| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Minerva Center for Calcium and Bone Metabolism, Nephrology Services, Hadassah-Hebrew University Medical Center, Jerusalem, Israel; and
Istituto Nazionale Ricerca sul Cancro, Genova, Italy
1Correspondence: Minerva Center for Calcium and Bone Metabolism, Hadassah Hebrew University Medical Center, PO Box 12000, Jerusalem, Israel 91120. E mail: tally{at}cc.huji.ac.il
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: AUF1 • Unr • ARE • calcium • phosphorus • chronic • kidney disease
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, whereas low serum phosphate has the opposite effects (2)
. Understanding the mechanism by which calcium and phosphate regulate PTH gene expression is important in physiology and in pathological states, especially chronic kidney disease. Experimental chronic kidney disease leads to secondary hyperparathyroidism, characterized by an increase in levels of PTH mRNA, serum PTH, and parathyroid cell proliferation similar to the secondary hyperparathyroidism due to dietary calcium depletion. The changes in PTH mRNA levels due to calcium or phosphorus depletion as well as experimental chronic kidney disease are post-transcriptional (3
, 4)
. They are mediated, in part, by regulated binding of stabilizing trans-acting factors to a 63-nt long cis-acting element located in the PTH mRNA 3'-untranslated region (3'UTR) (3
, 5)
. In vitro degradation assays (IVDA) using parathyroid extracts reproduce the differences in PTH mRNA stability induced by calcium and phosphate in vivo; calcium depletion leads to PTH mRNA stabilization, whereas phosphorus depletion leads to destabilization of PTH mRNA compared to parathyroid extracts from control rats (3)
. This correlates with PTH mRNA levels in vivo (6)
. The PTH mRNA 3'-UTR 63 nt element is both necessary and sufficient for this regulation (5)
. This element is a type III AU rich element (ARE) that does not contain the classical AUUUA pentamer (7)
. A 26 nt element within the 63 nt sequence is the minimal protein binding region and is conserved among species (5)
. AREs are targets for trans-acting proteins regulating mRNA stability and translation (8)
. A number of ARE binding proteins have been identified that can interact with AU- and U-rich regions. K-homology splicing regulator protein (KSRP) is an example of decay promoting factors (8
, 9)
. KSRP recruits the multiprotein 3'-5 exoribonuclease complex, exosome (10)
, to target mRNAs (9)
. The central part of the KSRP contains 4 adjacent K homology (KH) domains that are required to ensure its interaction with the decay-promoting machinery. Other proteins, such as HuR, are stabilizing factors (11)
. AU rich binding factor 1 (AUF1) promotes either decay or stabilization, depending on the mRNA and cell type (12)
. AUF1 and upstream of N-ras (Unr) bind to PTH mRNA, increasing its half-life (t1/2) (13
, 14)
. Recombinant AUF1 specifically stabilizes PTH mRNA when added to parathyroid extracts in IVDAs (13)
. Overexpression of Unr in human embryonic kidney (HEK) 293 cells specifically stabilizes co-transfected PTH mRNA or chimeric reporter GH mRNA containing the PTH mRNA 3'-UTR 63 nt ARE (14)
. Small-interfering RNA (siRNA) mediated knock-down of either AUF1 or Unr led to the opposite effect, decreasing PTH mRNA levels (14
, 15)
. However, the complete molecular mechanisms controlling PTH mRNA decay have not yet been defined. Herein, we identify KSRP as a PTH mRNA binding protein. KSRP/PTH mRNA interaction in intact parathyroid glands is regulated by changes in serum calcium or phosphate concentrations and experimental chronic kidney disease and correlates with instability of PTH mRNA. We show that the exosome is necessary for PTH mRNA decay in vitro and interacts with KSRP in parathyroid extracts. In transfected cells, KSRP binds to the PTH mRNA ARE and specifically decreases PTH mRNA levels. KSRP-PTH mRNA binding is blocked by the PTH mRNA binding and stabilizing protein AUF1 p45 isoform. Overexpression of this isoform also attenuates the KSRP mediated decrease in PTH mRNA steady-state levels in transfected cells. We suggest that KSRP-PTH mRNA interactions control PTH mRNA t1/2 by recruitment of a degradation complex to PTH mRNA.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Calcium restriction resulted in a serum calcium of 4.9 ± 0.4 mg/dl (n=20); serum calcium in control rats was 10.8 ± 0.3 mg/dl (n=10). Phosphorus restriction resulted in a serum phosphate of 4.1 ± 0.7 mg/dl and serum calcium of 12.8 ± 0.6 mg/dl (n=25); control rats serum phosphate was 9.9 ± 0.5 mg/dl (n=10). Adult male rats (100–120 g) were fed a control diet or a renal failure-inducing 0.75% adenine, high-phosphorus (1.5%) diet (4)
. At 7 days, serum creatinine levels were significantly increased (control rats: 0.27±0.03 vs. 0.60±0.07 µM in the adenine-fed rats). Serum phosphate was also significantly increased (control rats: 9.8±0.2 vs. 10.75±0.25 mg/dl in adenine-fed rats) and serum PTH was increased from 56 ± 28 pg/ml in control rats to 338 ± 113 pg/ml in the adenine-fed rats. All animal experiments were approved by the Institutional Animal Care and Use Committees.
