90 V for 1 h (Bio-Rad, Hercules, CA). Proteins were transferred from the gel onto 0.2-μM pore size polyvinylidene difluoride membranes (Bio-Rad), blocked for 1 h with 3% nonfat dry skimmed milk in Tris-buffered saline (TBS), and washed with TBS containing 0.05% Tween. Blots were incubated overnight with a concentration of 10 μg/ml SP-A1 antibody (IgY) and 50 ng/ml human SP-A antibody. Goat anti-chicken (IgY) HRP-conjugated secondary antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) for SP-A1 at 1:5,000 dilution and goat anti-rabbit (IgG) HRP-conjugated secondary antibodies (Santa Cruz Biotechnology) for SP-A at 1:5,000 dilution were used. Blots were exposed to X-ray film (Kodak, Rochester, NY), and antibody binding was localized by enhanced chemiluminescent detection.
Immunofluorescent Detection of SP-A
Stably transfected CHO mammalian cells, SP-A2 (1A0) and SP-A1 (6A2), were grown to 80% confluence in 3 ml of GMEM with MSX medium in six-well culture plates on a cover slip. The medium was aspirated, the cover slips were washed three times with PBS, and the cells were fixed with 1% paraformaldehyde for 10 min followed by three washes for 5 min with cold 1× PBS. Blocking was performed using 10% normal goat serum for 1 h at RT. Primary antibody, using 100 μl of 20 μg/ml of SP-A1 antibody and 4 μg/ml SP-A antibody was added to each cover slip and incubated for 1 h. Following the incubation, the cover slips were washed three times for 5 min each with PBS containing 0.2% Tween, and the appropriate FITC-labeled secondary antibody (100 μl) was added and incubated for 1 h. Cover slips were then washed three times for 5 min each with PBS containing 0.2% Tween, and the cover slips were mounted with Aqua Mount (Polysciences, Warrington PA). Slides were observed and photographed using a Leica TCS SP2 AOBS confocal microscope.
Statistical analysis was performed using Student's t-test or the Mann-Whitney rank sum test. Data were calculated as means ± SE from duplicates within each experiment and from three independent experiments. Nonparametric data were analyzed by Kruskal-Wallis one-way analysis of variance on ranks, with between-group differences determined by Dunn's method. Differences were considered significant when P < 0.05.
Generation, Purification, and Specificity of Chicken SP-A1 Gene-Specific Antibody (IgY)
The generation and purification of the SP-A1 (hSP-A1_Ab68-88_Col) antibody was carried out as described in materials and methods. Gene specificity was confirmed by Western blot analysis, ELISA, and immunofluorescence.
Western Blot Analysis
In vitro-expressed SP-A variants, SP-A1 (6A2) and SP-A2 (1A0), and human SP-A (hSP-A) from bronchoalveolar lavage of AP patients, were used for Western blot analysis (Fig. 2). The SP-A1 antibody successfully recognized the in vitro-expressed SP-A1 (6A2) protein in a concentration-dependent manner, as well as the hSP-A from bronchoalveolar lavage, but did not recognize the SP-A2 (1A0) in vitro-expressed variants at any SP-A concentration, indicating SP-A1 gene specificity (Fig. 2A). The hSP-A antibody recognized both the SP-A1 and SP-A2 in vitro-expressed variants and the hSP-A from BALF, as expected (Fig. 2B). Specificity of SP-A1 antibodies was also shown for all the SP-A1 variants (6A2, 6A4, and 6A), and no immunoreactivity was observed with SP-A2 variants (1A0, 1A1, and 1A) (Fig. 2C).
In vitro-expressed SP-A1 (6A2, 6A4, 6A) and SP-A2 (1A0, 1A1, 1A) variants were used for ELISA (Fig. 3) with 50 ng/well of each protein serially diluted. The SP-A1 antibody recognized all SP-A1 in vitro-expressed variant proteins (6A2, 6A4, 6A) but failed to recognize any of the SP-A2 variant proteins (1A0, 1A1, 1A) (Fig. 3A). As expected, the SP-A antibody recognized both SP-A1 and SP-A2 in vitro-expressed variants (Fig. 3B). Collectively, these data indicate that the SP-A1 antibody is gene specific.
