190
Biochimica et B iophvsica A cta, 1087 (1990) 190-I98 Elsevier
BBAEXP 92175
Expression and characterization of rat surfactant protein A synthesized in Chinese hamster ovary cells Francis X. McCormack 1,2, James H. Fisher 2, Akira Suwabe l, David L. Smith 1, John M. Shannon 1,2 and Dennis R. Voelker 1,2 I Lord and Taylor Laboratory for Lung Biochemistry, Department of Medicine, National Jewish Center for Immunology and Respiratory Medicine and the 2 Pulmonary Division, Department of Medicine, University of Colorado Health Sciences Center, Denver, CO (U. S. A. )
(Received 9 March 1990) (Revised manuscript received 12 June 1990)
Key words: Surfactant associated protein; Pulmonary surfactant; Recombinant DNA; Gene expression; Recombinant protein: (Chinese hamster ovary cell line); (Rat)
Rat surfactant protein A (SP-A) was expressed in a Chinese hamster ovary (CHO-KI) cell line and characterized for biologic activity using assays for receptor binding and modulation of phospholipid secretion from isolated type II cells. The CHO-K1 cell line was cotransfected with separate plasmids encoding for the rat SP-A, dihydrofolate reduetase and neomycin phosphotransferase, respectively. Antibiotic (Geneticin-G418)-resistant transformants were screened by ELISA for the secretion of recombinant SP-A into the media. Northern analysis of the transfected cell lines demonstrated the expression of both 1.6 kb and 0.9 kb mRNA species for SP-A, consistent with the proposed differential polyadenylation of the primary transcript. Amplification with methotrexate resulted in a dose-dependent increase in mRNA for SP-A and a 20-fold increase in the production of recombinant SP-A relative to untreated cells. Maximum production of SP-A was 370/~g of S P - A / I of media in a 4-day incubation. Recombinant SP-A was purified from the serum-free media of large scale cultures of transfected, amplified C H O cells by affinity chromatography on mannose-Sepharose. The recombinant SP-A migrated similarly to native SP-A by NaDodSO4-PAGE analysis under reducing and nonreducing conditions and under reducing conditions after digestion with N-glycanase. Recombinant SP-A effectively competed with 125I-native SP-A for binding to the high affinity receptor for SP-A on isolated plasma membranes from rat alveolar type II cells. The recombinant SP-A was as effective as native SP-A in the inhibition of secretion of phospholipid from isolated type II cells. We conclude that recombinant rat SP-A produced in Chinese hamster ovary cells is physically and functionally similar to native rat SP-A.
Introduction
Pulmonary surfactant is a complex mixture of phospholipid, protein and cholesterol which functions to reduce alveolar surface tension at the air/liquid interface [1]. The protein constituents of surfactant have recently been shown to be important contributors to the function and metabolism of surfactant [2]. The most abundant of these proteins in rat is a 26-38 kDa
Abbreviations: SP-A, pulmonary surfactant protein A; TPA, 12-O-tetradecanoylphorbol-13-acetate; ELISA, enzyme-linked immunosorbent assay; NaDodSO4-PAGE, sodium dodecylsulfate-polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline. Correspondence: F.X. McCormack, 1400 Jackson St., Denver, CO 80206, U.S.A.
glycoprotein (under reducing and denaturing conditions) designated surfactant protein A (SP-A) [3]. The precise functional role of SP-A in the alveolus is unknown although in vitro evidence suggests that the protein may act both in the formation of the surfactant monolayer and as an autocrine regulator of surfactant homeostasis [4]. The SP-A, in concert with the hydrophobic proteins; SP-B and SP-C, enhances the adsorption and spreading of surface active phosphohpid at the air/liquid interface [5]. The SP-A has been shown to enhance the uptake of liposomal phospholipid from the media of primary cultures of alveolar type II cells [6]. Several laboratories have demonstrated that SP-A functions to inhibit the secretion of phospholipid from isolated alveolar type II cells [7-9]. The inhibition of secretion of phospholipid is mediated, at least in part, through a high-affinity cell surface receptor for SP-A
0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
191 [10-12]. Taken together these observations suggest that SP-A functions in the assembly of the surfactant monolayer, in the modulation of secretion of surfactant phospholipid and in the recycling of phospholipids from the alveolar space. The amino acid sequence for dog [13], rat [14], rabbit [15] and human [16] SP-A have been deduced from the cDNAs isolated from each. Sequence homology of SP-A with the mammalian lectin, mannose-binding protein-A [17], has led to the elucidation of the carbohydratebinding properties of SP-A [18]. The overall structural organization of SP-A [19] is conserved and is similar to the organization of Clq [20] and mannose-binding protein [17]. SP-A from the rat [14,19] is comprised of a short globular amino terminal sequence followed by a unique collagen-like domain of 24 Gly-x-y repeats rich in hydroxyproline. A single interruption at the midpoint of this region introduces a bend into the molecule which presumably functions to separate the globular heads of the protein in space. The collagenous domain is followed by a putative phospholipid binding region [19] and the globular carbohydrate-binding domain where most of the sequence homology with mannose-binding protein resides. SP-A isolated from the rat resolves into three molecular weight species on one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (NaDodSO 4PAGE) under reducing and denaturing conditions [3]. These bands at 26, 32 and 38 kDa are due to differential glycosylation of the monomeric primary translation product [21]. Under nonreducing and denaturing conditions the protein forms disulfide-dependent oligomers with molecular masses of 50-70, 115 and 160 kDa and greater [9]. The conformation of the protein in the alveolus is unknown although gel permeation chromatography of delipidated, purified rat SP-A gives a molecular mass at 1.6 • 1 0 3 kDa [9]. Other investigators have found the molecular weight of SP-A from other species to be 6.5 • 102 kDa, suggesting an aggregate of 18 monomers [22]. Measurements of circular dichroism and melting temperature of canine and human SP-A support a collagen-like helical interaction among subunits [22, 23]. Human SP-A has been estimated by these methods to exist as an oligomer of 18 subunits arranged as 6 trimers of collagen-like triple helices, analogous to the structure of the complement component, Clq [20]. A recent electron microscopic study of canine SP-A isolated from bronchoalveolar lavage and human recombinant SP-A further supports this octadecameric arrangement of subunits [24]. SP-A undergoes an extensive amount of post translational modification resulting in the multitude of isoforms seen on two-dimensional polyacrylamide gel electrophoresis. These include: proteolytie cleavage of the signal peptide [14]; acetylation [25]; proline hydroxylation [14]; vitamin K-dependent carboxylation of
