| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Biochemistry and Molecular Biology, University of Medicine and Dentistry of New Jersey-New Jersey Medical School, Newark, New Jersey 07101
Address all correspondence and requests for reprints to: Melissa B. Rogers, Biochemistry and Molecular Biology (MSB E627), University of Medicine and Dentistry of New Jersey-New Jersey Medical School, 185 South Orange Avenue, P.O. Box 1709, Newark, New Jersey 07101-1709. E-mail: rogersmb{at}umdnj.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
BMP2 and Osteoporosis |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
Numerous Mechanisms Modulate BMP Signaling |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
BMP2 Single Nucleotide Polymorphisms (SNPs) Disrupt an Unusually Conserved Regulatory Region |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
We showed previously that the effect of the ultra-conserved region on BMP2 reporter gene activity in RA-treated F9 cells is significantly greater than any upstream or 5'-UTR region (35, 36, 37). Unlike the upstream elements, the 3'-UTR stimulatory effect is specific to RA-induced expression of BMP2 (37). We have now tested subclones containing the most highly conserved sequence from the 3'-UTRs of mouse, human, and dog. The homologous region from all three species strongly activates mouse promoter-driven reporter genes in the mouse model (Fig. 1
). Interchangeable 3'-UTR function between these mammals suggests that regulatory mechanisms are conserved. Therefore, mechanisms identified using mouse systems are highly relevant to how the human BMP2 gene is regulated.
|
). Thus, as observed previously with a promoter-only reporter transgene (62), reporter genes with the BMP2 3'-UTR are expressed in bone.
Activating Elements Act Posttranscriptionally
We made two constructs to distinguish between reporter gene activation due to cis-regulatory elements within the mRNA or transcriptional enhancers in the plasmid DNA. BmpLucBmp (Fig. 1C) contains the entire 3'-UTR and downstream sequences. The transcribed reporter mRNAs may end at the natural BMP2 poly(A) signals (37) or the simian virus 40 (SV40) poly(A) signal from pGL2-Basic. To distinguish between elements that act posttranscriptionally within the RNA molecule or as enhancers of transcription within the DNA, we placed the 2.2-kb fragment downstream of the SV40 poly(A) signal (BmpLucSVBmp; Fig. 1C). Because the SV40 sequence is a strong polyadenylation and cleavage signal, most mRNAs should end before the BMP2 sequence. The luciferase activity of BmpLucSVBmp-transfected cells was less than half that of BmpLucBmp. More importantly, relative RNA levels correlated with luciferase activities.
We also found that luciferase levels in cells transfected with a promoter-only construct, BmpLuc, declined faster after treatment of cells with the transcriptional inhibitor 5,6-dichloro-1-ß-ribofuranosyl benzimidazole (DRB) than in cells transfected with the 3'-UTR-containing BmpLucBmp (Fig. 1D). This may reflect a relatively less stable mRNA in the absence of the BMP2 3'-UTR. We previously showed that RNAs with the ultra-conserved region decayed more rapidly in vitro in extracts from F9 cells that do not express BMP2 relative to cells that express BMP2 (35). These data are consistent with two hypotheses: 1) the BMP2 3'-UTR contains elements that stabilize an mRNA in cells that do express BMP2 and/or 2) the BMP2 3'-UTR contains elements that destabilize an mRNA in cells that do not express BMP2.
Trans-Acting Factors Interacting with the Osteoporosis-Associated Polymorphic Region Influence BMP2 RNA Decay in Vitro
The nearly 400-nt ultra-conserved region probably contains multiple posttranscriptional regulatory motifs. One such well-characterized motif is the AU-rich element (ARE) which marks mRNAs whose half-life is regulated (63, 64, 65, 66). The ultra-conserved region in mammalian BMP2 genes contains 10 widely spaced AREs (AUUUA or AUUUUA) (35, 37). Interestingly, a conserved AUUUA element in the BMP2 RNA is altered to AUUUC in an osteoporosis-associated SNP (rs15705) (22).
