Sequences of a hairpin structure in the 3′-untranslated region mediate regulation of human pulmonary surfactant protein B mRNA stability

Abstract

The ability of pulmonary surfactant to reduce alveolar surface tension requires adequate expression of surfactant protein B (SP-B). Dexamethasone (DEX, 10⁻⁷ M) increases human SP-B mRNA stability by a mechanism that requires a 126-nt-long segment (the 7.6S region) of the 3′-untranslated region (3′-UTR). The objective of this study was to identify sequences in the 7.6S region that mediate regulation of SP-B mRNA stability. The 7.6S region was found to be sufficient for DEX-mediated stabilization of mRNA. Sequential substitution mutagenesis of the 7.6S region indicates that a 90-nt region is required for DEX-mediated stabilization and maintenance of intrinsic stability. In this region, one 30-nt-long element (002), predicted to form a stem-loop structure, is sufficient for DEX-mediated stabilization of mRNA and intrinsic mRNA stability. Cytosolic proteins specifically bind element 002, and binding activity is unaffected whether proteins are isolated from cells incubated in the absence or presence of DEX. While loop sequences of element 002 have no role in regulation of SP-B mRNA stability, the proximal stem sequences are required for DEX-mediated stabilization and specific binding of proteins. Mutation of the sequences that comprise the proximal or distal arm of the stem negates the destabilizing activity of element 002 on intrinsic SP-B mRNA stability. These results indicate that cytosolic proteins bind a single hairpin structure that mediates intrinsic and hormonal regulation of SP-B mRNA stability via mechanisms that involve sequences of the stems of the hairpin structure.

pulmonary surfactant is a complex, developmentally regulated lipoprotein that reduces alveolar surface tension and provides innate immunity to the lung (5, 18, 36). Surfactant is produced and secreted by alveolar type II epithelial cells of the lung and coats the lining of the alveolus (28). While the lipid component of surfactant, largely dipalmitoylphosphatidylcholine, is responsible for reduction of alveolar surface tension, the protein component of surfactant is critical in maintenance of surfactant function in the lung (29). Surfactant protein B (SP-B) is a small, hydrophobic protein that associates with surfactant lipids and functions to accelerate surfactant film formation and stabilize the surfactant monolayer (29). Premature birth precludes adequate production of surfactant and can lead to respiratory distress syndrome (RDS), the leading cause of neonatal morbidity and mortality in developed countries (1). In clinical situations where a premature delivery is imminent, antenatal glucocorticoids are given to accelerate fetal lung development and augment expression of pulmonary surfactant (11).

SP-B is essential in the function of surfactant to reduce alveolar surface tension in the lung (14). SP-B is a highly processed, positively charged 8-kDa dimeric protein expressed by alveolar type II epithelial cells (24, 35). Transcription of the SP-B gene is developmentally regulated; SP-B mRNA is detectable as early as 13 wk of gestation in human lung tissue (22). However, mature SP-B protein is not detectable until 31 wk of gestation (27). Inadequate expression of SP-B due to premature birth leads to lung dysfunction and RDS in neonates (1, 14). Expression of human SP-B is enhanced by prenatal administration of glucocorticoids, such as dexamethasone (DEX) (22). DEX enhances expression of human SP-B in a dose-dependent manner through a combination of increased transcription of the SP-B gene and increased stabilization of SP-B mRNA (2, 32, 34). However, the molecular mechanisms responsible for enhanced expression remain largely undefined. Our group is focused on describing and understanding the molecular mechanisms responsible for posttranscriptional regulation of SP-B mRNA stability.

Complex mechanisms underlie regulation of eukaryotic mRNA stability; in particular, the mechanisms involving steroid hormones, such as glucocorticoids, are highly variable (9, 10, 13, 17, 30). Previously, we reported that the effect of DEX to stabilize human SP-B mRNA requires a segment of the SP-B mRNA 3′-untranslated region (3′-UTR) called the 7.6S region (16). Cytosolic proteins derived from human alveolar epithelial cells specifically bind this 126-nt-long segment. DEX-induced stabilization of human SP-B mRNA does not require the activity of the glucocorticoid receptor or ongoing protein synthesis (31, 32). Those results suggest that the mechanism responsible for DEX-induced stabilization of SP-B mRNA is mediated by posttranscriptional or nongenomic effects of glucocorticoids on proteins that bind a specific region of the human SP-B mRNA 3′-UTR.

We previously described a reproducible, easily assayable plasmid-based method in which steady-state levels of SP-B mRNA reflect changes in mRNA stability (31). Hormonal regulation of human SP-B mRNA stability was characterized in cells transfected with plasmids encoding full-length SP-B cDNA and the green fluorescent protein (GFP) under transcriptional control of separate, identical promoters. With steady-state levels of GFP mRNA used as a reference, the effect of cell culture environment or deletion of SP-B mRNA sequences on mRNA stability could be determined by assay of steady-state levels of SP-B mRNA in transfected cells. In the present study, we have adapted this assay to functionally identify elements of the human SP-B mRNA 3′-UTR that are sufficient to impart DEX regulation to the stability of a heterologous reporter mRNA. We found that the 7.6S region of the human SP-B mRNA 3′-UTR is sufficient to impart DEX-mediated stability to a reporter mRNA. Distinct sequences of this region are necessary for proper intrinsic and DEX-regulated mRNA stability. One 30-nt-long element (002), predicted to form a stem-loop structure, was found to specifically bind cytosolic proteins and to mediate intrinsic and DEX-mediated stabilization of SP-B mRNA through specific mRNA sequences and structures, respectively. The results of the study suggest that the mechanism by which DEX stabilizes SP-B mRNA is mediated by a hairpin element involved in maintenance of intrinsic stability of SP-B mRNA.

MATERIALS AND METHODS

Materials.

DEX (catalog no. D1756) was purchased from Sigma Chemical (St. Louis, MO) and diluted in ethanol.