glutamic acid residues [26]; and N-linked glycosylation [21,27,28] with sialation [27] and sulfation [29] of the oligosaccharide moiety. The 1.6 kb c D N A for rat SP-A was isolated from a rat lambda gt 10 lung library using the human cDNA for SP-A as a probe [14]. This cDNA contains the entire coding region for rat SP-A as well as untranslated 3' and 5' regions [14]. Two species of m R N A for SP-A, 1.6 kb and 0.9 kb, were identified from RNA isolated from homogenates of whole lung and of isolated type II cells. There was complete agreement between the sequence of the N-terminal 58 amino acids of rat SP-A determined by Edman degradation and the translated nucleotide sequence of the isolated cDNA. The purpose of this study was to: (1) express the previously isolated cDNA for rat SP-A in a mammalian cell fine; (2) isolate the recombinant protein from the media; and (3) compare the recombinant protein with the native protein from rat lung lavage for the ability to inhibit phospholipid secretion and to bind to the highaffinity receptor of alveolar type II cells. Materials and Methods
Chemicals Reagents for chromatography were obtained from Pharmacia and Bio-Rad. Restriction endonucleases, ligases and DNA polymerases were purchased from International Biotechnologies. Radioisotopes were purchased from New England Nuclear ([3H]choline), ICN (32p) and Amersham (125I-Bolton-Hunter reagent). Dry chemicals were obtained from Sigma and culture media was prepared from powder obtained from GIBCO. Purification of native SP-A Surfactant was isolated from Sprague-Dawley rats 4 weeks after the intratracheal instillation of 20 mg of silica. The surfactant was purified as described by Hawgood et al. [21]. The SP-A was isolated and purified using previously published methods [9] by eentrifugation, affinity chromatography and gel permeation chromatography. Preparation of I :51-SP-A The 125I-SP-A used in the cell-free binding experiments was prepared using the reagent described by Bolton and Hunter [30]. A solution of 1-2 m g / m l of SP-A in 0.1 M sodium borate buffer (pH 8.5) was added to a 1 mCi vial of Bolton-Hunter reagent (Amersham) from which the organic solvent had just been evaporated. The reaction proceeded on ice for 30 min and unconjugated lz5I was removed by dialysis against 5 m M Tris buffer (pH 7.4). The 125I-SP-A was then repurified by affinity chromatography on mannose-Sepharose 6B [31]. The specific activity of the 125I-SP-A used ranged from 100-300 cpm/ng.
192
ELISA for SP-A The apoprotein content of tissue culture media containing native or recombinant SP-A was determined using a polyclonaI IgG against rat SP-A in a sandwich ELISA technique [9,32]. The antibody was raised in NZW rabbits by serial injections of purified SP-A and isolated by previously published methods [33]. The lower limit of sensitivity of the assay was approx. 0.5 n g / m l and the linear range extended from 0.5-10 ng/ml. Construction of the expression vector We previously reported the isolation and sequencing of the 1.6 kb c D N A for rat SP-A, cloned into the EcoRI site of p U C 19 [14]. This insert was excised with EcoRI and purified with preparative agarose gel electrophoresis. The overhanging ends of the insert were then filled in using the Klenow fragment of D N A polymerase and dNTPs and this construct was blunt end ligated to Sph I linkers. The modified insert was digested with SphI and ligated into the SphI site of the polylinker of p U C 19. The proper orientation (HindIII, 5' and XbaI, 3') was identified using restriction fragment length analysis with Barn HI. The insert was excised from this plasmid with HindIII and XbaI, and asymmetrically cloned into the expression vector PRSV.3 (obtained from Eric Long, N.I.H.). CHO K1 Cell Expression The transfeetion of a Chinese hamster ovary cell line (CHO-K1) was accomplished using a modification of the CaPO 4 method of Davis [34]. CHO-K1 cells were plated at a density of 3 • 105/60 mm plastic culture dish in complete medium ( D M E M supplemented with 10% FCS, 0.6 mM proline, 0.4 mM ascorbic acid, 100 t~g/ml penicillin, 100 # g / m l gentamicin) for 20 h prior to transfection. The PRSV.3/SP-A construct together with the plasmids pML neo (obtained from Alleem Siddiqui, University of Colorado Health Sciences Center), conferring resistance to aminoglycoside antibiotics and pSV2 dhfr [35], conferring resistance to methotrexate were used in the cotransfection of the C H O K-1 cells. 20 ~g of plasmid D N A in the relative mass ratio of 6 : 3 : 1 ; PRSV.3/SP-A, pSV2 dhfr and pML neo, respectively, were suspended in 0.5 ml of calcium phosphate solution [34]. The resultant precipitate was added slowly to the media over the monolayer of cells. After incubation of the monolayer with the precipitate for 30 min at 37 o C, 5 ml of fresh complete medium was applied and the cells were incubated at 37 °C for 24 h. A fresh application of complete media was applied at 24 h. After 48 h, fresh complete medium containing 1.6 m g / m l geneticin (G418) was added. Based on the initial mass ratios of plasmids used in the cotransformation, transfectants demonstrating resistance to the antibiotic geneticin were likely to also have incorporated the plasmids encoding for the SP-A and for dhfr expression. Antibiotic-re-
sistant transfectants were identified at 2 weeks and selected colonies were subcloned twice using cloning cylinders. The isolated colonies were then grown to confluence on 60 mm tissue culture dishes and screened for the secretion of immunoreactive SP-A into the media by ELISA.
