Pepsinogen C Proteolytic Processing of Surfactant Protein B^,

Abstract

Surfactant protein B (SP-B) is essential to the function of pulmonary surfactant and to lamellar body genesis in alveolar epithelial type 2 cells. The bioactive, mature SP-B is derived from multistep post-translational proteolysis of a larger proprotein. The identity of the proteases involved in carboxyl-terminal cleavage of proSP-B remains uncertain. This cleavage event distinguishes SP-B production in type 2 cells from less complete processing in bronchiolar Clara cells. We previously identified pepsinogen C as an alveolar type 2 cell-specific protease that was developmentally regulated in the human fetal lung. We report that pepsinogen C cleaved recombinant proSP-B at Met³⁰² in addition to an amino-terminal cleavage at Ser¹⁹⁷. Using a well described model of type 2 cell differentiation, small interfering RNA knockdown of pepsinogen C inhibited production of mature SP-B, whereas overexpression of pepsinogen C increased SP-B production. Inhibition of SP-B production recapitulated the SP-B-deficient phenotype evident by aberrant lamellar body genesis. Together, these data support a primary role for pepsinogen C in SP-B proteolytic processing in alveolar type 2 cells.

Alveolar patency is critical to gas exchange in the lung. The mixture of phospholipids, neutral lipids, and proteins that constitutes pulmonary surfactant facilitates alveolar patency by reducing surface tension of the alveolar lining fluid. Although phospholipids drive this biophysical process by organizing in a monolayer at the air-liquid interface in the alveolus, surfactant protein B (SP-B)2 has a vital function in pulmonary surfactant. The absence or deficiency of SP-B, due to developmental, genetic, or acquired disease processes, is associated with high surface tension in the alveolar space resulting in atelectasis, even in the presence of sufficient alveolar phospholipids (1). In animal models selective inactivation of SP-B from instilled monoclonal antibodies (2) in response to acute lung injury (3) or upon inactivation of SP-B gene expression in a conditional transgenic mouse model (4) is associated with poor surfactant biophysical profiles. Even when phospholipid levels were low in surfactant obtained from human patients, correction of phospholipid content alone was not sufficient to restore surfactant function, whereas supplementation with exogenous SP-B rapidly improved surface activity (5). Perhaps equally important, SP-B deficiency in vivo or in vitro is associated with abnormal type 2 cell ultrastructure due to the abnormal development of lamellar bodies, the lysosome-like structures that store and release surfactant within type 2 cells (6–10).

SP-B in its mature form is a small, lumenal, hydrophobic protein that commonly exists as a homodimer in the alveolar space (11, 12). Mature SP-B is the result of extensive proteolysis of a 381-amino acid proSP-B precursor that shares structural homology with the saposin/amoebapore family of proteins through the regular arrangement of intramolecular disulfide bridges (13). ProSP-B contains three saposin-like domains, with the mature SP-B protein residing in the second domain. Proteolysis of proSP-B after signal peptide cleavage involves the release of three saposin-like domains, amino terminus cleavage (domain 1) followed by carboxyl terminus cleavage (domain 3) and trimming of the amino- and carboxyl-terminal regions flanking mature SP-B (domain 2) (6, 14–16).

The first protease to be associated with proSP-B processing, cathepsin H, is responsible for the exopeptidase trimming of both the amino and carboxyl termini before release of the mature SP-B protein (6, 16). Inhibition of cathepsin H action either by protease inhibitors or RNA-mediated interference results in decreased production of mature SP-B. There is similar evidence using in vitro proteolysis and RNA-mediated interference that napsin A, an aspartic protease found in lung and kidney, is involved in removing the amino terminus from proSP-B between Leu¹⁷⁸ and Pro¹⁷⁹ (14, 16). However, the identity of the protease involved in cleavage of the carboxyl terminus has been elusive.

We recently described another aspartic protease, pepsinogen C, that is unique to alveolar type 2 cells in the lung (17). Pepsinogen C was highly induced during the process of in vivo and in vitro type 2 cell differentiation. Furthermore, in vitro transdifferentiation of type 2 cells into type 1-like cells was associated with abrupt cessation of pepsinogen C expression. We now report that pepsinogen C specifically cuts in both the amino- and carboxyl termini of proSP-B, resulting in a peptide Ser¹⁹⁷ to Met³⁰² containing mature SP-B. Knockdown of pepsinogen C results in the SP-B-deficient phenotype, whereas overexpression of pepsinogen C enhances SP-B production in a model of in vitro differentiation of alveolar type 2 cells. Together these findings establish pepsinogen C as an SP-B-processing enzyme.

EXPERIMENTAL PROCEDURES

Reagents—Dexamethasone, isobutylmethylxanthine (IBMX), and 8-bromo-cAMP were obtained from Sigma. Purified human pepsinogen C prepared from human gastric mucosa was obtained from Biodesign International (A38200H). Commercial antisera used herein included pepsinogen C (ab9013, Abcam), glyceraldehyde-3-phosphate dehydrogenase (MAB 374, Chemicon), and GST (554805, BD Biosciences), His (27-4710-01, GE Healthcare), and HA (2367, Cell Signaling Technology). A rabbit polyclonal antibody to human SP-B prepared from human alveolar proteinosis fluid was the kind gift of Dr. Aldo Baritussio, Padova, Italy. All other reagents were purchased from Fisher or Invitrogen.