Plasmid constructs.

Standard cloning, PCR techniques, and bacterial manipulations were used to construct the plasmids used in this study, and many were derived from pCMVGFP-hspB:N (31). The expression cassettes used in heterologous reporter assays are shown in Fig. 1A. pCMVGFP-RFP-MS2 was assembled by replacement of the SP-B sequences of pCMVGFP-hspB:N with sequences coding for red fluorescent protein (RFP), derived from pIRES2-DsRed2 (Clontech, Mountain View, CA). The 3′-UTR of the expressed mRNA consists of the human β-globin 3′-UTR derived from pTET-BBB with a 132-nt-long PCR product of cassette containing three MS2 RNA binding site hairpins (37, 39). These hairpin loops were included as a means to affinity-purify proteins that interact with the SP-B 3′-UTR and to tag the mRNA for subcellular localization. pCMVGFP-RFP-MS2:hspB3′-UTR was assembled by replacement of the β-globin 3′-UTR of pCMVGFP-RFP-MS2 with the full-length human SP-B 3′-UTR from pCMVGFP-hspB:N. pCMVGFP-RFP-MS2:7.6S was assembled by insertion of the 126-nt region of the human SP-B 3′-UTR, called 7.6S, into the β-globin 3′-UTR of pCMVGFP-RFP-MS2. pCMVGFP-RFP-MS2:002 was assembled by ligation of the 30-nt-long element 002 sequence, fabricated by hybridization of complementary oligonucleotides, into the β-globin 3′-UTR of pCMVGFP-RFP-MS2.

Fig. 1.

Regions of surfactant protein B (SP-B) mRNA 3′-untranslated region (3′-UTR) are sufficient to impart dexamethasone (DEX)-mediated stabilization to a heterologous mRNA. A: schematic of the expression cassette of pCMVGFP-RFP-MS2 and derivatives, shown in scale. Black boxes represent identical, independent cytomegalovirus (CMV) early 1 promoters, and arrows denote direction of transcription. Gray arrows represent coding regions for red and green fluorescent protein (RFP and GFP). Open boxes represent 3′-UTRs of rabbit β-globin mRNA (β-globin 3′-UTR) and SV40 large T mRNA (SV40pA). Position of MS2 binding sites is denoted by MS2. Sequences of human SP-B (hSP-B) 3′-UTR with appropriate restriction sites are denoted by thick lines, and bona fide polyadenylation site is indicated (AATAAA). Gray box indicates 30-nt-long element 002. B: effect of DEX on stability of RFP mRNA containing segments of hSP-B 3′-UTR. Steady-state levels of RFP and GFP mRNA were determined in A549 cells that were transfected with plasmids in A and incubated in the absence or presence of DEX (10⁻⁷ M). RFP-to-GFP mRNA ratio was determined in each sample. Average ratio in untreated samples was set as 1, and RFP-to-GFP mRNA ratio of DEX-treated samples was normalized to this average. Values (means ± SE) are normalized RFP mRNA levels in DEX-treated samples relative to levels in untreated samples (n ≥ 7 from ≥2 independent experiments). *P < 0.01 vs. control. #P < 0.01 vs. RFP-MS2 with DEX. C: effect of various segments of hSP-B 3′-UTR on intrinsic stability of RFP mRNA. Steady-state levels of RFP and GFP mRNA were determined in A549 cells that were transfected with plasmids in A. RFP-to-GFP mRNA ratio was determined in each sample. Average ratio in pCMVGFP-RFP-MS2 samples was set as 1, and RFP-to-GFP mRNA ratio of other samples was normalized to this average. Values (means ± SE) are normalized RFP mRNA levels relative to levels from pCMVGFP-RFP-MS2 samples (n ≥ 7 from ≥2 independent experiments). *P < 0.03 vs. RFP-MS2.

The plasmids used in scanning mutagenesis of the 7.6S region of the human SP-B mRNA 3′-UTR were also derived from pCMVGFP-hspB:N. Substitution mutagenesis of 30-nt segments of the proximal region of the human SP-B mRNA 3′-UTR was performed by recombinant PCR (15, 25, 26). Two 33-nt-long anchor PCR primers were designed to include the sequences surrounding NsiI and XbaI restriction endonuclease sites at the 5′ and 3′ ends of the SP-B 3′-UTR, respectively, in pCMVGFP-hspB:N. Two separate PCRs were performed using each primer with an internal mismatched primer. These 45-nt-long internal primers contained 15 nt of sequences complementary to the SP-B DNA at the 3′ end of the primer and 30 nt of the sequence 5′-GAATTCGAGCTCGGTACCCGGGGATCCTCT-3′ or 5′-AGAGGATCCCCGGGTACGAGCTCGAATTC-3′ at the 5′ end of the primer. These sequences were derived from multiple cloning sites of the common cloning plasmid pUC18. The substitutions began at nucleotide 7 downstream from the translational stop site (TGA) of SP-B. The resulting products were purified, denatured, and used as primers in extension. PCR was performed using the anchor primers. The secondary DNA product of full-length SP-B 3′-UTR could only be the result of hybridization and elongation through the substituted regions of the first DNA products. The final products were verified by sequencing, digested with NsiI and XbaI, and used to replace the wild-type SP-B 3′-UTR in pCMVGFP-hspB:N. The positions of the substitutions are depicted in Fig. 2A.

Fig. 2.