It is not known whether the 10 AREs function equivalently in any particular cell type. Because each ARE is embedded in highly distinct flanking sequence, binding affinity for proteins that regulate mRNA decay may vary. Additionally, the AREs may function additively, antagonistically, or synergistically. However, for all possibilities, the ARE/protein interactions may contribute to the sophisticated level of control required to fine-tune BMP2 expression in animals. ARE-interacting factors may promote or inhibit mRNA decay. A simple hypothesis is that more AREs recruit more decay proteins leading to more rapid decay. To begin to test this, we used a quantitative in vitro assay to measure the regulated decay of a synthetic, capped RNA in extracts (35, 67, 68, 69). Because these RNAs were not polyadenylated, this assay specifically measures the effect of ARE motifs on degradation rate, but not on deadenylation. Furthermore, we showed previously that RNAs with the well-characterized ARE from the short-lived TNF
mRNA, but not RNAs with a mutated TNF (Fig. 2B) (35, 70), decay rapidly. This indicates that F9 cells contain the factors required to regulate the stability of ARE-containing RNAs.
|
, A and B). This is shorter than the 15-min half-life of RNAs with the potent ARE from the short-lived TNF mRNA, which has six densely packed AUUUA motifs (Fig. 2B) (35, 70). Similar to that of TNF, the half-lives of three different BMP2 RNAs with five or six AREs are greater than 10 min (Fig. 2, A and B). A 35-nt RNA with three isolated AREs was relatively stable. This suggests that the number of AREs may influence decay rate in these RNAs.
Tissue-Specific ARE Binding Proteins Interact with the BMP2 RNA
Proteins that bind the 3'-UTR regulate many important regulatory processes; e.g. polyadenylation efficiency, nuclear mRNA export, cytoplasmic localization, and mRNA translation efficiency and decay. As a first step toward understanding the role of the ultra-conserved region, we used UV cross-linking to characterize the cytoplasmic proteins that interact with this RNA (Fig. 3, lanes 1, 4, 7, and 10). 32P-labeled RNAs containing the mouse BMP2 ultra-conserved region or the homologous human region UV cross-linked discrete proteins in cytoplasmic extracts from mouse cells (F9) or human cells (BEAS-2B immortalized human bronchial epithelial, A549 lung adenocarcinoma, or HeLa). Whereas the BMP2 RNA UV cross-links some proteins from all cells that migrate in similar positions (for example, p63), others are cell-type specific; for example, several proteins migrating in the 90- to 130-kDa range (Fig. 3A). We have also performed UV cross-linking assays using the mouse RNA with the three extracts from human cells and the human RNA with extracts from F9 cells and detected no differences in protein profiles (data not shown). Therefore, as might be expected from the sequence identity, the differences between the mouse and human cell protein profiles are cell-type specific rather than species specific.
|
mRNA are well characterized and can be used to make initial comparisons and tentative identifications. So, to begin to identify the proteins that bind the ultra-conserved region, we first examined proteins UV cross-linked to the TNF ARE (Fig. 3A, lanes 2, 5, 8, and 11). Although some proteins are unique to BMP2 in each cell type, several proteins of similar apparent molecular weights were UV cross-linked by both BMP2 and TNF. The apparent molecular weight of some of the proteins interacting with both RNAs suggests that several are previously identified proteins. For example, the set of 32- to 37-kDa proteins fall within the molecular weight ranges of known stability regulators including HuR (32 kDa) (71), the four AUF-1/hnRNP D isoforms (37–45 kDa) (72, 73, 74, 75), and tristetraproline (TTP) (35–55 kDa) (76, 77, 78). To test whether these proteins were indeed ARE-binding proteins, we attempted to UV cross-link them using an RNA containing a mutated TNF ARE (Fig. 3A, lanes 3, 6, 9, and 12) (79). The inability of this RNA to label the low-molecular-weight set of proteins suggests that they are bona fide ARE-binding proteins.