To further confirm SP-A1 gene specificity, ELISA was performed using different ratios of SP-A1 to SP-A2 (Table 1, Fig. 4). SP-A1 protein was detected with the SP-A1 gene-specific antibody in a concentration-dependent manner (Fig. 4A) regardless of the presence of the in vitro-expressed SP-A2 protein. The SP-A antibody in Fig. 4B (as expected) shows comparable reactivity for all SP-A1:SP-A2 combinations. Figure 4C shows immunoreactivity with SP-A1 antibodies at 50 ng/100 μl of mixed proteins for different dilutions of SP-A1 (6A2):SP-A2 (1A0) ranging from 1:100 to 100:1, as shown in Table 1.
Immunofluorescent Detection of SP-A1
Immunodetection of the SP-A1 protein in stably transfected CHO mammalian cells was performed by immunofluorescence on CHO cells expressing SP-A1 (6A2) and SP-A2 (1A0) variants (Fig. 5). Figure 5A shows specific staining by the SP-A1 antibody with cells expressing SP-A1 (6A2) variants but not with the cells expressing SP-A2 (1A0) variants, indicating that the antibodies are SP-A1 gene specific. In contrast, the SP-A antibody staining presented in Fig. 5B shows (as expected) a positive staining with both cells expressing SP-A1 (6A2) and SP-A2 (1A0). Similar results were obtained with cells expressing SP-A1 (6A4) and SP-A2 (1A1).
Determination of SP-A1/SP-A Ratios and/or Levels of SP-A1 in the BALF of HS, AP, CF, and NCF
As a function of age.
Individuals without known lung injury were grouped into four age categories (0–20, 21–40, 41–60, and 61–80 years) as shown in Fig. 6. A gradual but significant decrease in the ratio of SP-A1 to SP-A (0.33 ± 0.06 > 0.22 ± 0.06 > 0.20 ± 0.07 > 0.12 ± 0.04) was observed from the youngest to the oldest age group (P < 0.05), respectively. Multiple comparison procedures (Dunn's method) revealed a significance (P < 0.05) between the 0–20 (0.33 ± 0.06) and the 41–60 year groups (0.20 ± 0.07; see Fig. 6A). No significant differences were observed in either total SP-A or SP-A1 content among age groups (Fig. 6B).
As a function of lung health status.
The gene-specific antibodies described above were used for ELISA to determine the ratio of SP-A1-to-SP-A and SP-A1 levels in BALF samples in different subject groups (Fig. 7). A total of 47 samples from HS were compared with 17 AP patients, 16 children with CF, and 35 NCF children. The HS group contained a full range of age, and the CF and NCF groups were pediatric patients. The AP group contained mostly adults and only one pediatric patient. Variability in the levels of SP-A1 in the BALF samples was observed among HS, AP, CF, and NCF patients.
The ratio of SP-A1 to SP-A was higher in CF (0.49 ± 0.09) compared with AP (0.21 ± 0.07, P = 0.027) or HS (0.21 ± 0.04, P = 0.002) (Fig. 7A), even though total SP-A was lower in CF compared with all other groups and SP-A1 was lower in CF compared with AP (Fig. 7B). The ratio of SP-A1 to SP-A in NCF (0.30 ± 0.05) was lower compared with CF (P = 0.046) (Fig. 7A). No differences in the ratio were observed between AP (0.21 ± 0.07) and HS (0.21 ± 0.04), although the total levels of SP-A or SP-A1 between the two groups differed significantly (P = 0.001, Fig. 7B).