Sequential scanning mutagenesis of proximal hSP-B 3′-UTR indicates that distinct elements are necessary for DEX-mediated stabilization of SP-B mRNA and can act to reduce intrinsic stability. A: schematic of pCMVGFP-hspB:N and locations of substituted sequences of SP-B 3′-UTR shown in scale. Black boxes represent identical, independent CMV promoters, and arrows denote direction of transcription. Gray arrows denote coding region for SP-B (hSP-B) and GFP; thin black lines represent SP-B 3′-UTR. Open box represents 3′-UTR of SV40 large T mRNA (SV40pA). AATAAA indicates position of polyadenylation signal of SP-B mRNA. Dashed lines flank the region that was subjected to substitution mutagenesis, and positions of 30-nt substituted elements of 3′-UTR are indicated by open boxes. B: effect of DEX on steady-state levels of SP-B mRNA in A549 cells transfected with plasmids in A. SP-B-to-GFP mRNA ratio was determined in each sample. Average ratio in untreated samples was set as 1, and ratio from DEX-treated samples was normalized to this average. Values (means ± SE) are normalized SP-B levels in DEX-treated samples relative to levels in untreated samples. *P < 0.03 vs. control. C: effect of deletions of SP-B mRNA 3′-UTR on intrinsic stability of SP-B mRNA in A549 cells transfected with plasmids in A. SP-B-to-GFP mRNA ratio was determined in each sample. Average ratio in pCMVGFP-hspB:N samples was set as 1, and ratio from other samples was normalized to this average. Values (means ± SE) are normalized SP-B levels (n ≥ 7). *P < 0.01 vs. hspB.

The plasmids used in functional mutation analysis of element 002 of the human SP-B mRNA 3′-UTR were also derived from pCMVGFP-hspB:N. Site-directed mutagenesis of the element 002 sequences was performed using recombinant PCR, as described above. The specific mutations are shown in Fig. 5A. pCMVGFP-hspBΔ002 was created by recombinant PCR using primers that resulted in deletion of the element.

Fig. 5.

Distinct portions of the element 002 stem-loop structure are required for proper DEX-mediated and intrinsic SP-B mRNA stability. A: sequence of site-directed mutations of the element 002 structure. Boxes indicate limits of sequences that form the stem structure. Underlined sequences indicate mutated sequences. Schematic shows plasmids containing the mutated element 002 sequences, as well as a plasmid deleted in element 002 (pCMVGFP-hspBΔ002). B: effect of mutations of element 002 on DEX-mediated stabilization of SP-B mRNA. Steady-state levels of SP-B and GFP mRNA were determined in A549 cells that were transfected with plasmids in A and incubated in the absence or presence of DEX (10⁻⁷ M). SP-B-to-GFP mRNA ratio was determined in each sample. Average ratio in untreated samples was set as 1, and SP-B-to-GFP mRNA ratio of DEX-treated samples was normalized to this average. Values (means ± SE) are normalized SP-B mRNA levels in DEX-treated samples relative to levels in untreated samples (n ≥ 7 from ≥2 independent experiments). *P < 0.01 vs. control. C: effect of various mutations of element 002 on intrinsic stability of SP-B mRNA. Steady-state levels of SP-B and GFP mRNA were determined in A549 cells that were transfected with plasmids in A. SP-B-to-GFP mRNA ratio was determined in each sample. Average ratio in the pCMVGFP-hspB:N samples was set as 1, and SP-B-to-GFP mRNA ratio of other samples was normalized to this average. Values (means ± SE) are normalized SP-B mRNA levels relative to levels in pCMVGFP-hspB:N samples (n ≥ 7 from ≥2 independent experiments). *P < 0.03 vs. hspB.

Cell culture.

Human lung epithelial A549 cells (CCL-185, American Type Culture Collection) (20) were cultured in Weymouth's MB 752/1 medium (Invitrogen, Carlsbad, CA) containing fetal bovine serum (10% vol/vol) in a humidified incubator at 37°C with 5% CO₂. When used for investigations, the cells were preincubated in medium containing 2% charcoal-stripped serum (Invitrogen) for 24 h to prevent undue influence of serum-derived steroid hormones.

Cell transfection for mRNA stability assays.

Stability of specific mRNAs was assayed as described previously (31). Cells were transfected with plasmid DNA according to the protocol prescribed by the Lipofectamine Plus reagent (Invitrogen) with slight modifications; 4 μg of plasmid combined with 12 μl of Lipofectamine reagent and 12 μl of the Plus reagent were utilized to transfect 60-mm plates. A typical assay was performed as follows: cells were transfected with plasmid DNA for 4 h and allowed to recover overnight in medium containing 2% charcoal-stripped serum. DEX was added to the cells 18 h after transfection, and incubation continued for 36 h; then RNA was isolated for analysis.

Isolation of RNA.

RNA was isolated and purified from the cells using TRIzol reagent (Invitrogen). The RNA concentration was determined by measurement of absorbance at 260 nm.

Northern analysis of mRNA.

Northern analysis of SP-B, RFP, and GFP mRNA expression was performed as described in detail previously (3). Total RNA (20 μg) was electrophoresed, transferred to a nylon membrane (Zeta-Probe, Bio-Rad Laboratories, Hercules, CA), and probed using a radiolabeled GFP, RFP, or human SP-B cDNA. Signal was visualized and quantified using a phosphorimager (Storm 840, Amersham Biosciences, Piscataway, NJ).

RNA EMSAs.

RNA EMSAs (REMSAs) were performed as previously described (16). Cells were grown as described above in the absence or presence of DEX (10⁻⁷ M). Specific radiolabeled RNA probes were made from linearized plasmids with the Megascript T7 kit (Ambion, Austin, TX). Cytosolic proteins isolated from the cells were complexed with probes, and complexes were visualized after electrophoresis by autoradiography (37).

Data analysis.

At least two independent experiments were performed in each analysis. The data were analyzed using SigmaPlot (version 10, Systat Software, San Jose, CA). Differences between groups were assessed by Student's t-test. Values are means ± SE. Statistical differences were considered significant when P ≤ 0.05.

RESULTS

Human SP-B mRNA 3′-UTR is sufficient for DEX-mediated stabilization of mRNA.

