Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Pulmonary surfactant is a complex mixture of phospholipids and proteins at the air-water interface of the alveolus that decreases surface tension. Surfactant protein B (SP-B) is one of the constituents that plays a critical role in the biophysical properties of surfactant. The importance of SP-B was emphasized recently through the discovery of a genetic deficiency of SP-B in humans that is associated with congenital alveolar proteinosis and results in neonatal respiratory failure and death (1). Targeted deletion of the SP-B gene in mice leads to a similar phenotype that causes respiratory failure in newborns (2). Other features of the SP-B deficiency common to humans and experimental animals include aberrant processing and secretion of one of the other surfactant proteins, pro SP-C (3), and lack of characteristic surfactant ultrastructure features such as tubular myelin and lamellar bodies (2).

SP-B is synthesized by alveolar type II cells and Clara cells (nonciliated bronchiolar epithelial cells) (5). The primary translation product of human SP-B messenger RNA (mRNA) has been identified as a 40-kD preproprotein containing 381 amino acid residues (6). Mature surfactant protein B is a 79-aa amphipathic peptide with seven cysteine residues (9) which form three disulfide bridges. The high-molecular-weight preprotein is processed by cleavage of signal peptide in the endoplasmic reticulum and proteolysis of amino- and carboxyl-terminal peptides (10). Sequence comparison of the 79 amino acids of human, canine, porcine, and bovine SP-B peptide revealed a high degree of identity (11), suggesting evolutionary conservation and importance of these structures to its function.

In each of these species, the expression of SP-B shows the same degree of cell specificity (5, 12, 13). Although the generation of multiple mRNA isoforms derived from many single-copy genes has been well recognized (14), with the exception of rabbit SP-B, only a single normal SP-B mRNA form has been reported to date. In the rabbit SP-B, two isoforms, one with and one without an insertion in the 3' untranslated region, were identified (15). Low-abundance multiple SP-B transcripts were also identified in an SP-B-deficient infant (16), raising the possibility of alternative splicing of human SP-B precursor RNA. In this study, we report the identification of an alternatively spliced SP-B mRNA in normal murine lung but not in human, either from normal subjects or patients with several pulmonary disorders. The alternatively spliced SP-B mRNA lacks 69 base pairs (bp) in the coding region of the active peptide. Quantitation of the SP-B mRNA isoforms showed expression of both isoforms with ratio of short/long equaling approximately 0.3 in all strains of adult and fetal mice surveyed, adult rat, and rabbit. In the mouse, evidence was obtained for translation of the two SP-B mRNA isoforms.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Animal Sources and RNA Isolation

All adult mice were originally obtained from Jackson Laboratory (Bar Harbor, ME). Total RNA was isolated from snap-frozen mouse, rat, and rabbit lung tissues by using the guanidine thiocyanate-CsCl method (17). Rat lung was obtained from adult pathogen-free Sprague-Dawley rats. Rabbit lung was obtained from adult New Zealand White rabbits. Human lung was obtained at the time of transplant or biopsy and RNA prepared as described (1).

cDNA Amplification

Reverse transcription (RT) was performed using a cDNA synthesis kit (Invitrogen, San Diego, CA) with oligo (dT) at 42°C for 1 h. The products were further amplified by polymerase chain reaction (PCR) with KlenTaq DNA polymerase (Dr. Wayne Barnes, Washington University, St. Louis, MO). Thirty-five cycles of reactions were performed, with denaturing at 94°C for 1 min, annealing at 60°C for 2 min, and extension at 72°C for 2 min. The final extension was at 72°C for 10 min. The oligonucleotide primers used to generate SP-B complementary DNA (cDNA) clones are shown in Table 1.

DNA Cloning and Sequencing

PCR products of mouse SP-B cDNA were treated with T4 DNA polymerase (Fisher Scientific, Pittsburgh, PA), extracted with chloroform, and precipitated by ethanol. The fragments were purified from agarose gel by QIAEX Gel Extraction Kit (QIAGEN), and ligated to pBluescript II plasmid DNA (Stratagene, La Jolla, CA) linearized with Sma I. Recombinant clones were identified after transformation and DNA-purified by QIAGEN Plasmid Kit (QIAGEN, Santa Clarita, CA). Sequencing of double-strand DNA was performed on a 373A DNA Sequencer (PE Applied Biosystems, Norwalk, CT) with oligonucleotides complementary to the mouse SP-B cDNA.