The SP-A1 levels in AP patients (25.92 ± 4.93) were significantly higher compared with HS (7.44 ± 1.54, P = 0.01), CF (4.87 ± 1.04, P = 0.04), and NCF patients (6.02 ± 1.44, P = 0.01). The SP-A level was also noted to be significantly higher (P = 0.001) in AP patients (139 ± 27.34) compared with HS (52.26 ± 8.33), CF (15.73 ± 5.18), and NCF patients (32.19 ± 6.17). Higher SP-A levels were observed for HS compared with CF (P = 0.001) and CF compared with NCF (P = 0.025) (Fig. 7B).
As a function of bacterial cultures.
The ratio of SP-A1 to SP-A in BALF samples from pediatric CF and NCF patients was evaluated as a function of bacterial cultures. The SP-A1-to-SP-A ratio was significantly higher in the culture-positive samples (0.459 ± 0.067, P = 0.031) compared with culture-negative samples (0.258 ± 0.056) (Fig. 8A). As shown in the Fig. 8B, no significant difference in the level of either total SP-A or SP-A1 content was found between the culture-positive CF-NCF group and the culture-negative CF-NCF group. Although the ratio was also higher in CF culture-positive (0.59 ± 0.10) compared with NCF culture-positive (0.37 ± 0.07) groups (Fig. 8C), this difference did not reach statistical significance (P = 0.106). Also, no significant difference exists in either total SP-A or SP-A1 content in CF and NCF culture-positive groups (Fig. 8D).
The SP-A content of BALF has been shown to vary in individuals with various lung diseases (3–5). Moreover, qualitative and quantitative differences have been observed in vitro between the SP-A1 and SP-A2 genes (22, 36, 57, 60, 77–81), pointing to the possibility that the overall functional activity of SP-A may depend on the relative SP-A1-to-SP-A2 levels rather than on the total SP-A content. To address this possibility, it was essential to develop and characterize gene-specific antibodies. In this report, we have successfully produced an SP-A1-specific antibody and used this antibody along with an SP-A polyclonal antibody, to measure SP-A1 protein content and the ratio of SP-A1 to total SP-A as a function of age or lung infection status. The SP-A1 antibody specificity was assessed by Western blot analysis and ELISA using in vitro-expressed SP-A1 and SP-A2 variants and by immunofluorescence using stably transfected CHO cells expressing SP-A1 and SP-A2 variants. Moreover, using the SP-A1 gene-specific antibodies, we were able to show differences in the SP-A1-to-SP-A ratio in the study groups (HS, AP, CF, and NCF) under investigation. The SP-A1 and total SP-A content may differ in the various groups, and, in some cases, such as CF, even though the ratio was increased, total SP-A was significantly lower compared with some of the other groups, and SP-A1 was lower compared with AP. The SP-A1-to-SP-A ratio decreased as a function of age and increased in individuals with bacteria-positive culture or disease (i.e., CF) associated with bacterial infection. These data indicate that SP-A1 and/or presumably SP-A2 are regulated by factors that may be influenced by age and/or the lung microenvironment. We speculate that the relative levels of SP-A1 and SP-A2 rather than the total SP-A content better reflect the overall functional activity of SP-A that, in turn, may be more predictive of disease specificity.