Amplification with methotrexate In general, cotransfecting plasmids (e.g., PRSV.3/ SP-A and pSV2 dhfr) are incorporated into the host genome in tandem, permitting the methotrexate-induced amplification of loci which integrate adjacent to the plasmid derived dhfr locus [36]. We relied upon this tandem insertion for the methotrexate based amplification of the c D N A for SP-A. The concentration of methotrexate in 10 mM stock solutions was confirmed spectrophotometrically (E370 = 8100 M I cm-1). Complete media for methotrexate amplification was prepared as described above with the exception that the fetal calf serum was exhaustively dialyzed against PBS to remove thymidine prior to addition to the media. The transfected C H O K-1 cell line with the greatest level of production of SP-A (C5A) was subjected to a weekly doubling in the concentration of methotrexate from 0.01 /~M to 160/~M in complete medium. Isolation of recombinant rat SP-A The C5A cell line amplified as described above was grown to confluence in 100 m M dishes and seeded into roller bottles at 2-3 dishes/890 cm 2 bottle. Once the cells had reached confluence (5-7 days) in complete medium containing 160 ~M methotrexate, the serum containing media was removed and the cells were washed with sterile PBS. Serum-free media (same as complete medium except for the deletion of serum) was then incubated with the monolayer for 3-4 days, harvested and passed through a 0.2 micron nitrocellulose filter. This sequence was repeated for a second harvest. The 1-2 1 supernatant containing recombinant SP-A was passed over a 10 mt mannose-Sepharose 6B column in the presence of 5 m M Tris buffer (pH 7.4) containing 1 mM calcium. The column was washed with the T r i s / C a buffer and then eluted with 5 mM Tris buffer (pH 7.4) containing 2 mM EDTA. The fractions containing recombinant SP-A by ELISA were pooled and dialyzed against 10 mM ammonium bicarbonate (pH 7.8) overnight at 4 o C. The sample was lyophylized to dryness and resuspended in a minimal volume of 5 mM Tris buffer. Analysis of SP-A Protein samples were separated by sodium dodecyl sulfate 10% polyacrylamide gel electrophoresis (NaDodSO4-PAGE ) [37] and either stained with Coomassie blue or transferred to nitrocellulose for immunoblotting [38]. Membranes with adsorbed protein species were sequen-
193 tially incubated with 20 g g / m l polyclonal anti SP-A IgG and 10 6 cpm of 125I-labeled S. aureus protein A [38]. Immunoreactive species were imaged with autoradiography. For N-glycanase treatment, 10 gl of the native or recombinant protein at 0.5-1.0 m g / m l was boiled for 3 min in the presence of 0.5% SDS and 0.1 M fl-mercaptoethanol. The sample was then diluted in 0.2 M phosphate buffer (pH 8.6) and 1.0% NP-40. N-glycanase was added to 27 units/ml. The reaction mixture was incubated overnight at 37 °C. The products of the digestion were analyzed by NaDodSO4-PAGE under reducing and denaturing conditions.
Primary culture of alveolar type H cells and secretion of p h osphatidy lcholine Alveolar type II cells were isolated from male Sprague-Dawley rats by tissue dissociation with elastase and purification on metrizamide gradients [39]. The type II cells were seeded into 35 mm dishes at a density of 2 - 1 0 6 cells/dish and cultured for 21 h with [3H]choline (0.5 g C i / m l ) . Secretion of radiolabeled phosphatidylcholine was determined as described previously [9]. The phorbol ester 12-O-tetradecanoyl-phorbol-13-acetate (TPA) (10 -7 M) was used as a stimulant of surfactant secretion.
Cell-free plasma membrane binding An assay was developed for cell-free binding of SP-A to isolated plasma membrane. In this technique, isolated type II cells were homogenized with a Dounce apparatus and plasma membrane was purified on a discontinuous sucrose gradient by centrifugation at 80 000 ×gav for 3 h. This gradient system is a modification of that described by Duck-Chong [40]. The plasma membrane enriched fraction (between 0.9 M and 1.48 M sucrose) was washed and incubated (1.3 #g protein) with 125I-SP-A (0.16 gg) in a final volume of 160 gl of buffer A composed of 2 mM calcium, 100 m M NaC1 and 20 mg/rnl BSA in 50 mM Tris-HC1 (pH 7.4). Binding reactions were terminated by the addition of 0.8 ml of ice cold buffer B (same as buffer A except the albumin concentration is 2 mg/ml). The binding of 125I-SP-A to plasma membrane was determined utilizing filter-binding methods [41]. The filters were then washed four times with cold buffer B. The amount of bound 125I-SP-A was determined using a gamma counter. Nonspecific binding was defined as that amount of radioligand binding which occurred in the presence of 100 # g / m l of unlabeled SP-A. The time, temperature and SP-A concentration dependence of binding to isolated plasma membrane closely paralleled that seen in the intact cell (manuscript in preparation).
Solution hybridization The m R N A for SP-A was quantified using the technique described by Palmiter [42,43].
RNA analysis Total cellular R N A from freshly isolated type II cells, C H O cells and the transfected C5A cell lines was extracted in guanidinium isothiocyante. Northern analysis was performed on CsC1 purified R N A [44]. The Nytran membranes were hybridized with a rat SP-A 32p-labeled probe prepared from the 0.9 kb rat SP-A c D N A insert by random primer second-strand synthesis (Amersham kit). Results
Transformation and amplification of the cDNA for SP-A The expression vector chosen for the transformation, PRSV.3, contains the Rous sarcoma virus LTR promoter adjacent to the polylinker derived from P U C 19. This system has been successfully used for the expression of human globin genes [45]. The 1.6 kB c D N A insert for SP-A was asymmetrically ligated into the PRSV.3 vector following digestion of both components with HindlII and XbaI. The C H O K-1 cell line was utilized for the expression of recombinant SP-A. Following the cotransformation of this cell line with the plasmids PRSV.3/SP-A, pSV2 dhfr and p M L net, more than 50 colonies were identified on the plated monolayer of 3 • 105 cells grown in selective media. 40 colonies were cloned and screened by ELISA for the secretion of immunoreactive SP-A into the media. Eight SP-A secreting cell lines were identified ranging from 1.2-18.3 ng SP-A/ml with C5A cell line exhibiting the greatest level (18.3 ng/ml). By solution hybridization it was determined that the C5A cell line contained approx. 10% of the m R N A level found in freshly isolated type II cells (not shown). The C5A cell line was then subjected to a weekly doubling in the concentration of methotrexate from 0.01 # M to 160 gM. This resulted in a 20-fold increase in the concentration of detectable SP-A to 370 n g / m l of media. Northern analysis of the transfected C5A cell line subjected to increasing doses of methotrexate demonstrated a dose-dependent increase in the m R N A for SP-A (both the 1.6 and 0.9 kB forms) approaching that of freshly isolated type II cells (Fig. 1). With longer periods of exposure, m R N A species at 0.9 and 1.6 kb are evident for the unamplified cell line as well (not shown).