Ribonuclease Protection Assays

A fragment of mouse SP-B cDNA (599-866) spanning exons 6 and 7 (Figure 1A) was subcloned into pBluescript (Stratagene) for the synthesis of antisense RNA probes. Three human SP-B cDNA fragments (1-782, 659-1427, and 1152-1947) were generated by RT-PCR as described previously (16). A human SP-B cDNA clone (1-1749) was subsequently constructed by ligating RT-PCR fragments at appropriate restriction sites into pSP72 (Promega, Madison, WI). Reverse transcription of the plasmid linearized with EcoRI at the 3' end gave rise to IVT35 (in vitro transcript of normal SP-B coding strand) or IVT38 (in vitro transcript of 69 bp deletion SP-B coding strand); both were used as controls for the protection assay (Figure 2). The antisense human SP-B riboprobe was generated from a cDNA fragment (557-782) containing exons 6 and 7 (Figure 2A) in pSP72 (Promega) and linearized with Sma I. Two rabbit SP-B cDNA fragments (441-673, 1127-1430) from original type I cDNA were subcloned into pBluescript (Stratagene) and linearized with Xba I for the exon 7 construct (Figure 3A) and EcoRI for the 3' ultranslated region (UTR) construct (Figure 3B). The labeling reactions were performed in 20-µl solutions containing 0.5 mM rATP, rGTP, and rUTP; 12 µM rCTP; 5 µl [³²P]-CTP (800 Ci/mmol; DuPont, Wilmington, DE); and T7 RNA polymerase (Promega). The ribonuclease assays were carried out according to a standard method using the RPA II kit (Ambion, Austin, TX). Generally, 2 to 10 µg of total RNA were used for each assay and hybridization was performed at 42°C overnight. Following RNase A/T1 digestion at 37°C for 30 min, samples were precipitated and analyzed on 8 M urea 5% polyacrylamide gel along with a known sequencing ladder which served as the marker. Autoradiography was achieved by exposing the gel to X-ray film (Kodak, Rochester, NY) overnight at 70°C.

View larger version (16K): [in this window] [in a new window]   View larger version (51K): [in this window] [in a new window]   View larger version (67K): [in this window] [in a new window]   Figure 1.   RNase protection assay for murine SP-B mRNA. (A) Plasmid map of the SP-B cDNA construct used to generate antisense riboprobe with T7 RNA polymerase. (B) Lane 1, hybridization with 5 µg yeast RNA without RNase digestion, showing full-length probe; lane 2, same as lane 1 but with RNase digestion; lanes 3-6, hybridization with RNA isolated from adult mice lung tissues: BALB/cJ, SJL/J, B6SJLF1/J, C57B/6J, respectively. Arrows indicate the protected SP-B bands after RNase digestion ranging from 92 to 268 bp. Traces of incomplete digested probe appearing on top of each lane had little effect on quantitaion of the protected fragments. (C) Ratio of short/long SP-B isoform in mice at different developmental stages. Each bar represents mean value plus standard error (SE) of relative intensity of protected fragments (92 bp over 268 bp and 107 bp over 268 bp), adjusted for the number of radiolabeled nucleotides in each fragment. Four to six data points resulting from densitometry scanning of independent assays are included in each group. The standard error was calculated using StatWorksTM cricket software (Philadelphia, PA). All nonadult mice are B6SJLF1/J. E indicates days of embryogenesis.

View larger version (40K): [in this window] [in a new window]   Figure 2.   RNase protection assay for human SP-B mRNA. (A) Antisense human SP-B riboprobe and predicted protected fragments. (B) Lane 1, yeast RNA without RNase digestion, showing full-length probe; lane 2, same as lane 1 but with RNase digestion; lane 3, IVT35 (in vitro transcript of full-length human SP-B; see MATERIALS AND METHODS); lane 4, IVT38 (in vitro transcript of human SP-B containing 69-bp deletion in exon 7); lane 5, RNA from SP-B-deficient patient; lane 6, RNA from normal human donor lung tissues. The top band appearing in all lanes resulted from incomplete digestion of the probe.

View larger version (22K): [in this window] [in a new window]   View larger version (30K): [in this window] [in a new window]   Figure 3.   RNase protection assay for rabbit SP-B mRNA. (A) Top: antisense rabbit 3' UTR SP-B riboprobe and predicted protected fragments; bottom: lane 1, yeast RNA without RNase digestion, showing full-length probe; lane 2, same as lane 1 but with RNase digestion; lanes 3 and 4, rabbit lung RNA (1:1,000 and 1:100 dilution for RNase digestion, respectively). (B) Top: antisense rabbit SP-B exon 7 riboprobe and predicted protected fragments. The triangle represents the presumed boundary of rabbit SP-B exons 6 and 7 based on human and mouse SP-B gene. Bottom: ribonuclease assay of rabbit lung RNA. Material in each lane is same as in (A).