Construction, Expression, and Purification of Recombinant ProSP-B—A recombinant protein based the 39-kDa human ProSP-B protein (unglycosylated M_r) was developed for in vitro proteolysis by pepsinogen C. In the recombinant protein proSP-B is flanked on the amino terminal side by (in order) a GST tag, His tag, and thrombin cleavage site and is flanked on the carboxyl-terminal side by an HA tag (depicted in Fig. 1A). A DNA fragment encoding proSP-B was generated by PCR from a plasmid containing the human SP-B cDNA. The following synthetic oligonucleotide primers were used: forward, CTC TCA TAT GCA TGG CTG AGT CAC ACC TG, and reverse, CTC TGA ATT CTT AAG CGT AAT CTG GAA CAT CGT ATG GGT AAA GGT CGG GGC TGT GGA TAC, where restriction enzyme NdeI and EcoRI (underlines) were incorporated to facilitate cloning. The initiation and termination codons are indicated in bold, and the HA tag is indicated as italics. The resultant PCR product was cloned into pGEMT-Easy cloning vector, and the product confirmed both via sequencing and restriction analysis. The PCR product was digested with Ndel and EcoRI, gel-purified, and cloned into pAcGHLT-A vector. The resultant clones were confirmed via both restriction analysis and sequencing. Recombinant baculovirus were generated by co-transfecting ∼70% confluent Spodoptera frugiperda insect cells (sf9) with linearized BaculoGold baculovirus DNA (BD Bioscience) and the recombinant transfer vector pAcGHLT-AProSP-B using BD BaculoGold transfection kit reagents. Recombinant viruses were amplified for two rounds. The amplifications were done at a multiplicity of infection <0.5. The titer was checked using a rapid titer kit (BacPak BV; Clontech). Sf9 cells were routinely maintained as shake flask cultures grown in Sf-900II serum-free medium (Invitrogen) at a 3 × 10⁵ cells/ml, 125 rpm, and 27 °C. For expression of recombinant ProSP-B protein, cultures were infected with recombinant baculovirus at a cell density of 2 × 10⁶ cells/ml using a multiplicity of infection between 3 and 6. Production cultures were harvested 48 h post-infection by centrifugation (1000 × g, 10 min), at which time the culture viability remained high (∼90%). Cells were lysed, and the recombinant protein was recovered using mouse monoclonal anti-HA-agarose conjugate and HA peptide per the manufacturer's instructions (Sigma).

FIGURE 1.

Epitope-tagged recombinant human proSP-B. A, schematic diagram of recombinant human proSP-B. The recombinant protein contains GST and His tags separated from the amino terminus of proSP-B by a thrombin cleavage consensus sequence (X). An HA tag was added to the carboxyl terminus of proSP-B. Arrows indicate the positions of the peptides recognized by epitope-specific antisera to the amino and carboxyl termini of proSP-B that were used to characterize the recombinant protein. B, immunoreactivity of the recombinant human proSP-B. Composite recombinant proSP-B separated by SDS-PAGE under reducing conditions followed by immunoblotting using antisera to peptide tags, proSP-B epitopes, and mature SP-B. Molecular masses were determined by MALDI-TOF to be 145970.3 (a), 72709.7 (b), and 43543.1 daltons (c)(n = 2). C, thrombin cleavage of recombinant proSP-B. Recombinant proSP-B was incubated with thrombin to remove the GST and His tags. Products were separated by SDS-PAGE under reducing conditions, and the gel was either stained with Coomassie Blue or transferred for immunoblotting with CTermB and GST antibodies (n = 2).

Cleavage of Recombinant ProSP-B by Peptide N-Glycosidase F, Thrombin, and Pepsinogen C—To assist in the characterization of the recombinant protein, peptide N-glycosidase F (New England Biolabs) and thrombin (Novagen) cleavages were accomplished per the manufacturer's instructions. Proteolysis by pepsinogen C was accomplished in either 0.1 m sodium citrate buffer at pH 3 unless otherwise specified in the presence or absence of pepstatin (Roche Diagnostics). Phosphate-citrate buffers of 0.2 m dibasic sodium phosphate and 0.1 m citric acid were used for pH dependence enzyme studies.

Western Immunoblotting—Cells were harvested in lysis buffer (1% Triton X-100, 150 mm NaCl, 50 mm Tris-HCL, 5 mm EDTA, 5% glycerol, pH 8.0) mixed with 1– protease inhibitor (Roche Applied Science). Total protein was measured using the Bradford method (18). Western immunoblotting of cell lysates or in vitro proteolysis was performed using NuPAGE Bis-Tris gels with MES running buffer (Invitrogen) under reducing conditions and transferred to polyvinylidene difluoride membranes per the manufacturer's instructions. Immunoblotting was performed as previously described (19) using the following primary antibodies: Baritussio SP-B, hSP-B (also developed from SP-B isolated from human alveolar proteinosis fluid) (15), CTermB (rabbit polyclonal to Leu³³⁰–Leu³⁴⁹ peptide derived from human proSP-B), CFlank (Gly²⁸⁴–Ser³⁰⁴), NFlank (Gln¹⁸⁶–Gln²⁰⁰), NFProx (Ser¹⁴⁵–Leu¹⁶⁰) at 1:4000, pepsinogen C at 1:5000, and GAPDH at 1:20,000. Blots were probed with secondary antibodies conjugated to Alexa Fluor 680 (Molecular Probes) at a dilution of 1:10,000 to detect SP-B antibodies and pepsinogen C and conjugated to IRDye 800 (Rock-land) at a dilution of 1:10,000 to simultaneously detect GAPDH expression. Blotted proteins were detected, and densitometry was analyzed using the Odyssey infrared imaging system (Li-Cor).