We also observed that the intensity of UV cross-linking between the TNF and the BMP2 RNAs and proteins that appear to be the same size varies significantly. For example, in HeLa cells (lane 11), the 32-kDa protein that is strongly labeled by the TNF RNA with its multiple overlapping AUUUA motifs (see Fig. 2) is HuR (71). Although HuR is abundant in HeLa cells, the BMP2 RNA with its widely separated AUUUA or AUUUUA motifs (35, 37) cross-linked HuR poorly in the same extracts.
RNA oligonucleotide (oligo) competition can interfere with protein:RNA interactions at defined motifs such as AREs (80). The only sequence in common to the BMP2 and TNF RNAs is AUUUA. If the proteins UV cross-linked by both the BMP2 and the TNF RNAs that migrate similarly are indeed the same proteins, then an RNA oligo with the TNF ARE should compete for binding and decrease cross-linking. Figure 3, B and C, shows that the TNF ARE oligo (lanes 3–5, 10–14), but not a nonspecific oligo (lanes 6–8, 15–19), inhibited cross-linking of the 32- to 33-kDa proteins that are labeled by both the BMP2 and the TNF RNAs in HeLa cells. Thus, a common subset of proteins interacts with the classical ARE found in TNF and the BMP2 ultra-conserved region.
Different Proteins Bind the Normal RNA Sequence Relative to RNA with the rs15705 SNP in an ARE
Unlike the concatenated AREs in the TNF RNA, all but two of the BMP2 AREs in the ultra-conserved sequence are isolated. The conserved sequence flanking each ARE may alter the presentation of each ARE to stabilizing or destabilizing RNA binding proteins. We have begun to analyze the influence of each conserved ARE in the BMP2 RNA, beginning with an ARE that is altered in individuals with the rs15705 SNP. To test whether this SNP could alter the affinity of specific proteins for the BMP2 mRNA, we synthesized 35-nt RNA oligos that span three isolated AREs in a region that is identical between rodents and primates (37). The central ARE of this set is changed by the rs15705 SNP from AUUUA to AUUUC. These oligos were used to compete for UV cross-linking to 32P-labeled RNAs containing the full-length ultra-conserved sequences from mouse or human. The polymorphism could decrease the affinity of the RNA oligo for specific proteins and cause a loss of function. Alternatively, the polymorphism could increase the affinity of the RNA oligo for specific proteins and cause a gain of function.
First, the normal and the SNP oligos both inhibited the UV cross-linking of specific proteins. For example, 1 pmol of either oligo competed efficiently for the interaction between p63 present in five different cell types and the BMP2 RNAs (Fig. 4, A and B, and data not shown). UV cross-linking of other proteins (for example, several migrating above the 37-kDa marker) were never significantly affected by either oligo (n = 2–3 for each cell type). This suggests that the SNP has no effect on the affinity of the oligo for these proteins. However, in contrast to the wild-type oligo, the mutated oligo less efficiently inhibited UV cross-linking to proteins migrating between 32 and 37 kDa in F9, HeLa, A549 lung adenocarcinoma, BEAS-2B immortalized human bronchial epithelial, and ROS17/2.8 rodent osteosarcoma cells (compare lanes 1–5 with lanes 6–10 in each set; Fig. 4, A–C).
|
) using the TNF ARE oligo (lanes 4–6), the normal (lanes 7–10) or SNP (lanes 11–14) BMP2 oligos or the nonspecific oligo (lanes 15–17) confirmed that the proteins migrating at this position are ARE binding proteins. As might be expected from the concatenated nature of the AUUUA motifs in the TNF ARE, the TNF ARE competed more effectively than the BMP2 oligo at lower concentrations (Fig. 3D).
