An open reading frame element mediates posttranscriptional regulation of tropoelastin and responsiveness to transforming growth factor beta1

Abstract

Elastin, an extracellular component of arteries, lung, and skin, is produced during fetal and neonatal growth. We reported previously that the cessation of elastin production is controlled by a posttranscriptional mechanism. Although tropoelastin pre-mRNA is transcribed at the same rate in neonates and adults, marked instability of the fully processed transcript bars protein production in mature tissue. Using RNase protection, we identified a 10-nucleotide sequence in tropoelastin mRNA near the 5' end of the sequences coded by exon 30 that interacts specifically with a developmentally regulated cytosolic 50-kDa protein. Binding activity increased as tropoelastin expression dropped, being low in neonatal fibroblasts and high in adult cells, and treatment with transforming growth factor beta1 (TGF-beta1), which stimulates tropoelastin expression by stabilizing its mRNA, reduced mRNA-binding activity. No other region of tropoelastin mRNA interacted with cellular proteins, and no binding activity was detected in nuclear extracts. The ability of the exon-30 element to control mRNA decay and responsiveness to TGF-beta1 was assessed by three distinct functional assays: (i) insertion of exon 30 into a heterologous gene conferred increased reporter activity after exposure to TGF-beta1; (ii) addition of excess exon 30 RNA slowed tropoelastin mRNA decay in an in vitro polysome degradation assay; and (iii) a mutant tropoelastin cDNA lacking exon 30, compared to wild-type cDNA, produced a stable transcript whose levels were not affected by TGF-beta1. These findings demonstrate that posttranscriptional regulation of elastin production in mature tissue is conferred by a specific element within the open reading frame of tropoelastin mRNA.

FIG. 1

Expression of tropoelastin pre-mRNA persists in adult tissue. (A) Total RNA was isolated from lungs of 19-day fetuses, 3- and 11-day-old neonates, and 6-month-old adult rats and analyzed by Northern hybridization for tropoelastin and GAPDH mRNAs. Ethidium bromide staining of the gel before transfer (smaller panel) confirmed equivalent loading among lanes. (B) The same RNA samples were amplified by RT-PCR with intron-35 primers. Products were detected by Southern hybridization with ³²P-labeled intron-specific oligomeric probe. No signal was detected in samples processed without reverse transcriptase (data not shown). (C) Autoradiographic signal for tropoelastin mRNA and pre-mRNA was quantified and normalized to the signal for the 11-day-old neonate sample, which was arbitrarily set at 25. The autoradiographic signal density of tropoelastin mRNA is expressed relative to that of GAPDH mRNA (data not shown).

FIG. 2

Turnover of tropoelastin pre-mRNA and mRNA. (A) Second-passage neonatal (NN) and adult (Ad) rat lung fibroblasts were grown to confluence and were treated with 50 μM DRB for 0 to 60 min or 24 h. At the indicated times, total RNA was isolated, and levels of tropoelastin pre-mRNA were determined by RT-PCR. As seen in intact tissue (Fig. 1), equivalent levels of pre-mRNA were detected in control (0- and 24-h-DRB) neonatal and adult fibroblasts. In the presence of DRB, tropoelastin pre-mRNA levels decayed at about the same rate (t_1/2 = ∼15 min) in both neonatal and adult fibroblasts. (B) Neonatal and adult rat lung fibroblasts were treated with DRB for 1 or 24 h, and total RNA was isolated and analyzed by Northern hybridization. Tropoelastin mRNA in neonatal fibroblasts did not decay over the 24-h treatment. In contrast, tropoelastin mRNA in adult fibroblasts decayed rapidly to undetectable levels by 1 h post-DRB. In another experiment, fibroblasts were treated with 10 μg of actinomycin D (Act-D) per ml for 0.5 or 1 h. As with DRB, tropoelastin mRNA was stable in neonatal fibroblasts but decayed rapidly in adult cells.