In our previous study, we developed a plasmid-based expression system in which the steady-state levels of expressed SP-B mRNA reflect changes in SP-B mRNA stability (31). This assay method was found to recapitulate the magnitude of change in SP-B mRNA stability induced by DEX reported previously and demonstrated that stabilization requires elements of the SP-B mRNA 3′-UTR (16, 32). In the present study, the assay was adapted to determine the sufficiency of the SP-B mRNA 3′-UTR to mediate DEX stabilization of a heterologous reporter mRNA. A RFP expression cassette was fabricated as described in materials and methods and replaced the SP-B expression cassette of pCMVGFP-hspB:N (31) (Fig. 1A). To establish that the presence of DEX has no effect on the steady-state levels of reporter RFP mRNA, A549 cells were transfected with pCMVGFP-RFP-MS2 (Fig. 1A) and cultured in the absence or presence of DEX (see Experimental procedures). RNA was isolated and subjected to Northern analysis for visualization and quantitation of RFP and GFP mRNA, as described previously (31). The results of the assay (Fig. 1B) demonstrate that the steady-state levels of RFP mRNA are not significantly different in cells incubated in the absence or presence of DEX. These results indicate that reporter RFP mRNA can be used to determine if the presence of SP-B mRNA sequences can impart regulation of stability to a heterologous reporter RFP mRNA.

To determine if the human SP-B mRNA 3′-UTR is sufficient to mediate DEX stabilization of mRNA, the globin 3′-UTR in pCMVGFP-RFP-MS2 was replaced with the human SP-B 3′-UTR, resulting in pCMVGFP-RFP-MS2:hspB3′UTR (Fig. 1A). The effect of DEX on steady-state levels of RFP and GFP mRNA in cells transfected with this plasmid was determined as described above. The presence of DEX significantly increased steady-state levels of RFP mRNA compared with levels in cells incubated in the absence of DEX and with cells transfected with pCMVGFP-RFP-MS2 and incubated in the presence of DEX (Fig. 1B). These results demonstrate that the presence of the human SP-B 3′-UTR is sufficient to impart DEX-mediated stability to a heterologous reporter RFP mRNA.

The 7.6S region of the human SP-B mRNA 3′-UTR is sufficient for DEX-mediated stabilization of a heterologous mRNA.

Previously, we reported that a 126-nt-long segment of the human SP-B 3′-UTR, called the 7.6S region, is necessary for DEX-mediated regulation of SP-B mRNA stability (16). In the present study, the sufficiency of the 7.6S region to mediate DEX stabilization of mRNA was determined. The 7.6S region of the human SP-B 3′-UTR was inserted into the β-globin 3′-UTR of pCMVGFP-RFP-MS2, resulting in pCMVGFP-RFP-MS2:7.6S (Fig. 1A). A549 cells were transfected with the plasmid and cultured in the absence or presence of DEX (see Experimental procedures). The effects of DEX on steady-state levels of RFP and GFP mRNA were determined as described above. The presence of DEX significantly increased steady-state levels of RFP mRNA compared with levels in cells incubated in the absence of DEX and with cells transfected with pCMVGFP-RFP-MS2 and incubated in the presence of DEX (Fig. 1B). These results demonstrate that the 7.6S RNA segment is sufficient to impart DEX-mediated stability to a heterologous reporter RFP mRNA. The degree of change in steady-state levels of the RFP mRNA was not as great as that in the presence of the entire SP-B mRNA 3′-UTR, suggesting that other elements of the SP-B mRNA 3′-UTR may be involved in modulating the effect of DEX on stability.

Sequential scanning mutagenesis of the 7.6S segment of the SP-B mRNA 3′-UTR indicates that distinct elements are required for DEX-mediated stabilization of SP-B mRNA.

Previously, we reported that deletion of distinct segments of the human SP-B 3′-UTR abolishes DEX-mediated stabilization of SP-B mRNA stability (16, 31). In those studies, large regions (>200 bases) of the SP-B mRNA 3′-UTR were deleted. It has been reported that the spatial relationships of multiple mRNA elements to which proteins bind can be a critical factor in regulation of mRNA stability, and altering the spacing between the elements can potentially unduly influence regulation (12, 19, 33). To eliminate this possibility in our investigations, we sequentially substituted 30-nt-long elements of the proximal region of the human SP-B 3′-UTR with part of the sequence of the multiple cloning site of the cloning vector pUC18 (5′-GAATTCGAGCTCGGTACCCGGGGATCCTCTAGAGTC-3′). Substitution started 7 nt downstream of the stop codon of the SP-B coding region and included the region defined by 7.6S. A schematic of the expression cassette and relative locations and identities of the substitutions is shown in Fig. 2A. These substitutions were created by recombinant PCR mutagenesis of pCMVGFP-hspB:N (see materials and methods). A549 cells were transfected with the plasmids and cultured in the absence or presence of DEX, and the effects of DEX on steady-state levels of SP-B and GFP mRNA were determined (see Experimental procedures). The results of the assay are shown in Fig. 2B. Mutagenesis of many of the elements (021, 022, 001, 005, and 006) did not affect the ability of DEX to significantly increase steady-state levels of SP-B mRNA. However, the presence of DEX did not significantly increase steady-state levels of SP-B mRNA upon mutagenesis of elements 002, 003, and 004 of the 7.6S segment and element 007 or 008 of nearby segments. It seems that, upon mutagenesis of elements 002, 003, and 007, there is an effect of DEX to actively destabilize SP-B mRNA, although the reduction in SP-B mRNA expression elicited by mutagenesis of element 003 reaches statistical significance compared with its untreated control. These results suggest that not only do these elements actively participate in DEX-mediated stabilization of SP-B mRNA, they may act in maintenance of SP-B mRNA stability in the presence of DEX, perhaps by modulating another region of the SP-B mRNA that acts to destabilize it. Interestingly, these elements are consecutive, suggesting that a 90-nt region encompassed by elements 002, 003, and 004 and a 60-nt region encompassed by elements 007 and 008 are required for DEX-mediated stabilization of SP-B mRNA. Whether these elements can act individually or require interactions with the other elements is unclear.