Purification of recombinant SP-A The recombinant SP-A was isolated from the serumfree media of large scale cultures of the methotrexate amplified C5A cell lines using affinity chromatography on mannose-Sepharose. Under these conditions of methotrexate-induced gene amplification, the cell line secreted 300-400 ~tg SP-A/1 of serum-free media during a 3- to 4-day incubation. The cells continue to secrete SP-A in serum-free media for up to 6 days with the yield of SP-A being approx. 150-250/_tg/1 between
194 Discussion). The fractions of EDTA eluted material which contained SP-A as determined by ELISA were pooled and dialyzed against t0 mM ammonium bicarbonate. The samples were then lyophylized to dryness, resuspended in a minimal volume of 5 mM Tris buffer and frozen at - 20 ° C. Polyacrylamide gel electrophoresis of the native SP-A under reducing conditions demonstrated a triplet with bands of molecular mass 26, 32 and 38 kDa, with a heterogeneous band between the masses of 32 and 38 kDa (Fig. 2). These molecular species observed on NaDodSO4-PAGE were identified as SP-A by immunoblot analysis (Fig. 2). The immunoreactive band at approx. 62 kDa in the native rat SP-A lane under reducing conditions is consistent with the nonreducible dimeric form of SP-A which has t~een noted in the alveolar lavage of patients with alveolar proteinosis by several investigators. Treatment of the native protein with N-glycanase results in the simplification of molecular species on NaDodSO4-PAGE to bands at 26 and 32 kDa (Fig. 3). The band at 32 kDa is a glycosylated form of the 26 kDa primary translation product that is resistant to N-glycanase [46]. The recombinant SP-A also nfigrates as a triplet on NaDodSQ-PAGE. The polymorphism of the recombinant protein was simplified to
!
1.6 kb
-
-
.9 kb
-
-
0
ALV, TYPE II
60
CHO
120
160
CSA CELLS + METI'IOTREXATE(uM)
Fig. 1. Northern analysis of the amplified C5A cell line. Total cellular RNA from alveolar type 11 cells (type II), Chinese hamster ovary cells (CHO) and the transfected cell line C5A subjected to 0 160 /*M methotrexate was extracted and analysed as outlined in Materials and Methods. 5 btg of RNA was loaded per lane.
day 3 and 6 and 50-100/,g/1 between days 6 and 9. In general, we harvested 2 1 of media from 2-4 890-cm2 roller bottles at 3-day intervals over 6 days. This media was passed through a 0.2 micron filter and the recombinant protein was isolated and purified by mannoseSepharose affinity column chromatography. We routinely found that 50-85% of the recombinant SP-A contained in the media did not bind to the column (see
Mol. Wt. (kDa)
120
- -
60
--
43
--
30
A
B
--"
nSP-A
rSP-A
I
nSP-A
rSP-A
I I
REDUCING
NONREDUCING
nSP-A I
rSP-A
I
nSP-A I
REDUCING
- -
120
--
60
--
43
--
30
rSP-A
I
I
NONREDUCING
Fig. 2. Electrophoretic and immunoblot analysis of native and recombinant SP-A. Samples of native SP-A(nSP-A) and recombinant SP-A(rSP-A) were analyzed by NaDodSO4-PAGE (Panel A) as described in Materials and Methods using 10% polyacrylamide gels and either reducing or non-reducing conditions. The proteins were then electrophoretically transferred to a Nytran membrane overnight. The immunoreactive species were imaged by incubation of the membrane with polyclonal anti SP-A IgG and 1251-S. a u r e u s protein A followed by autoradiography (Panel B).
195 20
Mol. Wt.
(kDa)
NATIVE SP-A
RECOMBINANT SP-A =| I
A e,,
15 R e c o m b i n a n t SP-A
"N
c =•[ .c_ 10
10
m
4, ~ ~ _ a t i v e S P . A
Ft
s nonspecifi¢ binding
N-Glycanase
"
-I"
"1"
'4"
i 5
"!"
Fig. 3. Electrophoretic analysis of N-glycanase treated native and recombinant SP-A. Native and recombinant SP-A were digested with N-glycanase at 37 ° C overnight under reducing and denaturing conditions as described in Materials and Methods. Native SP-A and deglycosylated native and recombinant SP-A were then analyzed by NaDOdSO4-PAGE under denaturing and reducing conditions.
a pattern indistinguishable from that of the native protein when both were treated with N-glycanase (Fig. 3). We conclude that the heterogeneous band of the recombinant product on NaDodSO4-PAGE is due to posttranslational glycosylation. Under nonreducing and denaturing conditions the native protein migrates as disulfide-dependent oligomers with molecular masses of 50-70, 115 and 160 kDa and greater. The recombinant protein was indistinguishable from the native protein under these conditions (Fig. 2).
Recombinant SP-A competes for binding to the high affinity receptor for SP-A Studies from this laboratory have demonstrated that the inhibition of phospholipid secretion by SP-A is mediated, at least in part, by the binding of SP-A to a high-affinity cell surface receptor on isolated type II cells [10]. Binding of radiolabeled native SP-A to a cell surface receptor is readily demonstrated on whole alveolar type II cells in primary culture [10]. A cell-free receptor-binding assay has been developed, which demonstrates binding characteristics and kinetics that are indistinguishable from the intact cell surface receptor (manuscript in preparation). This technique greatly reduces the quantity of SP-A required for analysis of receptor binding. We found that recombinant SP-A effectively competes with 125I-SP-A for binding to isolated plasma membrane from type I! cells (Fig. 4). It is apparent from the data in this figure that the activity of recombinant SP-A was equal to that of native SP-A in competing for high-affinity binding. Recombinant SP-A inhibits the secretion of phospholipid from isolated type H cells Several laboratories have demonstrated that native SP-A inhibits the stimulated secretion of phospholipid from isolated alveolar type II cells [7-9]. We studied whether the recombinant SP-A synthesized in CHO
i 10
i 15
Protein (gg) Fig. 4. Competition for cell-free binding of native and recombinant SP-A to isolated plasma membranes from alveolar type II cells. Plasma membrane from isolated alveolar type II cells was purified on sucrose density gradients and washed as described in Materials and Methods. A 2-#g (protein) aliquot of plasma membrane was incubated with t25I-SP-A (1 # g / m l ) and varying amounts of unlabeled native and recombinant SP-A. The binding of radiolabeled SP-A to plasma membrane was determined by filter-binding methods. The data show both the specific and the non-specific components of binding.
cells was capable of inhibition of phospholipid secretion. Native SP-A reduced the TPA-stimulated secretion of [3H]phosphatidylcholine to half the maximal levels at a concentration of 0.1 gg of SP-A/ml and to below basal levels at I #g of SP-A/ml. We found that the recombinant product also inhibited the TPA-stimulated secretion of surfactant phospholipids from alveolar type II ceUs in an identical manner. The recombinant SP-A product was equivalent to native SP-A as an inhibitor of surfactant lipid secretion (Fig. 5).
r.