Densitometer Scanning Analysis

After autoradiography, the film was scanned with an LKB UltraScan XL densitometer with peak width = 1, X-width = 4, and Y-step = 2. The intensity of each band corresponding to protected RNA fragments was calculated by addition of the respective peak areas. The relative abundance of different forms of the message was determined after accounting for the amount of radiolabeled nucleotide in each protected fragment.

Generation of Anti-SPB Antiserum #1768

SP-B was isolated from sheep surfactant using methods previously established for the isolation of canine and rat SP-B (7, 18). Briefly, a single-phase butanol extraction was performed and the soluble phase was separated on a LH-60 column (Pharmacia, Piscataway, NJ) eluted isocratically with acidified chloroform:methanol (18). Fractions containing SP-B were detected by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), pooled, and used for immunization of New Zealand White rabbits using standard protocols. The resulting antiserum (#1768) had no reactivity against SP-A or SP-C. An IgG fraction was isolated from whole serum using a protein A affinity column (Pierce, Rockford, IL).

Production and Purification of SP-B Fusion Protein

Mouse mature SP-B fusion protein was produced in bacteria by cloning a cDNA fragment (599-866) corresponding to the region encoding mouse mature SP-B peptide into pMAL-c2 expression vector (New England Biolabs, Inc., Beverly, MA) to create an in-frame fusion maltose binding protein (MBP). Fusion protein was induced upon addition of isopropyl--D-1-thiogalactopyranoside and affinity-purified according to the manufacturer's instructions. The purified protein reacts with both anti-MBP antibody and anti-SP-B antibody by Western blot (data not shown).

Absorption of Anti-SP-B Antiserum

Nunc Immunotube was coated with purified MBP-SP-B fusion protein in phosphate-buffered saline (PBS) overnight at room temperature. The tube was rinsed with PBS 3 times after coating and blocked with 3% bovine serum albumin (BSA) in PBS for 2 h at 37°C. After rinsing with PBS, 1:500 diluted antiserum (#1768) was added to the tube and incubated overnight at room temperature. For control, the same dilution of the antiserum was added to a tube blocked with BSA but not coated with fusion protein and incubated for the same period.

Electrophoresis and Immunoblotting

B6SJLF1 mouse lung tissue was homogenized in ice-cold Tris saline buffer with protease inhibitors phenylmethylsulfonyl fluoride and leupeptin. After centrifugation, the supernatant was mixed with SDS sample buffer containing -mercaptoethanol and boiled for 3 min. Protein samples were run on 15% SDS-PAGE overnight at low voltage for better resolution and then transfered to Hybond enhanced chemiluminescence (ECL) membrane (Amersham, Arlington Heights, IL) using a transfer buffer of 20 mM Tris, 150 mM glycine, 20% methanol, and 0.1% SDS, pH 8.3. The membrane was blocked with 5% blocking reagent (Amersham) and incubated with the primary antibody (1:500 dilution) for 1 h at room temperature. Subsequently, 1:2,000 diluted antirabbit IgG antibody conjugated with horseradish peroxidase (Amersham) was used as secondary antibody. After washing, the immunoreactive proteins were detected using enhanced chemiluminescence reagents (Amersham).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Sequence Analysis of Murine SP-B cDNA

The coding sequence of murine SP-B was amplified by RT-PCR and cloned into pBluescript plasmid vector. Sequencing analysis revealed two types of clones: one was identical to the published mouse SP-B cDNA sequence (19); the other was similar except for the lack of 69 bp at the beginning of exon 7 (Figure 4). This cDNA segment contains a Kpn I restriction site, the lack of which can be used as an indicator for the deletion. In fact, when cloned RT-PCR fragments were digested with Kpn I restriction enzyme, only 50 to 70% were cut, suggesting that transcripts lacking the 69-bp segment in exon 7 were relatively abundant.

View larger version (12K): [in this window] [in a new window]   Figure 4.   Mouse SP-B gene structure and alternatively spliced mRNA. Shown in the middle is a partial map of mouse SP-B gene (according to Bruno [25]). Open boxes represent exons and lines represent introns. The long mRNA form as well as mature SP-B peptide sequence is shown on top. Both amphipathic helices are underlined, and the deleted sequence is boxed. The short mRNA form (which has a 69-bp deletion) and its predicted peptide sequence are shown at the bottom. The arrow denotes the junction where deletion occurs.