MALDI-TOF Mass Spectrometry—Protein samples were desalted using C₄ tips (Millipore) following the manufacturer's instructions. Samples were deposited onto the surface of a gold chip and allowed to dry followed by the addition of 1 μlof sinapinic acid (3.5 mg/ml in a mixture of water/acetonitrile (50/50) containing (0.1% trifluoroacetic acid). Samples were analyzed using a Ciphergen PBSIIC mass spectrometer. Analysis was performed using a laser energy of 230, sensitivity of 9, and a mass range set from 1,000 to 150,000 Da.

In-gel Tryptic Digestion—Wet Coomassie-stained gels were excised and cut into 1-mm³ cubes using a modified protocol (20). Briefly, gels were destained with 50% acetonitrile, 200 mm ammonium bicarbonate, reduced with 20 mm Tris(2-carboxy-ethyl)phosphine, and alkylated with 55 mm iodoacetamide. Gels were then washed with 25 mm ammonium bicarbonate followed by 50% acetonitrile, 25 mm ammonium bicarbonate. The gel pieces were then dried in a SpeedVac. Trypsin (20 ng/μl in 50 mm ammonium bicarbonate) was added to the gels until the gel was fully swelled and hydrated. Proteolysis was allowed to proceed at 37 °C overnight. After this, peptides were extracted with 20 mm ammonium bicarbonate followed by 5% formic acid, 50% acetonitrile. Samples were frozen at–80 °C until analysis.

LCMS/MS Analysis—Peptide digests were loaded directly onto a C₁₈ capillary column (75 μm × 100 mm; New Objective Proteoprep 2) isocratically in 2% acetonitrile, 0.1% formic acid at a flow rate of 1 μl/min using an Eksigent two-dimensional LC system. A linear gradient was then initiated at a flow rate of 300 nl/min (3–40% B over 42 min, 40–100% B over 3 min, then 5 min at 100% B). Buffer A was 0.1% formic acid, and buffer B was 80% acetonitrile, 0.1% formic acid. Mass spectrometry was performed on a Thermo-Finnegan LTQ mass spectrometer in a data-dependent fashion as peptides were eluted off of the capillary column. A top 5 method was performed in which one survey scan was followed by MS/MS analysis of the 5 most intense ions. MS thresholds were set to 1500, and the MS/MS was set to 500. A mass range of 800–2500 was implemented for all runs. A repeat count of 3 was selected such that after 3 MS/MS repeats this ion was placed onto an exclusion list for 0.5 min. An exclusion window was set to 0.5 daltons below the target m/z and 1.5 daltons above. MS/MS experiments were performed with an isolation width of 2 Hz, collision energy of 35, activation Q = 0.25, and an activation time of 30. Peptides that showed a neutral loss of 98, 49.5, or 33.2 daltons were subjected for a further round of sequencing (MS³).

Analysis of LCMS/MS Data and Data Base Searching—Raw Sequest files were searched against the species specific component of the Swissprot data base using both Sequest (Thermo) and MASCOT (Matrix Science). Two missed cleavages were allowed. A fixed modification of carbamidomethylation for cysteine and variable modification for methionine oxidation was used. A parent mass window of 1.2 and a fragment tolerance of 0.6 daltons were utilized for all ion-trap-based searches. A charge state of +2 and +3 was specified for all nanospray analyses. Peptide spectra were only accepted after SEQUEST analysis using the following criteria: Xcorr > 1.5 (z = 1), 2(z = 2), 2.5 (z = 3); ΔCn > 0.1 and a continuous ion series of at least 5 y-ions and b ions. Additionally, the five major peaks in the MS/MS spectrum needed to be identified. Semi-tryptic peptides, peptides in which either the carboxyl or amino terminus was non-tryptic, were considered in this analysis since non-tryptic cleavage of proSP-B by pepsinogen C was known to occur. Data were also searched using MASCOT. Only proteins with a MASCOT score above 60 were accepted. After the initial search, data were analyzed in detail using the program Scaffold (Proteome Software). Only proteins whose peptide probabilities were above 95% and protein probabilities above 80% were accepted. Any protein with a one-peptide match was manually inspected and only accepted if a contiguous y and/or b ion series was greater than or equal to five residues.