100
.o (J O (n
75
% ~ N a t - - - ..... - o
Recombinant
Ire
"10 t~
"5
50
,
Biochimica et B iophvsica A cta, 1087 (1990) 190-I98 Elsevier
BBAEXP 92175
Expression and characterization of rat surfactant protein A synthesized in Chinese hamster ovary cells Francis X. McCormack 1,2, James H. Fisher 2, Akira Suwabe l, David L. Smith 1, John M. Shannon 1,2 and Dennis R. Voelker 1,2 I Lord and Taylor Laboratory for Lung Biochemistry, Department of Medicine, National Jewish Center for Immunology and Respiratory Medicine and the 2 Pulmonary Division, Department of Medicine, University of Colorado Health Sciences Center, Denver, CO (U. S. A. )
(Received 9 March 1990) (Revised manuscript received 12 June 1990)
Key words: Surfactant associated protein; Pulmonary surfactant; Recombinant DNA; Gene expression; Recombinant protein: (Chinese hamster ovary cell line); (Rat)
Rat surfactant protein A (SP-A) was expressed in a Chinese hamster ovary (CHO-KI) cell line and characterized for biologic activity using assays for receptor binding and modulation of phospholipid secretion from isolated type II cells. The CHO-K1 cell line was cotransfected with separate plasmids encoding for the rat SP-A, dihydrofolate reduetase and neomycin phosphotransferase, respectively. Antibiotic (Geneticin-G418)-resistant transformants were screened by ELISA for the secretion of recombinant SP-A into the media. Northern analysis of the transfected cell lines demonstrated the expression of both 1.6 kb and 0.9 kb mRNA species for SP-A, consistent with the proposed differential polyadenylation of the primary transcript. Amplification with methotrexate resulted in a dose-dependent increase in mRNA for SP-A and a 20-fold increase in the production of recombinant SP-A relative to untreated cells. Maximum production of SP-A was 370/~g of S P - A / I of media in a 4-day incubation. Recombinant SP-A was purified from the serum-free media of large scale cultures of transfected, amplified C H O cells by affinity chromatography on mannose-Sepharose. The recombinant SP-A migrated similarly to native SP-A by NaDodSO4-PAGE analysis under reducing and nonreducing conditions and under reducing conditions after digestion with N-glycanase. Recombinant SP-A effectively competed with 125I-native SP-A for binding to the high affinity receptor for SP-A on isolated plasma membranes from rat alveolar type II cells. The recombinant SP-A was as effective as native SP-A in the inhibition of secretion of phospholipid from isolated type II cells. We conclude that recombinant rat SP-A produced in Chinese hamster ovary cells is physically and functionally similar to native rat SP-A.
Introduction
Pulmonary surfactant is a complex mixture of phospholipid, protein and cholesterol which functions to reduce alveolar surface tension at the air/liquid interface [1]. The protein constituents of surfactant have recently been shown to be important contributors to the function and metabolism of surfactant [2]. The most abundant of these proteins in rat is a 26-38 kDa
Abbreviations: SP-A, pulmonary surfactant protein A; TPA, 12-O-tetradecanoylphorbol-13-acetate; ELISA, enzyme-linked immunosorbent assay; NaDodSO4-PAGE, sodium dodecylsulfate-polyacrylamide gel electrophoresis; PBS, phosphate-buffered saline. Correspondence: F.X. McCormack, 1400 Jackson St., Denver, CO 80206, U.S.A.
glycoprotein (under reducing and denaturing conditions) designated surfactant protein A (SP-A) [3]. The precise functional role of SP-A in the alveolus is unknown although in vitro evidence suggests that the protein may act both in the formation of the surfactant monolayer and as an autocrine regulator of surfactant homeostasis [4]. The SP-A, in concert with the hydrophobic proteins; SP-B and SP-C, enhances the adsorption and spreading of surface active phosphohpid at the air/liquid interface [5]. The SP-A has been shown to enhance the uptake of liposomal phospholipid from the media of primary cultures of alveolar type II cells [6]. Several laboratories have demonstrated that SP-A functions to inhibit the secretion of phospholipid from isolated alveolar type II cells [7-9]. The inhibition of secretion of phospholipid is mediated, at least in part, through a high-affinity cell surface receptor for SP-A
0167-4781/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)
191 [10-12]. Taken together these observations suggest that SP-A functions in the assembly of the surfactant monolayer, in the modulation of secretion of surfactant phospholipid and in the recycling of phospholipids from the alveolar space. The amino acid sequence for dog [13], rat [14], rabbit [15] and human [16] SP-A have been deduced from the cDNAs isolated from each. Sequence homology of SP-A with the mammalian lectin, mannose-binding protein-A [17], has led to the elucidation of the carbohydratebinding properties of SP-A [18]. The overall structural organization of SP-A [19] is conserved and is similar to the organization of Clq [20] and mannose-binding protein [17]. SP-A from the rat [14,19] is comprised of a short globular amino terminal sequence followed by a unique collagen-like domain of 24 Gly-x-y repeats rich in hydroxyproline. A single interruption at the midpoint of this region introduces a bend into the molecule which presumably functions to separate the globular heads of the protein in space. The collagenous domain is followed by a putative phospholipid binding region [19] and the globular carbohydrate-binding domain where most of the sequence homology with mannose-binding protein resides. SP-A isolated from the rat resolves into three molecular weight species on one-dimensional sodium dodecyl sulfate polyacrylamide gel electrophoresis (NaDodSO 4PAGE) under reducing and denaturing conditions [3]. These bands at 26, 32 and 38 kDa are due to differential glycosylation of the monomeric primary translation product [21]. Under nonreducing and denaturing conditions the protein forms disulfide-dependent oligomers with molecular masses of 50-70, 115 and 160 kDa and greater [9]. The conformation of the protein in the alveolus is unknown although gel permeation chromatography of delipidated, purified rat SP-A gives a molecular mass at 1.6 • 1 0 3 kDa [9]. Other investigators have found the molecular weight of SP-A from other species to be 6.5 • 102 kDa, suggesting an aggregate of 18 monomers [22]. Measurements of circular dichroism and melting temperature of canine and human SP-A support a collagen-like helical interaction among subunits [22, 23]. Human SP-A has been estimated by these methods to exist as an oligomer of 18 subunits arranged as 6 trimers of collagen-like triple helices, analogous to the structure of the complement component, Clq [20]. A recent electron microscopic study of canine SP-A isolated from bronchoalveolar lavage and human recombinant SP-A further supports this octadecameric arrangement of subunits [24]. SP-A undergoes an extensive amount of post translational modification resulting in the multitude of isoforms seen on two-dimensional polyacrylamide gel electrophoresis. These include: proteolytie cleavage of the signal peptide [14]; acetylation [25]; proline hydroxylation [14]; vitamin K-dependent carboxylation of