We also noted that cross-linking to a protein migrating at approximately 100 kDa (p100) was competed more effectively by the SNP oligo in HeLa cells (Fig. 5A, lanes 3–11) and A549 cells (data not shown). This suggests that this protein may bind the SNP RNA with higher affinity. Because competition is an indirect, albeit quantitative, method of measuring RNA:protein affinity, we transcribed two capped and 32P-labeled RNAs containing the 35 nt of BMP2 sequence identical with the normal and rs15705 SNP-containing competitor RNA oligos. As expected, these two small transcribed RNAs UV cross-linked a subset of HeLa proteins relative to RNAs containing the entire ultra-conserved region (364 nt). Consistent with the increased ability of the SNP RNA oligo to compete for UV cross-linking by the full-length ultra-conserved region, the SNP RNA cross-linked the 100-kDa protein more than twice as well as the normal RNA (2.4 ± 0.2-fold; n = 6; Fig. 5A, lanes 1 and 2). Increased HeLa cell protein binding to the SNP RNA was also observed under the conditions of EMSAs (Fig. 5B, lanes 1 and 2). The BMP2 RNAs do not cross-link F9 cell proteins migrating at 100 kDa; however, EMSAs revealed that the SNP RNA bound proteins in both undifferentiated and differentiated cells more efficiently than the normal RNA (Fig. 5B, lanes 3–6). Taken together, the competitions, direct UV cross-linking, and EMSAs suggest that the SNP mRNA binds several proteins differently than normal mRNAs.
|
) (35), approximately 20% of this RNA remains after 30-min incubation. In contrast, nearly 40% remained when 5 or more pmol of competitor RNA oligo was added (Fig. 6, A and B). Similarly, this RNA oligo stabilized the human BMP2 RNA in HeLa cells (Fig. 6B). Thus, interfering with the proteins that bind this 35-nt region stabilized the 364-nt full-length BMP2 RNA in extracts from two cell types. Finally, we have directly compared the decay rate of the in vitro capped and transcribed small RNAs containing the BMP2 sequence of the normal or rs15705 SNP-containing RNA mimicking that of the competitor RNA oligos. The 35-nt SNP RNA decayed considerably more rapidly than the normal RNA in both undifferentiated F9 cells and HeLa cells (Fig. 6C).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The most significant aspect of our observations is that the single nucleotide change found in the rs15705 SNP markedly changes the affinity of several proteins for the BMP2 ultra-conserved region (Figs. 4 and 5). Specifically, an RNA with the polymorphism loses affinity for a 32- to 37-kDa set of proteins in five different extracts from human or mouse cells and gains affinity for a 100-kDa protein in human cells. More importantly, the stability of RNAs with the normal or the SNP sequence differs in both mouse and human cells (Fig. 6C). Thus, multiple and distinct assays showed that the normal and SNP RNAs function differently in vitro. We do not know how these proteins interact with the intact BMP2 mRNA in a living cell. However, it is reasonable to suppose that changes in BMP2 mRNA:protein interactions with the SNP mRNA would alter the posttranscriptional mechanisms controlling BMP2 expression in vivo.
We have not yet identified the 100-kDa protein that binds the SNP RNA better. However, based on the ability of the mutated TNF RNA to UV cross-link it (Fig. 3A) and the inability of the ARE RNA oligo to inhibit this cross-linking (Fig. 3B), it does not appear to be a classical ARE-binding protein. In contrast, the low-molecular-weight set of proteins with higher affinity for the normal RNA relative to the SNP RNA may be known RNA binding proteins such as HuR and various isoforms of hnRNPD/AUF1 and TTP. Previous studies (71) have shown that a protein that interacts with AREs and that migrates at 32 kDa in HeLa cell extracts is HuR, a known RNA stabilizer. Although one cannot predict the precise effect of a particular combination of ARE binding proteins in cells, the ability of these proteins to influence mRNA degradation in vitro and in cells is well established (71, 76, 81, 82, 89, 90, 91, 92, 93, 94, 95). Indeed, an increased affinity of the normal RNA for HuR would explain the increased stability of the normal RNA relative to the SNP RNA in F9 and HeLa cells (Fig. 6C). If the SNP also changes the molar ratio of proteins binding to the BMP2 mRNA in osteoblasts and chondrocytes, this might alter BMP2 protein levels at key stages of bone and cartilage differentiation and may explain the association of this SNP with osteoporosis and osteoarthritis (22, 25). Indeed, BMP2 is known to be aberrantly expressed in osteoarthritis (96, 97).