Raising SP-A gene-specific antibodies has been rather challenging due to high level of sequence similarity between the two genes and perhaps most importantly due to the location of the “core” aa (40) that distinguish the two gene products. These are located within the collagen-like domain of SP-A, a relatively compact, collagenous, triple helical structure that may impede exposure of critical epitopes. The successful generation of SP-A1 antibodies capable of recognizing human SP-A1-specific proteins under reduced and native conditions may be due primarily to a key aa difference between SP-A1 and SP-A2, namely residue 85. The primary structure of SP-A in the collagenous domain consists of 23 Gly-X-Y repeats, where Gly represents glycine and X or Y represents any other aa residue. However, an important difference exists between SP-A1 and SP-A2 in Gly-X-Y repeat number 18 (16, 22). SP-A1 has a Cys in the Y position (GEC) referred to as Cys85, whereas the SP-A2 has an Arg (GER) referred to as Arg85. Studies of host/guest triple helical peptides, where a single guest Gly-X-Y with substitutions in X and Y positions was studied while embedded in a collagen host peptide, showed that an Arg in the Y position associates with high stability of the host/guest peptide (9, 61, 73). Based on these studies, we have previously proposed (21, 22) that Arg85 (SP-A2) has a higher local microstability compared with Cys85 (SP-A1). This may result in local “microunfolding” and exposure of SP-A1-specific epitopes. Moreover, differences in thermal stability (22), degree of hydroxylation (16, 21, 54), and differences in the oligomerization patterns of SP-A1 and SP-A2 (77) may contribute to the differential exposure of potential epitopes and together may explain our success in generating SP-A1-specific antibodies.
Proline hydroxylation played a key role in the generation of SP-A1-specific antibodies that could recognize mammalian-expressed SP-A1. Antibodies raised against unhydroxylated peptides recognized insect cell-expressed SP-A1 protein but not the mammalian cell-expressed SP-A1 (unpublished data). Prolines are not hydroxylated in insect cells because these cells lack prolyl 4-hydroxylase activity required for proline hydroxylation (50, 53, 54). In general, in the molecules of the collectin family, the prolines at position Y are usually hydroxylated (9). However, since it is not known which prolines in SP-A collagen-like region are hydroxylated, a mixture of peptides containing randomly hydroxylated and unhydroxylated prolines was used as antigen. This approach enabled us to generate antibodies that could recognize in vitro-expressed SP-As from mammalian CHO cells vs. insect cells. Although the use of modified hydroxylated peptides to generate specific antibodies is not as common as the use of other modified peptides, such as phosphorylated peptides (13, 29), this approach proved essential in our case. Moreover, for SP-A, proline hydroxylation appears to be an important modification as this has been shown to affect its function (77) as well as other properties (21).
The SP-A1 antibody specificity was further assessed in pilot experiments (data not shown) using transgenic mice (TG) generated in the SP-A−/− background. Each TG mouse carries either a human SP-A1 or SP-A2 variant. The SP-A1 antibody stained alveolar type II cells in lung from the TG SP-A1 mouse and the human lung but not from the SP-A2 mouse; alveolar lung tissue from the SP-A−/− mouse was negative, as expected. The submucosal glands of the TG and SP-A−/− mice showed immunoreactivity with the SP-A1 antibody (data not shown), indicating that a cross-reacting antigen and/or a homolog of SP-A1 may be expressed in the submucosal glands. Full investigation of this is beyond the scope of this study.
Although the ratio of SP-A1 to SP-A2 has been suggested to be 2:1 based on the 2:1 composition of the SP-A trimer (76), studies of mRNA content in surgical lung tissues and fetal lung explants (38, 39) have shown that the gene-specific mRNA content varied from this proposed 2:1 ratio. This, assuming that SP-A mRNA levels reflect protein levels, raised the possibility that homooligomers of SP-A1 or SP-A2 do exist. Data from in vitro-expressed SP-A1 and SP-A2 variants showed that not only can homooligomers be formed, but these are also functional. In fact, functional (36, 57, 77) and regulatory (24, 39, 46, 56) differences between SP-A1 and SP-A2 and/or among SP-A variants (78, 79) have been observed. Therefore, we deemed it important to assess the SP-A gene-specific protein content in lung health and disease. The varying SP-A1-to-SP-A ratio observed in the present study in the various groups along with the published studies of functional and regulatory differences between SP-A1 and SP-A2 indicate that the overall SP-A activity may differ as a function of the health status of the lung.