glutamic acid residues [26]; and N-linked glycosylation [21,27,28] with sialation [27] and sulfation [29] of the oligosaccharide moiety. The 1.6 kb c D N A for rat SP-A was isolated from a rat lambda gt 10 lung library using the human cDNA for SP-A as a probe [14]. This cDNA contains the entire coding region for rat SP-A as well as untranslated 3' and 5' regions [14]. Two species of m R N A for SP-A, 1.6 kb and 0.9 kb, were identified from RNA isolated from homogenates of whole lung and of isolated type II cells. There was complete agreement between the sequence of the N-terminal 58 amino acids of rat SP-A determined by Edman degradation and the translated nucleotide sequence of the isolated cDNA. The purpose of this study was to: (1) express the previously isolated cDNA for rat SP-A in a mammalian cell fine; (2) isolate the recombinant protein from the media; and (3) compare the recombinant protein with the native protein from rat lung lavage for the ability to inhibit phospholipid secretion and to bind to the highaffinity receptor of alveolar type II cells. Materials and Methods
Chemicals Reagents for chromatography were obtained from Pharmacia and Bio-Rad. Restriction endonucleases, ligases and DNA polymerases were purchased from International Biotechnologies. Radioisotopes were purchased from New England Nuclear ([3H]choline), ICN (32p) and Amersham (125I-Bolton-Hunter reagent). Dry chemicals were obtained from Sigma and culture media was prepared from powder obtained from GIBCO. Purification of native SP-A Surfactant was isolated from Sprague-Dawley rats 4 weeks after the intratracheal instillation of 20 mg of silica. The surfactant was purified as described by Hawgood et al. [21]. The SP-A was isolated and purified using previously published methods [9] by eentrifugation, affinity chromatography and gel permeation chromatography. Preparation of I :51-SP-A The 125I-SP-A used in the cell-free binding experiments was prepared using the reagent described by Bolton and Hunter [30]. A solution of 1-2 m g / m l of SP-A in 0.1 M sodium borate buffer (pH 8.5) was added to a 1 mCi vial of Bolton-Hunter reagent (Amersham) from which the organic solvent had just been evaporated. The reaction proceeded on ice for 30 min and unconjugated lz5I was removed by dialysis against 5 m M Tris buffer (pH 7.4). The 125I-SP-A was then repurified by affinity chromatography on mannose-Sepharose 6B [31]. The specific activity of the 125I-SP-A used ranged from 100-300 cpm/ng.
192
ELISA for SP-A The apoprotein content of tissue culture media containing native or recombinant SP-A was determined using a polyclonaI IgG against rat SP-A in a sandwich ELISA technique [9,32]. The antibody was raised in NZW rabbits by serial injections of purified SP-A and isolated by previously published methods [33]. The lower limit of sensitivity of the assay was approx. 0.5 n g / m l and the linear range extended from 0.5-10 ng/ml. Construction of the expression vector We previously reported the isolation and sequencing of the 1.6 kb c D N A for rat SP-A, cloned into the EcoRI site of p U C 19 [14]. This insert was excised with EcoRI and purified with preparative agarose gel electrophoresis. The overhanging ends of the insert were then filled in using the Klenow fragment of D N A polymerase and dNTPs and this construct was blunt end ligated to Sph I linkers. The modified insert was digested with SphI and ligated into the SphI site of the polylinker of p U C 19. The proper orientation (HindIII, 5' and XbaI, 3') was identified using restriction fragment length analysis with Barn HI. The insert was excised from this plasmid with HindIII and XbaI, and asymmetrically cloned into the expression vector PRSV.3 (obtained from Eric Long, N.I.H.). CHO K1 Cell Expression The transfeetion of a Chinese hamster ovary cell line (CHO-K1) was accomplished using a modification of the CaPO 4 method of Davis [34]. CHO-K1 cells were plated at a density of 3 • 105/60 mm plastic culture dish in complete medium ( D M E M supplemented with 10% FCS, 0.6 mM proline, 0.4 mM ascorbic acid, 100 t~g/ml penicillin, 100 # g / m l gentamicin) for 20 h prior to transfection. The PRSV.3/SP-A construct together with the plasmids pML neo (obtained from Alleem Siddiqui, University of Colorado Health Sciences Center), conferring resistance to aminoglycoside antibiotics and pSV2 dhfr [35], conferring resistance to methotrexate were used in the cotransfection of the C H O K-1 cells. 20 ~g of plasmid D N A in the relative mass ratio of 6 : 3 : 1 ; PRSV.3/SP-A, pSV2 dhfr and pML neo, respectively, were suspended in 0.5 ml of calcium phosphate solution [34]. The resultant precipitate was added slowly to the media over the monolayer of cells. After incubation of the monolayer with the precipitate for 30 min at 37 o C, 5 ml of fresh complete medium was applied and the cells were incubated at 37 °C for 24 h. A fresh application of complete media was applied at 24 h. After 48 h, fresh complete medium containing 1.6 m g / m l geneticin (G418) was added. Based on the initial mass ratios of plasmids used in the cotransformation, transfectants demonstrating resistance to the antibiotic geneticin were likely to also have incorporated the plasmids encoding for the SP-A and for dhfr expression. Antibiotic-re-
sistant transfectants were identified at 2 weeks and selected colonies were subcloned twice using cloning cylinders. The isolated colonies were then grown to confluence on 60 mm tissue culture dishes and screened for the secretion of immunoreactive SP-A into the media by ELISA.