Alveolar Type 2 Cell Culture—Human fetal lung from 14–20-week therapeutic abortions came from the Birth Defects Laboratory in the Department of Pediatrics, University of Washington Medical Center (Seattle, WA) or from Advanced Biosciences Resources, Inc. (Alameda CA). All tissue was obtained and handled in compliance with state and federal law and under protocols approved by the Institutional Review Board, Children's Hospital of Philadelphia. Fetal lung epithelial cells were prepared as described (21). Cells were cultured on tissue culture plastic in 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm isobutylmethylxanthine (DCI medium) to induce and maintain type 2 cell differentiation.

siRNA Transfection—Naïve lung epithelial cells after isolation were transfected using the Amaxa Mammalian Epithelial Cell Nucleofector solution and Amaxa program M05 as recommended by the manufacturer (Amaxa). The siRNA (2–100 pmol) were added to 100 μl of nucleofector solution before executing the transfection protocol. Cells were then plated into 2 ml of Waymouth media with 10% fetal calf serum for a final siRNA concentration of 1–50 nm. After culture overnight to allow recovery and attachment, the medium was changed to Waymouth with 2% fetal calf serum to ensure that cells recovered completely and to minimize fibroblast overgrowth. After culture for 2 days, the medium was changed to Waymouth ± DCI without serum to induce differentiation of type 2 cells. The siRNA used included pepsinogen C and a scrambled pepsinogen C siRNA (Invitrogen), GAPDH (Ambion), and a fluorescence-conjugated negative control to assess transfection efficiency (siGLO, Dharmacon).

Real Time RT-PCR—Total cellular RNA was isolated with RNA STAT-60 Reagent (Tel-Test). Purity was verified by the absorbance 260:280 ratio. Integrity was screened using the Eukaryote Total RNA Nano assay on an Agilent 2100 bioanalyzer (Agilent). Real-time RT-PCR reactions using a single-plex strategy were performed using an ABI Prism 7900 system (Applied Biosystems). The two-step PCR protocol was performed as detailed elsewhere (17). The following primer/probe sets were used: SP-B Hs00167036_m1, pepsinogen C Hs00160052_m1, and GAPDH Hs99999905_m1, 18sHs99999901_ s1. All assays were determined to be in the linear amplification range by using cDNA standards developed from alveolar type 2 epithelial cells.

Ultrastructural Studies—Transmission electron microscopy samples were prepared and imaged as previously described (7). All transmission electron microscopy supplies were obtained from Electron Microscopy Sciences or Polysciences.

Construction and Expression of Human Pepsinogen C—RNA was prepared from human alveolar type 2 cells derived from in vitro differentiation of human fetal naïve epithelial cells and was reverse-transcribed into cDNA using the SuperScript™ First-Strand RT-PCR kit (Invitrogen). PCR was performed to obtain the human pepsinogen C cDNA using the following primers: forward, 5′-CAG TTG GGG ACC AGC ATC-3′; reverse, 5′-GTG CAG GGT CAA GAG GAA GA-3′. After inserting the cDNA into TA plasmid, the human pepsinogen C cDNA was subcloned into pcDNA3.1 plasmid by directional cloning using HindIII and XhoI sites.

Statistical Analysis—Results are given as the mean ± S.E. Analysis of variance with Dunnett's correction for multiple comparisons was performed using GraphPad Prism 4.00 for Macintosh (GraphPad). All RNA results were normalized to 18 S rRNA expression, and protein results were normalized to GAPDH expression.

RESULTS

Characterization of Recombinant ProSP-B—Coomassie-stained SDS-PAGE gels of recombinant human proSP-B revealed a major band at ∼70 kDa as expected (Fig. 1B). However, immunoblotting with antisera to epitopes within proSP-B, GST, His, and HA revealed three immunoreactive bands. MALDI-TOF revealed that these bands had masses of 145970.3, 72709.7, and 43543.1 daltons (Fig. 1B). The antiserum to the HA tag identified all three proteins, whereas the GST and His (His data not shown) antisera only identified the 72- and 145-kDa proteins.

We suspected that the 43-kDa band might arise from an alternative translation start site that eliminated the GST and His tags. In-gel trypsin digestion and mass spectrometry of each band identified 20 unique peptides covering 43% of the human proSP-B sequence. The recombinant proSP-B was then subjected to proteolytic cleavage by thrombin (Fig. 1C). Thrombin cleavage would remove the GST and His tags from the full-length recombinant protein, leaving a 43-kDa proSP-B with an HA tag. In the presence of thrombin both the 145- and 72-kDa proteins were reduced to a mass of 43 kDa. Together these data indicate that the dominant 72-kDa band is the full-length recombinant monomeric proSP-B, the 145-kDa band is a dimer of the full-length recombinant proSP-B that results from associations via the GST and/or His tags, and the 43-kDa band is the product of an alternate translation start site that excludes the GST and His tags.

In Vitro Cleavage of ProSP-B by Pepsinogen C—The specific pH profile for pepsinogen C cleavage of proSP-B was determined using phosphate-citrate buffers between pH 3 and 7. ProSP-B was digested by pepsinogen C over a broad pH range (Fig. 2A). Although the optimal pH was ≤5, pepsinogen C cleaved >25% of proSP-B at pH 6 after 1 h. In vitro digestion of proSP-B by pepsinogen C revealed two fragments with apparent molecular masses of 16 and 12 kDa. These fragments were recognized on immunoblots using antisera to mature SP-B and the CFlank epitope (Fig. 2B). By contrast, the more distal CTermB epitope-specific antibody only recognized the 16-kDa band. None of these fragments was identified by antisera to amino-terminal epitopes, GST, or His (not shown).

FIGURE 2.