Abnormal BMP2 expression also is specifically associated with lung, breast, colon, pancreatic, and prostate cancers (98, 99, 100, 101, 102). Loss of normal BMP2 expression occurs in prostate and breast cancers and in the microadenomas of familial adenomatous polyposis patients (99, 101, 103). Recombinant BMP2 can inhibit cell proliferation and induce apoptosis in diverse cell types (e.g. Refs. 6, 7, 51, 101 , and 104, 105, 106, 107, 108). Thus, normal BMP2 protein levels may suppress tumors in some tissues, and altering BMP2 expression may promote the transition from normality to malignancy. Mutations in potent regulatory elements like the AREs in the ultra-conserved region of the BMP2 3'-UTR, may alter the physiology of any BMP2-sensitive tissue and influence the onset or progression of diverse human pathologies.
According to the Single Nucleotide Polymorphism database (dbSNP; http://www.ncbi.nlm.nih.gov/SNP/), about one third of Caucasian and Asian populations carry the rs15705 SNP. This suggests that 9% of the population may be homozygous for this alteration in the ultra-conserved region. If this polymorphism changes the affinity of proteins for the BMP2 3'-UTR in vivo, the stoichiometry of mRNA binding proteins and the delicate balance of stabilizing and destabilizing factors controlling RNA half-life would be altered. Thus, the rs15705 or other 3'-UTR SNPs could alter BMP2 protein levels in bone and influence susceptibility to osteoporosis and diseases in other BMP2-expressing tissues.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Bam+Sac were described previously (35, 37).
BmpLucM (nt –1237 to 471 and nt 9585–9999)
The SacI site in pGL1.7XX was removed by digestion, blunting with T4 DNA polymerase in the presence of dNTPs, and religation to create pGL1.7XXSacI. A DraIII and XbaI fragment containing the 1702-bp BMP2 promoter excised from pGL1.7XXSacI was inserted in place of the DraIII and XbaI fragment in pGL2BasicBam+Sac, creating pGLB25'SacI+SacI. Finally, a 395-bp fragment was removed from pGEM4MouseKA (35) with PvuII and SacI, blunted, then inserted into the SacI site of pGLB25'SacI+SacI.
BmpLucH (nt –1237 to 471 and Human Sequence Homologous to Mouse nt 9604–9999)
A 385-bp fragment was removed from pGEM4HumanKA (35) with HincII, blunted, digested with SacI, and then ligated into SacI site of pGLB25'SacI+SacI. The ligation mixture was then blunted again and religated.
BmpLucD (nt –1237 to 471 and Dog Sequence Homologous to Mouse nt 9392–9999)
A 533-bp fragment was removed from pGEM4DogKA (35) with EcoRI, blunted, and inserted into the blunted SacI site of pGLB25'SacI+SacI.
In Vitro Transcription Plasmids
pGemB2-KA (nt 9455–9938) and plasmids containing the homologous regions from human and zebrafish (35) were linearized with BamHI. pGBmp2-SacPst (nt 9,397–10,202) and pGBmp2-PvuIIPst (9,574–10,202) were linearized with AccI to make sense probe spanning nt 9,397–9,938 or 9,574–9,938, respectively (35). pGBmp2-PvuIIRsa (nt 9574–9735), pGBmp2-RsaAcc (nt 9735–9938), or wild-type or mutated TNF plasmids (79) were linearized with HindIII.
For a ribonuclease (RNase) protection probe specific to luciferase, a fragment was PCR amplified using the pGL2-basic plasmid as template and primers specific to the luciferase gene (5'-ATGGAAGACGCCAAAAACATAAAG-3' and 5'-ATAGCTTCTGCCAACCGAAC-3'). This fragment was cloned into the pCRII-TOPO cloning vector (Invitrogen, Carlsbad, CA) according to the manufacturer’s instructions. This plasmid (pCRII-Luc) and the M13 forward and reverse primers were used for PCR to generate T7 polymerase template for probe.