Amplification with methotrexate In general, cotransfecting plasmids (e.g., PRSV.3/ SP-A and pSV2 dhfr) are incorporated into the host genome in tandem, permitting the methotrexate-induced amplification of loci which integrate adjacent to the plasmid derived dhfr locus [36]. We relied upon this tandem insertion for the methotrexate based amplification of the c D N A for SP-A. The concentration of methotrexate in 10 mM stock solutions was confirmed spectrophotometrically (E370 = 8100 M I cm-1). Complete media for methotrexate amplification was prepared as described above with the exception that the fetal calf serum was exhaustively dialyzed against PBS to remove thymidine prior to addition to the media. The transfected C H O K-1 cell line with the greatest level of production of SP-A (C5A) was subjected to a weekly doubling in the concentration of methotrexate from 0.01 /~M to 160/~M in complete medium. Isolation of recombinant rat SP-A The C5A cell line amplified as described above was grown to confluence in 100 m M dishes and seeded into roller bottles at 2-3 dishes/890 cm 2 bottle. Once the cells had reached confluence (5-7 days) in complete medium containing 160 ~M methotrexate, the serum containing media was removed and the cells were washed with sterile PBS. Serum-free media (same as complete medium except for the deletion of serum) was then incubated with the monolayer for 3-4 days, harvested and passed through a 0.2 micron nitrocellulose filter. This sequence was repeated for a second harvest. The 1-2 1 supernatant containing recombinant SP-A was passed over a 10 mt mannose-Sepharose 6B column in the presence of 5 m M Tris buffer (pH 7.4) containing 1 mM calcium. The column was washed with the T r i s / C a buffer and then eluted with 5 mM Tris buffer (pH 7.4) containing 2 mM EDTA. The fractions containing recombinant SP-A by ELISA were pooled and dialyzed against 10 mM ammonium bicarbonate (pH 7.8) overnight at 4 o C. The sample was lyophylized to dryness and resuspended in a minimal volume of 5 mM Tris buffer. Analysis of SP-A Protein samples were separated by sodium dodecyl sulfate 10% polyacrylamide gel electrophoresis (NaDodSO4-PAGE ) [37] and either stained with Coomassie blue or transferred to nitrocellulose for immunoblotting [38]. Membranes with adsorbed protein species were sequen-
193 tially incubated with 20 g g / m l polyclonal anti SP-A IgG and 10 6 cpm of 125I-labeled S. aureus protein A [38]. Immunoreactive species were imaged with autoradiography. For N-glycanase treatment, 10 gl of the native or recombinant protein at 0.5-1.0 m g / m l was boiled for 3 min in the presence of 0.5% SDS and 0.1 M fl-mercaptoethanol. The sample was then diluted in 0.2 M phosphate buffer (pH 8.6) and 1.0% NP-40. N-glycanase was added to 27 units/ml. The reaction mixture was incubated overnight at 37 °C. The products of the digestion were analyzed by NaDodSO4-PAGE under reducing and denaturing conditions.
Primary culture of alveolar type H cells and secretion of p h osphatidy lcholine Alveolar type II cells were isolated from male Sprague-Dawley rats by tissue dissociation with elastase and purification on metrizamide gradients [39]. The type II cells were seeded into 35 mm dishes at a density of 2 - 1 0 6 cells/dish and cultured for 21 h with [3H]choline (0.5 g C i / m l ) . Secretion of radiolabeled phosphatidylcholine was determined as described previously [9]. The phorbol ester 12-O-tetradecanoyl-phorbol-13-acetate (TPA) (10 -7 M) was used as a stimulant of surfactant secretion.
Cell-free plasma membrane binding An assay was developed for cell-free binding of SP-A to isolated plasma membrane. In this technique, isolated type II cells were homogenized with a Dounce apparatus and plasma membrane was purified on a discontinuous sucrose gradient by centrifugation at 80 000 ×gav for 3 h. This gradient system is a modification of that described by Duck-Chong [40]. The plasma membrane enriched fraction (between 0.9 M and 1.48 M sucrose) was washed and incubated (1.3 #g protein) with 125I-SP-A (0.16 gg) in a final volume of 160 gl of buffer A composed of 2 mM calcium, 100 m M NaC1 and 20 mg/rnl BSA in 50 mM Tris-HC1 (pH 7.4). Binding reactions were terminated by the addition of 0.8 ml of ice cold buffer B (same as buffer A except the albumin concentration is 2 mg/ml). The binding of 125I-SP-A to plasma membrane was determined utilizing filter-binding methods [41]. The filters were then washed four times with cold buffer B. The amount of bound 125I-SP-A was determined using a gamma counter. Nonspecific binding was defined as that amount of radioligand binding which occurred in the presence of 100 # g / m l of unlabeled SP-A. The time, temperature and SP-A concentration dependence of binding to isolated plasma membrane closely paralleled that seen in the intact cell (manuscript in preparation).
Solution hybridization The m R N A for SP-A was quantified using the technique described by Palmiter [42,43].
RNA analysis Total cellular R N A from freshly isolated type II cells, C H O cells and the transfected C5A cell lines was extracted in guanidinium isothiocyante. Northern analysis was performed on CsC1 purified R N A [44]. The Nytran membranes were hybridized with a rat SP-A 32p-labeled probe prepared from the 0.9 kb rat SP-A c D N A insert by random primer second-strand synthesis (Amersham kit). Results
Transformation and amplification of the cDNA for SP-A The expression vector chosen for the transformation, PRSV.3, contains the Rous sarcoma virus LTR promoter adjacent to the polylinker derived from P U C 19. This system has been successfully used for the expression of human globin genes [45]. The 1.6 kB c D N A insert for SP-A was asymmetrically ligated into the PRSV.3 vector following digestion of both components with HindlII and XbaI. The C H O K-1 cell line was utilized for the expression of recombinant SP-A. Following the cotransformation of this cell line with the plasmids PRSV.3/SP-A, pSV2 dhfr and p M L net, more than 50 colonies were identified on the plated monolayer of 3 • 105 cells grown in selective media. 40 colonies were cloned and screened by ELISA for the secretion of immunoreactive SP-A into the media. Eight SP-A secreting cell lines were identified ranging from 1.2-18.3 ng SP-A/ml with C5A cell line exhibiting the greatest level (18.3 ng/ml). By solution hybridization it was determined that the C5A cell line contained approx. 10% of the m R N A level found in freshly isolated type II cells (not shown). The C5A cell line was then subjected to a weekly doubling in the concentration of methotrexate from 0.01 # M to 160 gM. This resulted in a 20-fold increase in the concentration of detectable SP-A to 370 n g / m l of media. Northern analysis of the transfected C5A cell line subjected to increasing doses of methotrexate demonstrated a dose-dependent increase in the m R N A for SP-A (both the 1.6 and 0.9 kB forms) approaching that of freshly isolated type II cells (Fig. 1). With longer periods of exposure, m R N A species at 0.9 and 1.6 kb are evident for the unamplified cell line as well (not shown).