In vitro proteolysis of recombinant proSP-B by pepsinogen C. A,pH dependence of pepsinogen C cleavage of proSP-B. Recombinant proSP-B was incubated with purified pepsinogen C in phosphate-citrate buffers of pH 3–7. Densitometry was performed on the recombinant proSP-B band (top) to determine the % proSP-B remaining after a 1-h digest (n = 3 experiments). B, pepsinogen C proteolysis of recombinant proSP-B. Recombinant proSP-B was incubated with pepsinogen C in sodium citrate buffer, pH 3, for 1 h in the presence or absence of pepstatin. The products were separated by SDS-PAGE under reducing conditions, and the gel was either stained with Coomassie or transferred for immunoblotting with SP-B, CFlank, and CTermB antibodies (asterisk, activated pepsinogen C enzyme; arrowhead, 12-kDa cleavage product; n = 5). C, carbohydrate modification of recombinant proSP-B and cleavage products. Recombinant proSP-B was incubated with peptide N-glycosidase (PNGase F) before or after proteolysis by pepsinogen C. The products were separated by SDS-PAGE under reducing conditions, and the gel was transferred for immunoblotting with CFlank antibody. Arrowhead, 12 kDa cleavage product; n = 2.

ProSP-B is typically glycosylated at Asn¹²⁹ and Asn³¹¹ of the amino and carboxyl termini, respectively. Neither the 72-kDa nor 145-kDa bands exhibited perceptible changes in migration after treatment with peptide N-glycosidase F (Fig. 2C). However, the 42-kDa proSP-B migrated faster after peptide N-glycosidase F. ProSP-B fragments resulting from pepsinogen C digestion were subjected to peptide N-glycosidase F digestion to determine whether any cleavage product was N-linked glycosylated. The cleavage product with an apparent molecular mass of 12 kDa (Fig. 2B, arrowhead), strongly immunoreactive to CFlank but not to CTermB, did not shift after peptide N-glycosidase F (Fig. 2C, arrowhead). Together these data indicate that the 12-kDa cleavage product resulting from pepsinogen C cleavage of proSP-B consists of mature SP-B with additional carboxyl-terminal amino acids recognized by CFlank antibody but does not include the glycosylation site at Asn³¹¹.

Mass Spectrometry of Recombinant proSP-B Cleavage Products—In-gel tryptic digestion and mass spectrometry of the 12-kDa band resulting from pepsinogen C digests of proSP-B revealed 20 unique peptides mapping to human proSP-B, many of which were contained within the amino acid sequence of mature SP-B (Fig. 3, underlined). Each tryptic fragment from the 12-kDa band was also identified in tryptic digests of full-length recombinant proSP-B. There were 3 semi-tryptic peptides identified from the 12-kDa band; in other words these peptides were tryptic at one end but non-tryptic at the other. The identities of these peptides were ¹⁹⁷SEQQFPIPLPYCWLCR, SPT-GEWLPRD²⁹⁵, and DSECHLCM³⁰² (Fig. 3, italicized sequence). None of the semi-tryptic peptides was identified in tryptic digests of full-length recombinant proSP-B, indicating that they were the result of pepsinogen C cleavage. The LCMS/MS analysis also identified the full tryptic peptide SPTGEWLPR²⁹⁴, making the SPTGEWLPRD²⁹⁵ peptide somewhat ambiguous and possibly the result of off-target proteolysis. This was not observed for the other two semi-tryptic peptides. The estimated M_r from Ser¹⁹⁷–Met³⁰² is 11,826 Da, in good agreement with the immunoblotting results. The schematic diagram in Fig. 3C illustrates the position of the 12-kDa cleavage product in reference to full-length proSP-B and the position of the epitope-specific antisera.

FIGURE 3.

Mass spectroscopy analysis of proSP-B cleavage products after pepsinogen C proteolysis. ProSP-B (50 μg) was digested with pepsinogen C, the products were separated by SDS-PAGE, and the 12-kDa band (see Fig. 2B) excised from the gel for in-gel tryptic digestion. Peptides were analyzed by LCMS/MS. Amino acid sequence coverage of peptides identified (A) and a representative spectrum (B) are shown (black, peptides identified; gray, proSP-B sequence not identified by LCMS/MS; underline, mature SP-B amino acid sequence). Three semi-tryptic fragments (SEQQFPIPLPYCWLCR, SPTGEWLPRD, DSECHLCM) representing pepsinogen C cleavage sites are shown in bold italics. C, schematic diagram of the position of the 12-kDa product of proSP-B cleavage by pepsinogen C based on LCMS/MS and immunoblotting. Amino acid positions refer to full-length human proSP-B. Arrows indicate the position of N-linked glycosylation sites. CTermB, CFlank, and hSP-B indicate the position of proteins/peptides used to develop antisera.

In Vivo Silencing of Pepsinogen C in Alveolar Type 2 Cells— To determine whether pepsinogen C accomplishes proSP-B processing in alveolar type 2 cells, naïve human lung epithelial cells were transfected with siRNA followed by in vitro differentiation into type 2 cells using 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm IBMX (DCI) for 72 h. We have previously shown that DCI treatment of naïve human lung epithelial cells results in increased expression of SP-B over 5 days of culture in addition to promoting the expression of lamellar bodies (21). Using siGLO, a fluorescence-tagged unrelated siRNA, >90% of plated cells were successfully transfected. The effect of siRNA on pepsinogen C RNA expression was dose-responsive between 2 and 100 pmol of siRNA in 2 separate experiments (Fig. 4), with poor pepsinogen C knockdown at 2 pmol of siRNA. Because cell viability by visual inspection and plating efficiency was decreased with 60 and 100 pmol of siRNA, all subsequent experiments were performed using the 20 pmol of siRNA.