FIG. 3

Sequences in exon 30 bind a cytosolic factor present in adult lung fibroblasts. (A) Map of tropoelastin mRNA and summary of the binding data. Tropoelastin mRNA is transcribed from 36 exons, which alternate between regions coding for hydrophobic or cross-linking domains. Exon 36 codes for a conserved hydrophilic domain and the fairly large 3′ UTR. The horizontal lines under the mRNA map indicate the RNA probes used protection assays; the numbers designate which exons the probes represent. For exon 36, two probes were made: 3′ UTR/L include all of the exon-36 sequences up to the first polyadenylation signal; 3′ UTR/S is a truncated version of 3′ UTR/L. +, Protected band was detected; ×, no detectable binding. (B) ³²P-labeled RNA probes were transcribed in vitro and incubated with fixed amounts (10 or 30 μg of total protein) of nuclear or cytosolic extracts from adult lung fibroblasts. After a 30-min incubation, unbound RNA was digested by T1 RNase, and protected products were resolved by electrophoresis and visualized by autoradiography. No protected fragment or residual probe was seen in reactions containing T1 RNase without cytosolic extract (Probe+T1). A protected band (arrowhead) was produced with an RNA probe covering exons 17 to 36 incubated with cytosolic extract (Cyto) but not when incubated with nuclear extracts (Nucl). In contrast, no protected band was detected with an RNA probe covering exons 1 to 18. (C) A protected band was detected with exon-30 RNA and cytosolic extract. Because gels were not all the same dimension or run for the same time, the migration of the protected band differs among experiments. (D) No protected bands were detected with antisense RNA probes. (E) Yield of the protected band produced with RNA probe 17-36 was inhibited with a 20- or 60-fold excess of unlabeled exon 30 RNA (Ex30). Unlabeled RNA transcribed from pGEM plasmid sequences (pG) did not inhibited production of the protected band. (F) Sequence of rat tropoelastin exon 30. The bases in lowercase letters represent a 72-bp insert found only in the rodent gene. Progressively smaller ³²P-labeled RNA probes were transcribed from insert linearized with the indicated restriction enzymes and were incubated with adult fibroblast cytosolic extract. All probes produced the same protected fragment, and an example is shown in the next panel. (G) Incubating ³²P-labeled RNA probe transcribed from BsrSI-linearized exon-30 cDNA with 30 μg of cytosolic extract from adult lung fibroblasts produced a protected fragment. The size of the protected fragment was identical to that produced with a full-length exon-30 RNA probe. Binding was specifically competed with excess cold exon-30 RNA but not with RNA transcribed from pGEM plasmid sequences. (H) RNA probe 27-34 was incubated with (+) or without (−) adult fibroblast cytosolic extract before addition of T1 RNase. The protected products were resolved by electrophoresis, excised from the gel, and extracted in phenol-chloroform. The resulting ³²P-labeled RNA fragment was resolved on a sequencing gel, and its size was determined by comparison to the migration of single-stranded DNA markers. One prominent band migrating at ca. 9 to 11 nt was detected. Other bands common to both lanes likely represent undigested RNA.

FIG. 4

Binding of the tropoelastin mRNA-binding protein relies more on RNA sequence than structure. (A) RNA oligomers 1 to 7 of various lengths were synthesized to cover overlapping regions of the 5′ 50-nt of rat tropoelastin exon 30. The oligomer lengths are as follows: 1, 21 nt; 2, 23 nt; 3, 40 nt; 4, 18 nt; 5, 13 nt; 6, 12 nt; and 7, 20 nt. A protected product was produced only with oligomer 4 (+), which is identical to the AluI RNA probe used in the experiments summarized in Fig. 3F. (B) Sequences of oligomer 4 and three mutant RNA oligomers (M1, M2, and M3), with the mutated bases underlined. (C) ³²P-labeled RNA oligomers were incubated with cytosolic extract from control adult lung fibroblasts (−) or cells treated with 50 pM TGF-β1 (+) for 48 h. A specific protected band was detected only with wild-type oligomer 4, and binding activity was decreased by treatment with TGF-β1.

FIG. 5

Insertion of exon 30 increases expression of a heterologous gene in response to TGF-β1. Rat tropoelastin exon-30 cDNA was inserted into a SmaI site in either a sense or an antisense orientation into a luciferase gene just 3′ of the translation stop codon (TAA). Adult lung fibroblasts were transiently transfected with these constructs or the parental plasmid and a CMV–β-Gal (β-gal) construct and, 24 h later, half of the dishes were treated with 50 pM TGF-β1. After an additional 24 h, cells were harvested, and lysates were assayed for reporter gene activity. The data shown are the mean ± the standard deviation of triplicate dishes for each condition from three separate experiments.