Transgene Construct (pGLB25'3'LacZloxpCNS)
This plasmid contains in order, the BMP2 distal promoter region (nt –1,237 to 471), the lacZ gene, a loxP site, BMP2 3'-UTR with the most highly conserved region (nt 9,392–10,200), another loxP site, BMP2 3'-UTR with poly(A) sites and downstream sequence (nt 10,200–11,444). Plasmid pBSB2–3'-UTR (35) was digested with PstI to remove nt 10,200–11,444, and then religated to create pBSB2–3'-UTRPst. Nucleotides 9,392–10,200 of BMP2 were removed by digesting this plasmid with SacI, and then blunted and then digested with EcoRI. pLoxp3'-UTR was created by inserting this 817-bp fragment into pGKLoxp, which had been digested with HindIII, blunted, and then digested with EcoRI. pGL1.7XXSacI was digested with KpnI and XhoI to release a 1712-bp BMP2 promoter-containing fragment. This was used to replace the promoter-containing KpnI and XhoI fragment in pGLB25'3'. This created pGLB25'3'SacI, which is identical with pGLB25'3' except the SacI site in the polylinker was deleted. pLoxp3'-UTR was digested with PacI and blunted. This 885-bp fragment containing nt 9,392–10,200 flanked by LoxP sites was inserted into pGLB25'3'SacI digested with PstI and SacI (blunted) to create pGLB25'3'Sac loxpCNS. The lacZ-containing plasmid, pBI-3 (109) was digested with XbaI, partially filled in with Klenow in presence of dTTP and dCTP, and then digested with XmnI, yielding a 3155-bp fragment containing the lacZ gene fused in-frame with a nuclear localization signal. pGLB25'3'Sac loxpCNS was digested with HindIII, partially filled in with Klenow in presence of dATP and dGTP, and EcoNI (blunted), and the 7949-bp fragment containing the BMP2 promoter and 3'-UTR regions without luciferase was isolated. The lacZ fragment was inserted to create pGLB25'3'LacZloxpCNS in which the HindIII, XbaI, XmnI, and EcoNI sites were destroyed.
Transgenic Mice
All animals were handled in accordance with the Guidelines for Care and Use of Experimental Animals and approved by the New Jersey Medical School Institutional Animal Care and Use Committee (protocol nos. 04086 and 00100). The transgene construct pGLB25'3'LacZloxpCNS was digested with SalI and KpnI to remove all but 24 bp of vector sequence. Linearized DNA fragment was microinjected into oocytes in University of Medicine and Dentistry of New Jersey-New Jersey Medical School transgenic/knockout facility (Dr. Carlos Molina, director). The ß-gal expression pattern of the reporter gene in transgenic mice was detected by X-Gal staining (110).
F9 cell culture, differentiation, and transfection have been described in Refs. 37 and 111 .
In Vitro Stability Measurements
After linearization, plasmids were transcribed with SP6 RNA polymerase with 7meGpppG and 32P-UTP (70, 79, 112). Capped and labeled RNAs were incubated in S18 cytoplasmic extracts as described in Refs. 67, 68, 70 , and 79 . RNAs and degradation products were visualized and quantified on an 8 M urea-containing 5% polyacrylamide (37.5:1, acrylamide:bis-acrylamide) gel (67, 68) using a Molecular Dynamics (Sunnyvale, CA) PhosphorImager and ImageQuant software.
UV Cross-Linking Reactions
UV cross-linking analysis was performed essentially as described (113). Briefly, 20–50 fmol of radiolabeled RNA was incubated in the cell extract under in vitro decay assay conditions for 10 min at 37 C. EDTA was added to a concentration of 1 mM to prevent RNA decay while proteins bound to the RNA. The reaction mixtures were irradiated with UV light for 10 min using a 15-W germicidal lamp at 2 µJ/sec at room temperature. A total of 100 ng of RNase A was added and incubated for 15 min at 37 C to degrade the labeled RNA. Labeled proteins were solubilized by heating at 95 C for 5 min in 8% glycerol, 1.0% SDS, 75 mM dithiothreitol, 0.05% Bromophenol Blue, 62.5 mM Tris (pH 6.8), analyzed on a 10% polyacrylamide gel (37.5:1, acrylamide:bis-acrylamide) in running buffer (192 mM glycine, 25 mM Tris-HCl, 0.1% SDS), and quantified as above.