FIGURE 4.

Dose-responsiveness of pepsinogen C siRNA. Representative dose-response experiment to determine optimal siRNA dose (n = 3). Naïve lung epithelial cells were transfected with 2–100 pmol of siRNA to pepsinogen C or a scrambled siRNA to pepsinogen C (PGC) followed by induction of type 2 cell differentiation. Shown are real time RT-PCR results examining pepsinogen C RNA (A) and composite immunoblot comparing pepsinogen C and GAPDH protein expression (B).

Pepsinogen C siRNA at 20 pmol for 72 h resulted in an average 70.8 ± 7.3% knockdown of pepsinogen C RNA with no effect on the expression of SP-B or GAPDH RNA (Fig. 5). Scrambled pepsinogen C siRNA had no effect on the expression of pepsinogen C, SP-B, or GAPDH RNA.

FIGURE 5.

RNA expression after in vivo silencing of pepsinogen C in alveolar type 2 cells. Naïve lung epithelial cells were transfected using no siRNA, a scrambled pepsinogen C siRNA (scrPGC), or siRNA to pepsinogen C (siPGC). Transfected cells were then cultured in Waymouth media alone (no DCI) or in media supplemented with 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm IBMX (DCI) to induce type 2 cell differentiation for 3 days. Real time RT-PCR results (mean ± S.E.) from four experiments are shown for pepsinogen C (PGC; A), SP-B (B), and GAPDH (C). Results were normalized to 18 S ribosomal RNA, expressed as % of cells receiving DCI alone, and compared by one-way analysis of variance with Dunnett's correction (p value shown when comparisons with the DCI control group were significant).

Pepsinogen C siRNA resulted in 67.3 ± 6.2% knockdown of pepsinogen C protein levels at 72 h (Fig. 6B) and had no effect on GAPDH protein levels (Fig. 6A). In addition, pepsinogen C knockdown resulted in 57.3 ± 5.6% knockdown of 8-kDa SP-B protein (Fig. 6E). At early time points pepsinogen C knockdown was associated with increased proSP-B and 25 kDa intermediate (supplemental Fig. S1), but by 72 h there was no effect of pepsinogen C knockdown on proSP-B (Fig. 6C) and a small but significant decrease in the 25-kDa intermediate (Fig. 6D). Cells transfected with scrambled siRNA to pepsinogen C exhibited no change in the levels of proSP-B, 25-kDa SP-B intermediate, or mature SP-B protein at all time points examined.

FIGURE 6.

Protein expression after in vivo silencing of pepsinogen C in alveolar type 2 cells. Naïve lung epithelial cells were transfected using no siRNA, a scrambled pepsinogen C siRNA (scrPGC), or siRNA to pepsinogen C (siPGC). Transfected cells were then cultured in Waymouth media alone (no DCI) or in media supplemented with 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm IBMX (DCI) to induce type 2 cell differentiation for 3 days. Cell lysates (30μg) were immunoblotted for GAPDH, pepsinogen C (PGC), CTermB, and hSP-B. A composite immunoblot from a representative experiment (A) and densitometry results for pepsinogen C (B), proSP-B (C), 25-kDa SP-B intermediate (D), and 8 kDa SP-B (E) from 4 experiments are shown. Densitometry of immunoblots results (mean ± S.E.) were normalized for GAPDH, expressed as % of cells receiving DCI alone, and compared by one-way analysis of variance with Dunnett's correction (p value shown when comparisons with the DCI control group were significant).

We have previously shown that interruption of SP-B processing using protease inhibitors to cathepsin H disrupted lamellar body formation (6). Electron microscopy of cells transfected with pepsinogen C siRNA for 72 h exhibited abnormal membrane-limited, electron-dense inclusions (Fig. 7, C and F) when compared with cells transfected in the absence of siRNA (Fig. 7, B and E) and to cells transfected with scrambled siRNA (not shown), both having well developed lamellar bodies.

FIGURE 7.

Type 2 cell morphology after in vivo silencing of pepsinogen C in alveolar type 2 cells. Naïve lung epithelial cells were transfected using no siRNA, a scrambled pepsinogen C siRNA (not shown), or siRNA to pepsinogen C (siPGC). Transfected cells were then cultured in Waymouth media alone (no DCI) or in media supplemented with 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm IBMX (DCI) to induce type 2 cell differentiation for 3 days. Composite electron micrographs of cells in the no DCI (A and D), DCI (B and E), and siPGC (C and F) treatment groups from a representative experiment are shown (n = 3).

Overexpression of Pepsinogen C in Alveolar Type 2 Cells—To determine whether increased pepsinogen C enhanced production of SP-B in alveolar type 2 cells, naïve human lung epithelial cells were transfected with plasmid containing the cDNA for human pepsinogen C (pcDNA-PGC) followed by in vitro differentiation into type 2 cells using 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm IBMX (DCI) for up to 96 h. Pepsinogen C protein levels were increased, and this was associated with increased mature SP-B (Fig. 8). Enhanced mature SP-B was not associated with changes in proSP-B or the 25-kDa intermediate (data not shown).

FIGURE 8.