FIG. 6

Competition of tropoelastin mRNA degradation by exon 30. Polysomes and S100 extracts were isolated from two lines of neonatal lung fibroblasts (NLF-1 and -2) and two lines of adult lung fibroblasts (ALF-1 and -2) and were combined and incubated with (solid symbols) or without (open symbols) an excess of in vitro-transcribed exon-30 RNA. Samples were harvested, and mRNA levels for tropoelastin and GAPDH were assessed by RT-PCR. For tropoelastin mRNA, sequences coded by exons 35 and 36 were amplified. Tropoelastin mRNA in NLFs degraded with a half-life of ca. 6 h, and this rate was only minimally increased in the presence of exon-30 RNA. In contrast, tropoelastin mRNA degraded rapidly in ALFs with a half-life of less than 1 h and was essentially undetectable by 2 h. In the presence of excess exon-30 RNA, degradation of tropoelastin mRNA in ALF slowed but was nearly completely degraded by 7 h. The bottom panel shows the 2-, 5-, and 7-h ALF datum points graphed on an expanded y axis.

FIG. 7

Exon 30 confers regulated mRNA turnover and responsiveness to TGF-β1. (A) Human PE cells were transfected a full-length (exons 1 to 36) or a truncated (exons 1 to 29) bovine tropoelastin cDNA under the control of a CMV promoter. Stable cell lines were selected and pooled. For the experiment shown, some cells of each group were treated with 50 pM TGF-β1 and, 24 h later, some control and treated cells were harvested (0 h). The remaining cells were treated with 10 μg of actinomycin D per ml with or without TGF-β1 and were harvested 16 and 24 h later. Tropoelastin and GAPDH mRNAs were assayed by RT-PCR and Southern hybridization. (B) PE cells were transfected with full-length (1-36) bovine tropoelastin cDNA or with a deletion mutant lacking exon 30 (Δ30), and stable cell lines were selected and pooled. Cells of each group were treated with 50 pM TGF-β1 and, 24 h later, RNA was isolated. Tropoelastin mRNA was assessed by Northern hybridization. Loading equivalence among lanes was demonstrated by ethidium bromide staining (not shown).

FIG. 8

Characterization of the tropoelastin mRNA-binding protein. (A) Cytoplasmic extracts from fetal (F), neonatal (N), and adult (A) rat lung fibroblasts were incubated with RNA probe 17-36. Before T1 RNase digestion, half of the samples were incubated with proteinase K (+Prot K), which eliminated all binding activity. In addition, binding activity was developmentally regulated, being greatest in extracts from adult fibroblasts. (B) Cytoplasmic extracts from fetal, neonatal, and adult rat lung fibroblasts were incubated with RNA probe 17-36 and treated with T1 RNase. The samples were then exposed to UV light, and cross-linked products were resolved by denaturing SDS-PAGE. A band migrating at ca. 53 kDa (arrow) was seen in all lanes, and its abundance increased with age, being most prominent in adult fibroblasts extracts. The large, faster-migrating band seen in all lanes likely represents non-cross-linked probe and nonspecific cross-linked products. (C) Fibroblasts were isolated from neonatal and adult rat skin and lung and were treated with 50 pM TGF-β1 for 48 h before the isolation of cytosolic extracts. Extracts were incubated with RNA probe 17-36. Binding activity was greater in adult fibroblast extracts and was reduced after exposure to TGF-β1, but only in adult extracts. (D) Adult rat lung fibroblasts were treated with 0, 0.05, 0.5, 2, or 5 pM TGF-β1. After a 48-h incubation, cells were harvested and used for isolation of total RNA and cytosolic extract. Levels of tropoelastin (TE) and GAPDH mRNAs were assessed by Northern hybridization (upper panels), and tropoelastin mRNA-binding protein (TE mRNA BP) activity was assessed by incubation with labeled exon-30 RNA (lower panels).