In HS, the ratio of SP-A1 to total SP-A decreased with age, although there was no significant change in either the total SP-A or SP-A1 content. The latter is in agreement with published observations where although decreased SP-A levels as a function of age were found in bronchial lavage, no alterations in SP-A levels were observed in alveolar lavage (7). Of interest, in the present study, we observed that the proportion of SP-A1 declined with age. This may reflect the difference of regulatory mechanisms between human SP-A1 and SP-A2 during aging. Indeed, it has been observed that regulatory differences between human SP-A1 and SP-A2 exist in the transcriptional/pretranslational levels (38, 39) as well as at the translational level (78, 79). SP-A1 has been shown in vitro to be less active in host defense activities (36, 57, 80, 81) compared with SP-A2. Whether and how the decreased SP-A1-to-SP-A ratio in aged individuals contributes to altered host defense observed with aging (2, 12, 30) remains to be determined.
In individuals with AP, there was no change in the ratio compared with HS, but significant changes in the total SP-A content were observed between HS and AP subjects; the latter exhibited increased content. Berg et al. (6) using Edman degradation and mass spectrometry, showed that SP-A from AP consists mainly of the SP-A2 gene product and only a small portion of the SP-A1 gene product. It is important to note that the SP-A in the Berg study is derived from a single individual, and differences in the relative levels of SP-A1 and SP-A2 among AP individuals may exist. In our study, we observed a range of 0.009–0.562 for the SP-A1-to-total SP-A ratio with AP. The finding indicates a wide range of SP-A1-to-total SP-A ratio in vivo rather than the proposed 2:1 ratio (SP-A1:SP-A2). This may indicate that assembly of SP-A protein in vivo is found in multiple forms, including heterotrimers of SP-A1 and SP-A2 and homotrimers of either SP-A1 or SP-A2. This notion is also supported by our previous results from in vitro experiments in which single SP-A gene products from in vitro expression contained multiple-size oligomers (77).
However, an increase in the ratio was observed in CF subjects compared with HS, AP, or NCF, but the total SP-A and SP-A1 content in CF was decreased compared with all other study groups and AP, respectively. In CF, a chronic inflammatory disease of the airways with recurrent exacerbations and pulmonary deterioration (37, 42, 58), the higher ratio of SP-A1 to SP-A and the reduced total SP-A content may contribute to impaired host defense due to the reduced ability of SP-A1 to enhance proinflammatory cytokine production (80, 81) and/or phagocytosis by alveolar macrophages (57). In addition, the imbalance of proteases and antiproteases in CF patients (8, 23, 28) and the impact of this imbalance on the degradation of SP-A may further contribute to impaired host defense in CF patients. In fact, products of SP-A degradation have been observed in the BALF of CF patients but not in NCF patients (74). These data together indicate that differences in the relative SP-A gene-specific products exist in BAL samples among different subject groups, and this warrants further study. The availability of the SP-A1-specific antibody could facilitate further studies of samples from a variety of well-defined clinical patient groups.
In summary, we report the development and characterization of an SP-A1 gene-specific antibody and show for the first time at the protein level that the SP-A1-to-total SP-A content may differ in specimens from healthy and aged individuals as well as individuals with lung disease. The availability of the SP-A1 gene-specific antibody will enable the measurement of the relative SP-A1 gene-specific products in various well-defined clinical groups, and this may not only provide a more accurate assessment of the overall SP-A activity in the lung, compared with the measurement of the total SP-A content, but may also provide a better predictor for disease specificity.
This study was supported by National Heart, Lung, and Blood Institute Grants HL-68947 and HL-34788 and a research grant from the Barsumian Trust, Penn State University College of Medicine.
We gratefully acknowledge the Gift of Life Donor Program (Philadelphia, PA) and the generosity of the organ donor families for allowing these organs, which are not suitable for transplantation, to be utilized to advance the understanding of human disease, and Dr. X. Guo for technical assistance.
Present address of M. J. Thomassen: Program in Lung Cell Biology and Translational Research, Division of Pulmonary and Critical Care Medicine, East Carolina University, Brody School of Medicine, Greenville, NC 27834.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society