RT-PCR Analysis of SP-B mRNA

To assay the prevalence of alternatively spliced SP-B mRNA, we undertook RT-PCR analysis in RNA from multiple animals in five strains of mice to exclude a rare event. Specific oligonucleotides were designed to generate the cDNA fragment spanning a segment of the coding sequence containing exon 7 (Table 1). SP-B cDNA fragments were analyzed by agarose gel electrophoresis (Figure 5A, lanes 1-4). In all cases, two distinct major products of the predicted sizes were observed. RT-PCR of rat (Figure 5A, lane 5) and rabbit (data not shown) lung RNA also generated the same two-banded pattern similar to that in mice, suggesting the short isoform of SP-B message also exists in rat lung. Restriction digestion of rabbit RT-PCR products indicates one isoform with and the other without a Kpn I restriction site which is in the first 69 bp of exon 7 (data not shown).

View larger version (61K): [in this window] [in a new window]   Figure 5.   RT-PCR analysis of murine SP-B mRNA. (A) RT-PCR with oligonucleotides complementary to mouse (and rat) SP-B cDNA (see MATERIALS AND METHODS). RNA was isolated from adult mouse lung tissues. Lane 1, Swiss Black mice; lane 2, C57/B6 mice; lane 3, same strain as in lane 2 but independent isolation; lane 4, B6SJLF1/J mice; lane 5, rat. FL indicates PCR product from full-length message; AS indicates PCR product from alternatively spliced message. (B) RT-PCR with mouse actin oligonucleotides as control. Lanes 1-4 correspond to same RT material as in (A). Lane 5 is a negative control.

Ribonuclease Protection Assay of Murine SP-B

To quantitate the relative abundance of the 69-bp deleted message of mouse SP-B, RNase protection assays were conducted. The antisense riboprobe was generated as described (Figure 1A). The full-length probe is 335 bp. Protection of normal message yielded a protected band at 268 bp, whereas protection of the 69-bp deleted message yielded two bands, at 107 bp (in exon 7) and 92 bp (in exon 6). The pattern of protected SP-B mRNA is similar among all the mice tested (Figure 1B), suggesting that the alternative splicing event is common if not universal in this species. Quantitation of the relative abundance of the two message forms was achieved by densitometry scanning. In five strains of adult mice, the ratio of SP-B short to SP-B long mRNA is approximately 30%. There were no significant differences among these five strains of mice. The alternative splicing of SP-B mRNA was also observed at early developmental stages (Figure 1C). The ratio of short/long SP-B isoform is slightly greater in all prenatal samples, but the difference from adult ratios is relatively small, suggesting little if any developmental change in the generation of the SP-B short isoform.

Ribonuclease Protection Assay of Human and Rabbit SP-B

Aberrant splicing of SP-B has been observed in human patients with SP-B deficiency (16). This event was thought to be associated with extremely low steady-state concentrations of SP-B mRNA in the context of a frame-shift mutation (20). In order to ascertain if this 69-bp deleted SP-B mRNA isoform was expressed in humans, ribonuclease protection assay was carried out in 17 human lung RNA samples (data not shown), including five normal lung donors; several patients with different pulmonary disorders such as pulmonary hypertension, pulmonary vein stenosis, and pulmonary vascularities; several patients with congenital alveolar proteinosis but not SP-B deficiency; and a previously described SP-B-deficient patient (16). No SP-B mRNA containing the 69-bp deletion in exon 7 was detected in human lung tissue by this assay (Figure 2).

A similar analysis of rabbit SP-B mRNA showed the previously reported SP-B mRNA isoforms (15) generated by alternative splicing in the 3' untranslated region (Figure 3A) and, in addition, a 69-bp deletion in exon 7 (Figure 3B). The 3'UT variants occur at a ratio of 3'UT long/ 3'UT short (Type II/Type I) 0.78, in rough agreement with an earlier estimate (0.66). The exon 7 short/long transcript ratio is about 0.32.