EMSAs
Binding reactions contained 1 mM spermidine, 2 µg of poly(A), 50 fmol of capped and 32P-labeled RNA probe, 8 µl of cell extract, and 1 mM EDTA (to prevent RNA decay) in a final volume of 10 µl. All components were incubated at room temperature for 10 min. After adding glycerol to a final concentration of 30%, reactions were loaded on 4.5% nondenaturing polyacrylamide gels containing 1x Tris-borate EDTA, and electrophoresed at 200 V, constant voltage.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
D.T.F., S.J., J.X., and M.B.R. have nothing to declare.
First Published Online February 23, 2006
1 D.T.F. and S.J. contributed equally to this work. 
Abbreviations: ARE, AU-rich element; BMP, bone morphogenetic protein; nt, nucleotide; oligo, oligonucleotide; RA, retinoic acid; RNase, ribonuclease; SNP, single nucleotide polymorphism; SV40, simian virus 40; TTP, tristetraproline; UTR, untranslated region.
Received for publication November 23, 2005. Accepted for publication February 14, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONBMP2 and OsteoporosisNumerous Mechanisms Modulate BMP...BMP2 Single Nucleotide...RESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and 
in the regulation of bmp-2 and bmp-4 in F9 embryonal carcinoma cells. Cell Growth Differ 7:115–122[Abstract]
B specifically activates BMP-2 gene expression in growth plate chondrocytes in vivo and in a chondrocyte cell line in vitro. J Biol Chem 278:29130–29135
, nt –1237 to 471 relative to the promoter) upstream of the luciferase (LUC) gene alone or with the ultra-conserved region (nt 9574–9938) of the murine 3'-UTR (M) or the homologous human (H) or dog (D) sequence downstream of the pGL2-Basic luciferase gene. F9 cells were treated with RA, dibutyryl cAMP, and theophylline for 4 d, and cotransfected with a plasmid constitutively expressing ßgal to normalize transfection efficiency. These conditions induce maximal BMP2 expression. Average reporter activity is shown ± SEM, n 
4. B, Reporter transgenes with the murine BMP2 promoter region (nt –1237 to 471) and the 3'-UTR and downstream sequences (nt 9,392–11,604) flanking the lacZ gene were injected into oocytes. Toes removed to mark individual pups were whole-mount stained for ßgal activity. The dark staining indicates ßgal activity in the claw bed and phalanges (Ph) of transgenic mice, but not in the adjacent soft tissues or in the dense claw bed and phalanges of nontransgenic mice. C, Reporter genes with the murine Bmp2 promoter (nt –1237 to 471) upstream of the luciferase (LUC) gene alone (BmpLuc) or with nt 9,392–11,604 of mouse BMP2 3'-UTR and downstream sequence between the pGL2-Basic luciferase gene and the SV40 pA signal (SV) of pGL2-Basic (BmpLucBmp). "M" and "A A" mark the ultra-conserved region and the two BMP2 poly(A) sites, respectively. In BmpLucSVBmp, the 3'-UTR and downstream sequence was inserted after the SV40 pA signal. Average reporter activity (gray bars) at 4 d of differentiation is shown ± SEM, n 
) shows the position of the AUUUA that is converted to AUUUC in the rs15705 SNP. The number of AREs (AUUUA or AUUUUA) and in vitro half-lives (t1/2 ± SEM) of the RNAs in extracts from F9 cells are shown to the right of each bar. The t1/2 value for the longest BMP2 RNA (9455–9938) (n = 4) and RNAs containing the wild-type ARE from TNF
) or nonspecific (NS, 
) competitor oligos. D, RNAs with the human BMP2 ultra-conserved region were UV cross-linked in HeLa extracts with 5, 10, or 25 pmol of TNF