Sequential scanning mutagenesis of the 7.6S segment of the SP-B mRNA 3′-UTR indicates that distinct elements act to reduce intrinsic stability of SP-B mRNA.

In our previous report, the plasmid-based expression system was used to determine the effect of these deletions on intrinsic stability of SP-B mRNA, which is accomplished by comparing the steady-state levels of SP-B mRNA containing these mutations with levels of wild-type SP-B mRNA (pCMVGFP-hspB:N) (31). Deletion of the 7.6S segment of the SP-B mRNA 3′-UTR increased intrinsic stability 2.3-fold, suggesting that the segment acts to destabilize SP-B mRNA in the absence of DEX. In the present studies, the effect of mutagenesis of 30-nt elements on intrinsic stability was determined by normalization of the expression of altered mRNA species to unaltered SP-B mRNA (expressed from pCMVGFP-hspB:N), all incubated in the absence of DEX. The results of this type of analysis are shown in Fig. 2C. Mutagenesis of any of the contiguous elements 001–005, encompassing the entire 7.6S region, resulted in a significant increase in steady-state levels of SP-B mRNA, while mutagenesis of the other elements had no effect on steady-state levels of SP-B mRNA. These data agree with our previous results. Interestingly, elements 002–004 of this region of the SP-B mRNA 3′-UTR are also necessary for DEX-mediated stabilization of SP-B mRNA, suggesting that this region has a dual role in regulation of SP-B mRNA. The results of Fig. 2C suggest that individual 30-nt elements of the 7.6S region act individually or cooperatively to mediate destabilization of SP-B mRNA by some unknown mechanism. The most dramatic effect on mRNA stability, intrinsic and DEX-mediated, was elicited by mutagenesis of element 002, which is the focus of the next series of investigations.

Element 002 is predicted to form a stem-loop structure.

Much of the 7.6S region of the human SP-B mRNA 3′-UTR is necessary for DEX-mediated stabilization of SP-B mRNA and acts to decrease intrinsic stability of SP-B mRNA in the absence of DEX, especially element 002 (Fig. 2, B and C). When the sequence of the 7.6S region was subjected to theoretical folding of the RNA using the internet-based Mfold program (http://mfold.rna.albany.edu) (40), the sequences of element 002 consistently resulted in the stem-loop structure shown in Fig. 3A. The structure has a purine-rich loop and a ΔG of −12.4 kcal/M. When the program is used to fold the entire 834-nt-long human SP-B mRNA 3′-UTR, this stem loop occurs in 66% of the final folded structures (data not shown).

Fig. 3.

Element 002 is predicted to form a stem-loop structure with a purine-rich loop and forms specific complexes with cytosolic proteins derived from A549 cells. A: structure of element 002 of human SP-B mRNA 3′-UTR as predicted by the web-based Mfold program. B: schematic of the template used for in vitro transcription of RNA. Black boxes indicate location of MS2 binding sites; gray box indicates location of element 002. T7 denotes location and direction of T7 RNA polymerase, and appropriate restriction endonuclease restriction sites are shown. C: RNA EMSA (REMSA) of in vitro transcribed RNA in B with cytosolic extracts from A549 cells. MBP indicates addition of purified protein that binds MS2 sequences and is used as control to demonstrate binding. Arrows indicate complex formation of RNA with proteins. Discontinuous presentation of the autoradiograph is shown by white space. D: REMSA of RNA containing element 002 in the presence of various competitors to demonstrate specific binding. Radiolabeled RNA containing element 002 was subjected to REMSA with cytosolic proteins in the absence or presence of 100× excess cold competitor. 7.8, Template derived from the distal end of SP-B 3′-UTR and used as a nonspecific competitor (Comp); 7.6S, RNA from SP-B 3′-UTR containing element 002. Arrow indicates position of protein-RNA complexes.

Element 002 of the human SP-B mRNA 3′-UTR is sufficient for DEX-mediated stabilization of a heterologous RFP mRNA and acts to reduce intrinsic stability.

While the results shown in Fig. 2, B and C, suggest that element 002 plays a role in the mechanism by which the 7.6S region regulates human SP-B mRNA stability, they do not imply that element 002 is sufficient for this activity. To determine if the element is sufficient for mediating DEX stabilization of SP-B mRNA and for reducing intrinsic mRNA stability, element 002 was inserted into the 3′-UTR of the RFP reporter mRNA. The resulting plasmid, pCMVGFP-RFP-MS2:002, is shown in Fig. 1A. A549 cells were transfected with the plasmid and cultured in the absence or presence of DEX, and the effects of DEX on steady-state levels of RFP and GFP mRNA were determined as described above. The presence of DEX significantly increased (∼1.4-fold) steady-state levels of RFP mRNA containing element 002 compared with cells incubated in the absence of DEX (Fig. 1B). In addition, steady-state levels of RFP mRNA in cells transfected with pCMVGFP-RFP-MS2:002 and incubated in the presence of DEX were significantly higher than levels of RFP mRNA in cells transfected with pCMVGFP-RFP-MS2 and incubated in the presence of DEX. These results demonstrate that the 30-nt-long element 002 is sufficient to impart DEX-mediated stability to a heterologous mRNA.

The results in Fig. 2C indicate that element 002 acts as a potent destabilizer of SP-B mRNA in the absence of DEX. The ability of element 002 to destabilize a heterologous mRNA was determined using pCMVGFP-RFP-MS2 and pCMVGFP-RFP-MS2:002. The results of the assay are shown in Fig. 1C. The presence of the 30-nt-long element 002 significantly decreased the stability of RFP RNA in transfected cells incubated in the absence of DEX, indicating that element 002 acts to reduce mRNA stability. Interestingly, no significant change in steady-state RFP mRNA levels was observed when the 7.6S region was assayed for the ability to regulate intrinsic stability of reporter RFP mRNA (Fig. 1C). These results suggest that while element 002 is sufficient to destabilize mRNA, the presence of other sequences of the SP-B mRNA 3′-UTR may modulate instability as well.