Western Blot Analysis of Mouse SP-B Protein

To see if the alternatively spliced SP-B mRNA can be translated to protein product, homogenated mouse lung tissue was analyzed by Western blot. Using a sensitive ECL method, antibody #1768 detected two bands at the MW 40-42 kD which corresponds to the precursor form of mouse SP-B protein (Figure 6, lanes 1 and 2). The smaller protein product corresponds to an apparent size that would be predicted from the transcript with the deletion of 69-bp. The intensity of both bands was greatly reduced after absorption of the antibody against purified fusion SP-B protein (Figure 6, lanes 3 and 4), suggesting that the bands are specific.

View larger version (41K): [in this window] [in a new window]   Figure 6.   Western blot analysis of mouse SP-B protein. Equal amounts of proteins from mouse lung homogenates were loaded on adjacent lanes of 15% SDS-polyacrylamide gel, separated by electrophoresis, and then transferred to nitrocellulose. Stripes of the membrane were subjected to Western blot analysis using either control anti-SP-B antibody (lanes 1 and 2) or absorbed antibody (lanes 3 and 4). Bands were revealed by ECL. The position of precursor SP-B protein and mature protein are marked by arrows.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

Alternative splicing of pre-messenger RNA has been recognized in recent years as a mechanism for regulating gene expression and for generating isoform diversity (14). Many examples have been reported. In most cases, alternative splicing gives rise to protein isoforms sharing extensive regions of identity but varying in specific domains, thus allowing fine modulation of protein function and economy of information storage in the genome. Splicing variants of surfactant protein mRNA have been reported for human SP-A, which includes alternative 5' untranslated exons (21). Two forms of rabbit SP-B cDNAs were also cloned. These differ by 69-bp in the 3' untranslated region (15).

In this study we demonstrated two isoforms of murine pulmonary SP-B mRNA which are likely the result of alternative splicing. One isoform corresponds to the published cDNA sequence; the other lacks the first 69-bp of exon 7. The exon 7 short message is present in five strains of mice as well as in rat and rabbit, and in similar relative amounts in mouse embryos. The 3' splice-site consensus is YAG/G (Y is a pyrimidine) (14). A comparison of the splice junction in the exon 6/7 boundary for the two isoforms shows sequence similarity in the splice acceptor sites (Table 2), including identity of the critical AG doublet, which are absolutely conserved bases at the intron/exon boundary (14). However, sequence differences near the splice acceptor sites do not account for species differences because humans and mice have identical junction sequences (Table 2) yet the short SP-B is found in mice but not in humans. Likewise, rabbit and rat differ from mouse yet all three species express approximately the same abundance of the short isoform. Other elements can also contribute the species variation, such as sequences flanking the splice site and/or higher order RNA structure (22). The size of the intron 6 can also be considered as a possible variable determining the difference between murine and human SP-B multiple transcripts.

One can only speculate why the short isoform of murine SP-B mRNA was not identified in the initial cloning of the cDNA (19). We considered the possibility of a significant variation in expression of the short isoform among different mouse strains. However, this is an unlikely explanation because we found a similar ratio of SP-B short/long in BALB/cJ, a strain genetically quite close to the BALB/ cByJ used to generate the cDNA library (19). One possibility is that shorter clones or clones not corresponding to the same restriction map were not isolated and sequenced. It is possible that this alternative splicing of SP-B has a function at the mRNA level; i.e., for regulating the mRNA stability or translation efficiency or at the protein level; i.e., alteration of protein function. Presently, there is no evidence for regulation of the ratios of the two SP-B isoforms.

The 23 amino acids encoded by the region deleted in one of the SP-B isoform are within the mature peptide. When translated, and if normally processed, the resulting SP-B protein isoform would likely have a different secondary and tertiary structure from SP-B long form. The presence of four internal repeats with a conserved periodicity of cysteines in each repeat (23) probably contributes to a characteristic folding of three domains in SP-B (18). Thus, loss of 23 amino-acid residues which include two cysteines would not only shorten the distance between the two amphipathic helixes but also disrupt the formation of one internal disulfide bond. This change in higher-order structure could then affect the secretion, localization, and/or function of SP-B protein. Previous biosynthesis studies of SP-B in rat type II cells (18, 24) showed heterogeneity in the precursor form of rat SP-B protein. Moreover, in the studies by Weaver and Whitsett, the pattern of multiple bands persists even after deglycosylation (24). In the present study, a slightly lower MW mouse SP-B precursor was observed (Figure 6) and its identity established by blocking with purified fusion SP-B protein. These data suggest that the alternatively spliced SP-B mRNA is in fact translated. However, the fate of the precursor protein product cannot be determined with certainty. Thus, the biologic consequences of this short form of SP-B have yet to be defined.