Overexpression of pepsinogen C in alveolar type 2 cells increases production of mature SP-B. Naïve lung epithelial cells were transfected with pcDNA3-PGC or pEGFP-N1 and were cultured for up to 96 h in media supplemented with 10 nm dexamethasone, 0.1 mm 8-Br-cAMP, and 0.1 mm IBMX to induce type 2 cell differentiation. A composite immunoblot for GAPDH, pepsinogen C (PGC), and SP-B is shown using antisera (n = 3).

DISCUSSION

SP-B is a critical component of surfactant and plays an essential role in lamellar body genesis in alveolar type 2 cells (for review, see Ref. 1). The biological properties of SP-B are largely due to its hydrophobicity, rendering it particularly fusogenic toward phospholipid membranes. It has been proposed that these properties are limited within cells by the containment of SP-B in a larger proprotein, proSP-B, such that proteolytic processing to the mature protein ensures that the fusogenic activities are released in an appropriate environment, specifically the multivesicular and lamellar bodies. Furthermore, the developmental regulation of proteolytic processing ensures that lamellar body genesis proceeds at an appropriate time during fetal lung development to allow for sufficient surfactant stores to accumulate in preparation for the adaptation to air-breathing at birth.

Pepsinogen C (pepsinogen II, progastricsin) is an aspartyl protease initially identified as a gastric zymogen with functions similar to pepsinogen A (23). Unlike pepsinogen A, pepsinogen C has a broad expression profile, including more extensive gastrointestinal tract expression as well as more distant sites like lung, pancreas, genital tract, and several neoplasms including breast cancer (24). When secreted extracellularly, the inactive zymogen is activated in an acidic environment for action on extracellular propeptides. Although the action of pepsinogen C in the stomach is to degrade dietary proteins, pepsinogen C in seminal fluid is used to activate defensins for local antimicrobial defense (25). Intracellular proprotein activation in an acidic organelle, as in proSP-B proteolytic processing in the secretory pathway of alveolar type 2 cells, is a novel function for pepsinogen C.

Our previous work and studies by others have begun to identify the proteases essential to post-translational processing of proSP-B to the bioactive 8-kDa SP-B protein. We employed three strategies to examine the role of pepsinogen C in proSP-B processing; they are in vitro proteolysis of recombinant proSP-B, in vivo knockdown of pepsinogen C by siRNA during type 2 cell differentiation, and in vivo overexpression of pepsinogen C in type 2 cells. In vitro proteolysis of recombinant proSP-B would allow us to identify specific pepsinogen C cleavage sites but would require an appropriately folded recombinant protein. Preparations of recombinant proSP-B protein expressed in Escherichia coli have been difficult to work with, requiring chemical denaturation and refolding, and are not glycosylated, potentially leading to exposure of proteolytic sites not typically exposed in vivo (26). Because N-linked glycans assist in protein folding, orient epitopes toward the surface of proteins, and provide local protection from proteolysis (27), we chose to use a baculovirus system to produce large quantities of glycosylated proSP-B for our studies.

The recombinant proSP-B was a mixture of three different proteins existing largely as a monomer. Recombinant proSP-B was stable in solution and was immunoreactive to all of the inserted epitope tags as well as to epitope-specific antisera for regions of proSP-B. There was a minimal amount of dimeric recombinant protein. Dimerization was unlikely the result of disulfide bonding at Cys²⁴⁸ since the dimer was readily detectable under reducing conditions upon gel electrophoresis. Furthermore, the dimer migrated at 42-kDa after removal of the GST and His tags by thrombin cleavage, suggesting that dimer formation was facilitated by epitopes within these tags and was not due to any intermolecular associations involving proSP-B. As anticipated, the recombinant proSP-B was also glycosylated.

It was not surprising to find multiple cleavage products from the in vitro proteolysis of proSP-B by pepsinogen C. Instead, it was surprising that pepsinogen C cleavage in vitro left the mature SP-B core intact. The reported consensus sequence for pepsinogen C (EC 3.4.23.3; cleavage at Y-X or W-X) suggested at least three potential cleavage sites within proSP-B in addition to five potential sites within GST and one within the HA tag, but none of these predicted separation of the C-terminal peptide from proSP-B. However, pepsinogen C did cleave the recombinant proSP-B, producing a 12-kDa product from Ser¹⁹⁷– Met³⁰² containing mature SP-B protein with an additional small amino-terminal fragment and larger carboxyl-terminal fragment.

An amino-terminal pepsinogen C cleavage site between Leu¹⁹⁶ and Ser¹⁹⁷ was suggested by the presence of the peptide SEQQFPIPLPYCWLCR after in-gel tryptic digest of the 12-kDa proSP-B cleavage product. This described a unique and more distal amino-terminal cleavage site than had been demonstrated for napsin A (between Leu¹⁷⁸ and Pro¹⁷⁹) or cathepsin H (between A¹⁸⁷ and Arg¹⁸⁸ (14, 16). The presence of residual amino-terminal peptide after pepsinogen C cleavage would require an additional protease to trim the amino terminus to Phe-²⁰¹ of mature SP-B.

A more complex carboxyl-terminal proteolytic profile was observed from pepsinogen C. Two putative pepsinogen C cleavage sites were observed between Asp²⁹⁵ and Ser²⁹⁶ and between Met³⁰² and Ser³⁰³, as identified by the LCMS/MS peptides SPTGEWLPRD and DSECHLCM. The concomitant presence of SPTGEWLPR, a typical tryptic fragment, sheds doubt that the Asp²⁹⁵–Ser²⁹⁶ cleavage is a pepsinogen C cleavage site. However, the identification of the DSECHLM peptide was unambiguous and established at least one pepsinogen C cleavage site between Met³⁰² and Ser³⁰³ in the carboxyl terminus of proSP-B.