Cytosolic proteins specifically bind to element 002 of the human SP-B mRNA 3′-UTR.

The ability of cytosolic proteins to specifically interact with the 7.6S segment of the human SP-B mRNA 3′-UTR has been previously reported (16). Using REMSA, we then determined the ability of proteins to interact with element 002. In vitro transcribed RNA lacking or containing element 002 was generated using the T7 expression cassette template shown in Fig. 3B. Digestion of the template with EcoRI results in a 164-nt-long transcript, which contains three MS2 hairpins (4), and this RNA acts as a nonspecific control for nonspecific binding of protein, as it does not contain element 002 sequences. Digestion of the template with SacI results in a 210-nt-long transcript that contains element 002 at the end of the transcript, in addition to the MS2 sites. REMSA was performed using cytosolic extracts from A549 cells and/or an MS2 binding protein (MBP) expressed and isolated from Escherichia coli (39). MBP was used to demonstrate that RNA-protein complexes can form. The results are shown in Fig. 3C. Transcripts containing only the MS2 RNA elements form complexes with MBP, but not with proteins derived from A549 cells. However, addition of the 30-nt-long element 002 to the transcripts allows formation of additional complexes in the presence of A549 cytosolic proteins.

The specificity of the interaction of cytosolic proteins with element 002 is further evidenced in Fig. 3D. REMSA was performed as described above using labeled element 002 Sac RNA probes and proteins from A549 cells. In some samples, 100× excess unlabeled in vitro transcribed RNA was added to compete for binding. These RNAs included element 002 Sac (self) and 7.6S (containing element 002) and element 002 Eco and 7.8 (329 nt of the distal region of the SP-B 3′-UTR) (16), which serves as a nonspecific competitor in the assays. Addition of excess RNA containing element 002 efficiently competed for binding to element 002 Sac, as expected. In contrast, addition of excess RNA lacking element 002 competed some binding, but specific complexes remain.

Binding characteristics of cytosolic proteins to element 002 are not affected by the presence of DEX or time of exposure to DEX.

One mechanism by which mRNA stability can be regulated by steroid hormones is induction of protein binding to a specific sequence when cells are incubated in the presence of the hormone. Such is the case of the protein vigilin, which binds to an element in the vitellogenin RNA in the presence of estrogen (7, 8). Since element 002 is sufficient to mediate DEX-induced RNA stabilization, REMSA was used to determine if DEX alters binding of proteins to element 002. Cytosolic extracts were isolated from A549 cells grown in the absence or presence of DEX (see materials and methods). Proteins were isolated at various times after addition of DEX. REMSA was performed with in vitro transcribed radiolabeled element 002 Sac RNA in the presence of 100× excess cold element 002 Eco RNA. The results are shown in Fig. 4A. The absence or presence of DEX has no effect on specific binding of proteins to element 002: there is no increase or decrease in binding activity, nor are additional complexes formed. Upon densitometric analysis of the complexes (Fig. 4B), binding activity is relatively unchanged as a function of time of exposure to DEX. These results suggest that the mechanism by which proteins bind element 002 and increase SP-B mRNA stability in the presence of DEX does not require formation of new complexes in the presence of DEX and that the mechanism underlying regulation of SP-B mRNA involves posttranslational changes to the proteins bound to element 002.

Fig. 4.

Nature of complexes of cytosolic proteins with element 002 is unaffected by the presence of DEX. A: REMSA of radiolabeled in vitro transcribed element 002 Sac riboprobes incubated in the presence of A549 proteins. Proteins were isolated from cells incubated in the absence or presence of DEX for 0–48 h. 002 Eco, presence of 100× cold nonspecific competitor; arrow, position of specific complexes. FP, free probe. B: densitometric analysis of specific complexes. Intensity of signal of specific complexes of REMSA in A was measured and plotted as a function of time of exposure to DEX.

Proximal stem sequence of the element 002 stem-loop structure is necessary for DEX-mediated stabilization of SP-B mRNA.

The sequence of element 002 has the potential to form a strong stem-loop structure (Fig. 3A). The stem is GC-rich, and the loop is composed exclusively of purines, predominantly adenosine. The next set of investigations sought to determine if sequences or structures of the stem loop are critical in regulation of SP-B mRNA stability. Nucleotide transitions were designed as A > T, T > A, G > C, and C > G, to keep the thermodynamic nature of secondary structures as similar to wild-type as possible. Distinct elements of the structure were targeted: the loop (Loop, changed from purine-rich to pyrimidine-rich), the sequences of the proximal stem arm (Stem1), the sequences of the distal stem arm (Stem2), the sequences of both stems (Both, which regenerates the stem with different sequences), and the entire structure (Total). The sequences of the mutations are shown in Fig. 5A. The resulting expression cassettes, as well as a cassette entirely lacking element 002, are shown in Fig. 5A. A549 cells were transfected with the plasmids and cultured in the absence or presence of DEX, and the effects of DEX on steady-state levels of SP-B and GFP mRNA were determined as described above. The presence of DEX significantly increased steady-state levels of SP-B mRNA expressed from pCMVGFP-hspB:N (Fig. 5B), as expected. Also, as expected, DEX did not increase steady-state levels of SP-B mRNA when element 002 was eliminated (pCMVGFP-hspBΔ002), again demonstrating that element 002 is crucial in appropriate hormonal regulation of SP-B mRNA stability. Mutagenesis of the entire loop had no effect on the ability of DEX to significantly increase SP-B mRNA stability. On the other hand, mutagenesis of Stem1 prevented significant DEX-mediated stabilization of SP-B mRNA, while mutagenesis of Stem2 did not, indicating that the sequences of Stem1 are necessary for hormonal regulation of SP-B mRNA stability. This result is reinforced when mutagenesis of both stem sequences is tested; in all cases where Stem1 is altered, DEX-mediated stabilization of mRNA stability is absent, despite the fact that the stem structure can be reformed (albeit with different sequences). These results suggest that the proximal sequences (5′-AGCUGCCAGGCU-3′) are necessary for DEX-mediated stabilization of SP-B mRNA.