Purification of recombinant SP-A The recombinant SP-A was isolated from the serumfree media of large scale cultures of the methotrexate amplified C5A cell lines using affinity chromatography on mannose-Sepharose. Under these conditions of methotrexate-induced gene amplification, the cell line secreted 300-400 ~tg SP-A/1 of serum-free media during a 3- to 4-day incubation. The cells continue to secrete SP-A in serum-free media for up to 6 days with the yield of SP-A being approx. 150-250/_tg/1 between
194 Discussion). The fractions of EDTA eluted material which contained SP-A as determined by ELISA were pooled and dialyzed against t0 mM ammonium bicarbonate. The samples were then lyophylized to dryness, resuspended in a minimal volume of 5 mM Tris buffer and frozen at - 20 ° C. Polyacrylamide gel electrophoresis of the native SP-A under reducing conditions demonstrated a triplet with bands of molecular mass 26, 32 and 38 kDa, with a heterogeneous band between the masses of 32 and 38 kDa (Fig. 2). These molecular species observed on NaDodSO4-PAGE were identified as SP-A by immunoblot analysis (Fig. 2). The immunoreactive band at approx. 62 kDa in the native rat SP-A lane under reducing conditions is consistent with the nonreducible dimeric form of SP-A which has t~een noted in the alveolar lavage of patients with alveolar proteinosis by several investigators. Treatment of the native protein with N-glycanase results in the simplification of molecular species on NaDodSO4-PAGE to bands at 26 and 32 kDa (Fig. 3). The band at 32 kDa is a glycosylated form of the 26 kDa primary translation product that is resistant to N-glycanase [46]. The recombinant SP-A also nfigrates as a triplet on NaDodSQ-PAGE. The polymorphism of the recombinant protein was simplified to
!
1.6 kb
-
-
.9 kb
-
-
0
ALV, TYPE II
60
CHO
120
160
CSA CELLS + METI'IOTREXATE(uM)
Fig. 1. Northern analysis of the amplified C5A cell line. Total cellular RNA from alveolar type 11 cells (type II), Chinese hamster ovary cells (CHO) and the transfected cell line C5A subjected to 0 160 /*M methotrexate was extracted and analysed as outlined in Materials and Methods. 5 btg of RNA was loaded per lane.
day 3 and 6 and 50-100/,g/1 between days 6 and 9. In general, we harvested 2 1 of media from 2-4 890-cm2 roller bottles at 3-day intervals over 6 days. This media was passed through a 0.2 micron filter and the recombinant protein was isolated and purified by mannoseSepharose affinity column chromatography. We routinely found that 50-85% of the recombinant SP-A contained in the media did not bind to the column (see
Mol. Wt. (kDa)
120
- -
60
--
43
--
30
A
B
--"
nSP-A
rSP-A
I
nSP-A
rSP-A
I I
REDUCING
NONREDUCING
nSP-A I
rSP-A
I
nSP-A I
REDUCING
- -
120
--
60
--
43
--
30
rSP-A
I
I
NONREDUCING
Fig. 2. Electrophoretic and immunoblot analysis of native and recombinant SP-A. Samples of native SP-A(nSP-A) and recombinant SP-A(rSP-A) were analyzed by NaDodSO4-PAGE (Panel A) as described in Materials and Methods using 10% polyacrylamide gels and either reducing or non-reducing conditions. The proteins were then electrophoretically transferred to a Nytran membrane overnight. The immunoreactive species were imaged by incubation of the membrane with polyclonal anti SP-A IgG and 1251-S. a u r e u s protein A followed by autoradiography (Panel B).
195 20
Mol. Wt.
(kDa)
NATIVE SP-A
RECOMBINANT SP-A =| I
A e,,
15 R e c o m b i n a n t SP-A
"N
c =•[ .c_ 10
10
m
4, ~ ~ _ a t i v e S P . A
Ft
s nonspecifi¢ binding
N-Glycanase
"
-I"
"1"
'4"
i 5
"!"
Fig. 3. Electrophoretic analysis of N-glycanase treated native and recombinant SP-A. Native and recombinant SP-A were digested with N-glycanase at 37 ° C overnight under reducing and denaturing conditions as described in Materials and Methods. Native SP-A and deglycosylated native and recombinant SP-A were then analyzed by NaDOdSO4-PAGE under denaturing and reducing conditions.
a pattern indistinguishable from that of the native protein when both were treated with N-glycanase (Fig. 3). We conclude that the heterogeneous band of the recombinant product on NaDodSO4-PAGE is due to posttranslational glycosylation. Under nonreducing and denaturing conditions the native protein migrates as disulfide-dependent oligomers with molecular masses of 50-70, 115 and 160 kDa and greater. The recombinant protein was indistinguishable from the native protein under these conditions (Fig. 2).
Recombinant SP-A competes for binding to the high affinity receptor for SP-A Studies from this laboratory have demonstrated that the inhibition of phospholipid secretion by SP-A is mediated, at least in part, by the binding of SP-A to a high-affinity cell surface receptor on isolated type II cells [10]. Binding of radiolabeled native SP-A to a cell surface receptor is readily demonstrated on whole alveolar type II cells in primary culture [10]. A cell-free receptor-binding assay has been developed, which demonstrates binding characteristics and kinetics that are indistinguishable from the intact cell surface receptor (manuscript in preparation). This technique greatly reduces the quantity of SP-A required for analysis of receptor binding. We found that recombinant SP-A effectively competes with 125I-SP-A for binding to isolated plasma membrane from type I! cells (Fig. 4). It is apparent from the data in this figure that the activity of recombinant SP-A was equal to that of native SP-A in competing for high-affinity binding. Recombinant SP-A inhibits the secretion of phospholipid from isolated type H cells Several laboratories have demonstrated that native SP-A inhibits the stimulated secretion of phospholipid from isolated alveolar type II cells [7-9]. We studied whether the recombinant SP-A synthesized in CHO
i 10
i 15
Protein (gg) Fig. 4. Competition for cell-free binding of native and recombinant SP-A to isolated plasma membranes from alveolar type II cells. Plasma membrane from isolated alveolar type II cells was purified on sucrose density gradients and washed as described in Materials and Methods. A 2-#g (protein) aliquot of plasma membrane was incubated with t25I-SP-A (1 # g / m l ) and varying amounts of unlabeled native and recombinant SP-A. The binding of radiolabeled SP-A to plasma membrane was determined by filter-binding methods. The data show both the specific and the non-specific components of binding.
cells was capable of inhibition of phospholipid secretion. Native SP-A reduced the TPA-stimulated secretion of [3H]phosphatidylcholine to half the maximal levels at a concentration of 0.1 gg of SP-A/ml and to below basal levels at I #g of SP-A/ml. We found that the recombinant product also inhibited the TPA-stimulated secretion of surfactant phospholipids from alveolar type II ceUs in an identical manner. The recombinant SP-A product was equivalent to native SP-A as an inhibitor of surfactant lipid secretion (Fig. 5).
r.
100
.o (J O (n
75
% ~ N a t - - - ..... - o
Recombinant
Ire
"10 t~
"5
50
,