Participation of Napsin A in carboxyl-terminal processing has been controversial. In one instance, multiple napsin A carboxyl-terminal cleavages were identified from proSP-B and lamellar body proteolytic products, whereas fusion peptides consisting of the carboxyl terminus of proSP-B and EGFP exhibited no proteolysis from recombinant Napsin A. Our data suggest that pepsinogen C cleavage would leave a lengthy carboxyl-terminal peptide (Asp²⁸⁰–Cys³⁰¹) attached to mature SP-B for subsequent cleavage by another protease. A cathepsin H cleavage site between Met²⁷⁹ and Asp²⁸⁰ (16) would be a likely candidate.

The proteolytic specificity of pepsinogen C has not been as completely studied as pepsinogen A. Initial reports indicated that pepsinogen C was more reactive against hemoglobin and demonstrated activation at higher pH, albeit more slowly (28). These studies also demonstrated retention of 95–100% of activity after exposure to pH as high as 6.9. Based upon initial characterization of purified gastricsin (29) and cleavage of oxidized insulin B chain (28), glucagon, and oxidized ribonuclease A (30), it has been accepted that pepsinogen C prefers a Tyr in the P₁ position. All three putative pepsinogen C cleavage sites in proSP-B involved a Ser at the P′₁ position and Leu, Asp, or Met in the P₁ position. Although Leu, Asp, and Met are common in the P1 position of pepsin A (31), detailed information on pepsinogen C is lacking for these amino acids. However, pepsinogen C cleavage between Glu-Ser, Gln-Ser, Ser-Ser, and Met-Ser has been reported (30).

To show that pepsinogen C was involved in SP-B processing in vivo, we utilized a robust model of in vitro differentiation of type 2 cells to study the effects of pepsinogen C knockdown and overexpression. We have used this model previously in conjunction with adenovirus-mediated antisense SP-B expression and cysteine protease inhibitors to perturb SP-B expression (6, 7). Naïve lung epithelial cells are devoid of mature SP-B protein and lamellar bodies and contain very low levels of SP-B RNA (21) and no pepsinogen C RNA or protein (17). This permitted us to study the onset of both SP-B expression and lamellar body genesis without background expression interfering with our ability to detect changes in proSP-B, mature SP-B, or intermediate forms of SP-B peptides. Furthermore, the onset of pepsinogen C expression is more rapid than SP-B expression with in vitro differentiation, allowing us to establish intracellular siRNA and pepsinogen C overexpression before endogenous pepsinogen C expression.

Pepsinogen C knockdown was specific and dose-dependent. The siRNA to pepsinogen C had no untoward effects on expression of the housekeeping gene GAPDH or on the expression of SP-B RNA. Scrambled siRNA had no effects on the expression of any of the genes examined. Therefore, the effects of pepsinogen C knockdown on the expression of mature SP-B or on lamellar body genesis was due to the interference with pepsinogen C function in type 2 cells.

SP-B is essential for lamellar body transformation from multivesicular bodies. In the absence of SP-B in genetic knockout animals and human patients with inherited deficiency of SP-B, lamellar bodies are absent, and multivesicular bodies are more numerous (10, 32). We have previously demonstrated aberrant lamellar body formation in association with knockdown of mature SP-B expression in our in vitro model system either by knocking down SP-B gene expression or by inhibiting cathepsin H (6, 7). In our present studies pepsinogen C knockdown resulted in prominent, membrane-bound organelles with heterogeneous electron dense inclusions similar to multivesicular bodies and similar to organelles seen with cathepsin H inhibition (6).

Based on our prior studies inhibiting cathepsin H in type 2 cells (6), we expected that pepsinogen C knockdown would result in not only decreased SP-B but increased proSP-B and/or 25-kDa intermediate. This was only evident at early time points after induction of type 2 cell differentiation. The difference between cathepsin H inhibition and pepsinogen C knockdown may reflect the site of proteolysis. Pepsinogen C precursors are likely to be in more proximal cellular compartments (33) that can be more easily secreted from the cell (22) or degraded.

Together, our in vitro and in vivo data establish pepsinogen C as a necessary protease in SP-B processing. Although it is possible that pepsinogen C could be necessary for activating some other process in SP-B proteolysis, the surprising specificity of cleavages in vitro that resulted in an intact mature SP-B protein, the presence of substrate accumulation early in the course of in vivo knockdown, and the enhanced production of SP-B in the context of pepsinogen C overexpression are highly suggestive that pepsinogen C participates directly in proSP-B proteolysis. Together with our prior observation that pepsinogen C is expressed by alveolar type 2 cells and not by bronchiolar Clara cells, we make a compelling argument that the differences in the forms of SP-B produced by these two cell types is due to the differential expression of pepsinogen C. The combination of tight spatial and developmental regulation of pepsinogen C and its critical role in SP-B proteolysis likely contribute to the rapid availability of mature SP-B to promote lamellar body genesis in the early third trimester of human gestation. This burst in mature, functional SP-B production and in lamellar body genesis facilitates the availability of sufficient surfactant stores for newborn lungs in the transition to air breathing.