Sequence and formation of the stem structure are necessary for the ability of element 002 to reduce intrinsic stability of SP-B mRNA.

Since element 002 can also function to reduce intrinsic stability, the mutants of element 002 shown in Fig. 5A were assessed for their effect on this activity. A549 cells were transfected with the plasmids and cultured (in the absence of DEX), and the effects of the mutations on steady-state levels of SP-B mRNA were determined as described above. Mutagenesis of the loop had no effect on intrinsic SP-B mRNA stability (Fig. 5C), clearly demonstrating that the loop has absolutely no role in the mechanisms that regulate SP-B mRNA stability. However, mutagenesis of either or both stem sequences abolished the ability of element 002 to reduce intrinsic stability of SP-B mRNA. As expected, elimination of the entire structure of element 002 increased intrinsic stability of SP-B mRNA. These results clearly indicate that intrinsic stability of SP-B mRNA mediated by element 002 requires formation of the stem structure and the proper sequences of the structure.

Cytosolic proteins specifically bind to proximal stem sequences of the element 002 stem-loop structure.

Figure 3C clearly shows that proteins specifically bind intact element 002, and Fig. 5B suggests that regulation of SP-B mRNA stability by DEX requires the presence of the proximal stem sequences (Stem1). To determine if distinct components of the element 002 stem-loop structure are required for protein binding, REMSA was performed (see Experimental procedures) on in vitro transcribed RNA containing element 002 with the mutations shown in Fig. 6. To demonstrate specific binding, the samples were incubated in 100× unlabeled excess element 002 Eco RNA. The results of the analysis are shown in Fig. 6D. A nonspecific interaction can be visualized in all samples but is very strong when element 002 and Stem2 probes are used. However, the presence of 100× excess cold competitor (element 002 Eco) demonstrates that Stem1, Both, and Total do not specifically bind cytosolic proteins, as the interactions are greatly diminished or absent. While interactions of element 002 and Stem2 probes are diminished in the presence of excess competitor, specific interactions are clearly visible. These results of REMSA indicate that protein interactions with the element 002 structure occur through the proximal stem sequence, and since DEX-mediated stabilization of SP-B mRNA occurs only when this sequence is present, these findings suggest a mechanism by which mRNA stabilization by DEX occurs through a protein(s) that binds the proximal stem sequence 5′-AGCUGCCAGGCU-3′.

Fig. 6.

Cytoplasmic proteins specifically and efficiently bind the element 002 structure only when the sequence of the proximal stem is present. REMSA shows radiolabeled in vitro transcribed riboprobes of element 002 containing mutations described in Fig. 5A and incubated in the presence of A549 proteins. −, Addition of lysate to the riboprobes; +, incubation in the presence of 100× cold nonspecific competitor (Eco). Arrow indicates position of specific complexes. Discontinuous presentation of the autoradiograph is indicated by white space.

DISCUSSION

The transition from a fluid-filled intrauterine environment to an air-breathing extrauterine environment requires expression of SP-B for adequate pulmonary function (14). Expression of SP-B is developmentally regulated, and premature infants born before the onset of adequate expression of pulmonary surfactant and SP-B are susceptible to multiple lung injuries, such as RDS, which are associated with increased morbidity and decreased life expectancy (1). Administration of glucocorticoids accelerates fetal lung development and surfactant synthesis (21). Corticosteroids are used to prevent the onset and severity of RDS (21, 38). However, controversy has arisen concerning the efficacy of antenatal glucocorticoids, as well as detrimental effects of glucocorticoids on neuronal development (6, 23). Thus, understanding the molecular mechanisms by which glucocorticoids increase expression of the components of pulmonary surfactant, such as SP-B, may provide alternative avenues for treatment of individuals with inadequate surfactant without the use of corticosteroids.

The effect of glucocorticoids to increase expression of human SP-B is complex; the presence of DEX increases transcription of SP-B and SP-B mRNA stability (2, 32, 34). Our laboratory has focused on defining the molecular mechanisms by which glucocorticoids increase human SP-B mRNA stability. We previously reported that the SP-B mRNA 3′-UTR is necessary for stabilization of SP-B mRNA by DEX, indicating that this region contains RNA sequences necessary for regulation of SP-B mRNA stability (16, 31). The focus of this investigation was to identify specific elements of human SP-B mRNA involved in regulation of SP-B mRNA stability. Here we show that the presence of the entire SP-B 3′-UTR is sufficient to impart DEX-mediated stability to a heterologous mRNA. Previously, a 126-nt-long segment of the 3′-UTR, called the 7.6S region, was found to be necessary for stabilization of SP-B mRNA by DEX (16, 31). In the present study, this region was found to be sufficient to impart hormonal regulation of stability to a heterologous mRNA.

Sequential substitution mutagenesis of this region of the 3′-UTR, in which 30-nt-long segments were replaced with sequences of the pUC18 multiple cloning site, was used to delineate specific elements involved in regulation of SP-B mRNA stability. Consecutive elements (002–004 and 007–008) of the 7.6S region were required for DEX-induced stabilization of SP-B mRNA. Interestingly, mutation of elements 003 and 004 also resulted in an increase in mRNA stability in the absence of DEX, suggesting that they also act to reduce intrinsic SP-B mRNA stability. Mutagenesis of one element, 002, resulted in a large (∼9-fold) increase in stability of SP-B mRNA in the absence of DEX. This surprising characteristic became the focus of subsequent investigations. When the 30-nt-long sequence of element 002 was subjected to theoretical folding based on the web-based Mfold program, which is used to predict the two-dimensional structure of RNA, a stem-loop structure was evident when the sequence was analyzed alone or in the context of the 7.6S region or the entire SP-B 3′-UTR. Similar to the 7.6S segment, element 002 has the ability to increase stability of a heterologous mRNA in the presence of DEX. This element also has the ability to reduce intrinsic stability of a heterologous mRNA in the absence of DEX. These results indicate that a single structure in the SP-B mRNA 3′-UTR can act in two different modes of regulation of mRNA stability, intrinsic and hormonal.

Site-directed mutagenesis of specific sequences of element 002 that change the nature of the stem-loop structure was used to investigate the possibility that distinct elements of the structure are responsible for the two modes of regulation. Mutations were introduced into element 002, where G and C were switched and A and T were switched. While these mutations altered the sequence of the specific elements of the structure, any potential hybridization should be thermodynamically equal to the original sequence. These mutations altered the sequence of the proximal arm of the stem, the loop, the distal arm of the stem, both arms (while retaining hybridization), and complete element 002 sequences while maintaining the structure. The results indicate that the sequence of the loop has no role in regulation of SP-B mRNA stability. DEX-mediated stabilization of SP-B mRNA required the sequence of the proximal stem (5′-AGCUGCCAGGCU-3′), even when compensatory mutation regenerates formation of the stem. It appears that any mutation of the stem sequence abolishes the activity of the element 002 structure to reduce intrinsic stability, regardless of the formation of a stem structure. These results suggest that a single structure in the SP-B mRNA 3′-UTR mediates two modes of regulation of stability, DEX-mediated stabilization and intrinsic stability, through different components of the structure.

We previously showed through REMSA that the 7.6S region forms specific complexes with cytosolic proteins from alveolar epithelial cells (16, 31). When REMSA was performed on element 002, proteins isolated from human alveolar epithelial cells formed specific complexes with the hairpin structure. These results portend the possibility that binding of a protein or protein complex depends on the absence or presence of DEX. Estrogen-mediated stabilization of vitellogenin mRNA in Xenopus liver is the best example of such a mechanism. Activation of the estrogen receptor by estrogen induces a protein that binds to a particular segment of the 3′-UTR of the vitellogenin mRNA containing the sequence ACUGUA, increasing its stability 30-fold (8). The binding of this protein, vigilin, prevents site-specific endonucleolytic cleavage of the mRNA. In these studies, REMSA demonstrated that cytosolic proteins bind specifically to element 002 and that the mobility and intensity of these complexes are not altered when proteins are isolated from cells incubated in the absence or presence of DEX. These results indicate that the presence of DEX promotes neither association nor dissociation of proteins that bind to element 002, suggesting that regulation of SP-B mRNA stability by DEX may be mediated by mechanisms involving posttranslational modifications of the proteins involved.

A schematic model of the elements of the 7.6S region and portions of element 002 responsible for intrinsic or hormonally regulated SP-B mRNA stability is shown in Fig. 7. In the SP-B mRNA 3′-UTR, a 90-nt-long region, including the element 002 stem loop, is necessary for DEX-mediated stabilization and intrinsic stability (Fig. 7A). These studies also demonstrate that DEX-mediated stabilization of SP-B mRNA requires the sequence of the proximal “leg” of the stem (5′-AGCUGCCAGGCU-3′) and that the ability to reduce intrinsic mRNA stability requires formation of the stem with its native sequence (Fig. 7B). While both modes of regulation appear to be distinct in mechanism and independent, the possibility that they require common factors, such a protein, remains. The fact that elements 002, 003, and 004 are involved in both modes of regulation suggests that regulation of intrinsic and hormone-mediated stability of SP-B mRNA has the potential to share common mechanisms. On the basis of these and previous findings, one potential mechanism of regulation of human SP-B mRNA stability is as follows: interactions of proteins binding to the sequences of the 7.6S region and element 002 destabilize SP-B mRNA to maintain intrinsic stability. In the presence of DEX, the proteins that bind the sequence undergo a posttranslational change that dampens this destabilizing effect, thereby increasing the stability of SP-B mRNA.

Fig. 7.

Model of SP-B mRNA regulation mediated through the 7.6S segment and element 002. A: based on results of the study, elements 001–004 of the 7.6S region of the human SP-B mRNA 3′-UTR are involved in regulation of intrinsic stability of SP-B mRNA. Elements 002–004 are involved in regulation of DEX-mediated stabilization of SP-B mRNA. B: in element 002, the proximal stem sequence appears to be important in DEX-mediated stabilization, while sequences and structure of the stem are required for maintenance of intrinsic stability.

The goal of our research is to define the complex molecular mechanisms by which SP-B mRNA stability is regulated by hormones through identification of specific mRNA sequences and proteins involved in regulation of SP-B mRNA stability by glucocorticoids. Here we have identified specific sequences of the 3′-UTR involved in regulation of SP-B mRNA stability, and our results suggest that a posttranscriptional mechanism is responsible for DEX-induced stabilization of SP-B mRNA. This information has the potential to allow identification of the proteins that bind these sequences. Identification of the RNA sequences and binding proteins involved in regulation of SP-B mRNA stability will provide a more complete description of the mechanisms by which glucocorticoids increase mRNA stability, as well as a new model of hormonal regulation of mRNA stability.

GRANTS

This research was supported by National Heart, Lung, and Blood Institute Grant R01-HL-068116 to J. L. Alcorn.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

H.W.H., S.R.B., and J.L.A. are responsible for conception and design of the research; H.W.H., D.E.P., W.B., S.P., and S.R.B. performed the experiments; H.W.H., D.E.P., S.R.B., and J.L.A. analyzed the data; H.W.H., D.E.P., S.R.B., and J.L.A. interpreted the results of the experiments; H.W.H., S.R.B., and J.L.A. edited and revised the manuscript; H.W.H., D.E.P., W.B., S.P., S.R.B., and J.L.A. approved the final version of the manuscript; J.L.A. prepared the figures; J.L.A. drafted the manuscript.