Cell cultures and transient transfection
HEK293 cells were transiently transfected with the indicated plasmids using a Ca-phosphate transfection kit (Sigma-Aldrich, St. Louis, MO, USA). siRNA oligonucleotides were transiently co-transfected with expression plasmids, using Lipofectamine 2000 Reagent (Invitrogen, Carlsbad, CA, USA). siRNAs were used according to the manufacturer’s instructions. Transient transfections were performed in 24-well plates in duplicate or triplicate for RNA analysis and in 10 cm plates for protein extractions. Total amount of DNA in co-transfection experiments was maintained constant using an empty vector, as indicated.
RNA immunoprecipitation (RIP)
Thyroparathyroid glands (a pool of 6) were roughly chopped and cross-linked with 1% formaldehyde in PBS for 1 h at room temperature. The reaction was stopped by 0.25 M glycine (pH 7) at room temperature for 5 min followed by a wash with ice-cold PBS. Tissue was suspended in radioimmunoprecipitation assay (RIPA) buffer supplemented with ribonuclease (RNase) inhibitors (Promega Corp., Madison, WI, USA) and homogenized using a polytron homogenizer. Equal amounts of whole cell extracts were immunoprecipitated with Protein A agarose–bound anti-KSRP or AUF1 antibody beads (EMD Biosciences Inc., San Diego, CA, USA) or IgG as control. The beads were washed with modified RIPA buffer supplemented with 1 M NaCl, 1% sodium deoxycholate, 1 mM EDTA, and 2 M urea, suspended with 50 mM Tris (pH 7.5), 5 mM EDTA, 10 mM dithiothreitol (DTT), and 1% SDS and heated to 70°C for 1 h to reverse cross-linking. RNA was extracted using Tri Reagent (Molecular Research Center, Inc., Cincinnati, OH, USA) (16)
, reverse transcribed with random hexamer primers using a Maxime RT premix kit (iNtRON Biotechnology, Gyeonggi-do, Korea), and analyzed by real-time quantitative polymerase chain reaction (qPCR) (conducted using ABI Prism 7901 Sequence Detection System; Applied Biosystems, Foster City, CA, USA; and SYBR Green ROX Mix; ABgene, Epsom, UK). In some experiments, transfected HEK293 cells were analyzed by RIP and reverse transcriptase-polymerase chain reaction (RT-PCR) products run on agarose gels.
Protein extractions
For IVDA, postmitochondrial extracts were prepared. Cultured cells or parathyroid glands (pools of 6) were incubated on ice for 10 min in an extraction buffer containing 0.25 M sucrose, 30 mM Tris HCl (pH 7.5), 2 mM DTT, and a protease inhibitor mix. Tissue samples were homogenized, and the supernatant was cleared by centrifugation at 15,000 g for 15 min (4°C). For Western blots and RIP, cultured cells or parathyroid glands (pools from 5 rats) were prepared using RIPA buffer containing 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), and protease inhibitors. For UV cross-linking experiments, cells were suspended in a buffer containing 10 mM NaCl, 20 mM 1,4-piperazinebis(ethane sulfonic acid (pH 7), 0.5% Nonidet P-40, 0.05% 2-mercaptoethanol, and 5 mM EDTA and then centrifuged for 10 min at 4°C. Extracts were stored in aliquots at –80°C.
Exosome depletion
Parathyroid postmitochondrial extracts (400 µg) were immunoprecipitated using antibodies against exosome components hRrp40, hRrp41, and PM/Scl100 simultaneously (17)
or with IgG (mock depletion). Exosome immunodepletion was confirmed by immunoblots with antibody against Rrp46.
Immunoblots
Proteins were analyzed by SDS-PAGE immunoblots, as described elsewhere (15)
.
RNA transcription and labeling
Unlabeled and uniformly [
-32P]UTP-labeled transcripts were prepared as described elsewhere (5)
. For IVDA, a transcript for the full-length mRNA was transcribed from the full-length PTH cDNA cloned into pBluescript II KS (Stratagene, La Jolla, CA, USA) and linearized by SmaI (3)
or a PTH cDNA from which the ARE (90 nt) was removed. A construct containing the truncated rat PTH cDNA in which the 90 nt was inserted in the reverse orientation was also used. The resulting template cDNAs contained a stretch of 
150 bp of dT at the 3'-end. In some experiments, uniformly labeled and capped RNAs were transcribed in vitro in the presence of m7GpppG cap analog (Promega, Madison WI, USA) according to the manufacturer’s instructions.
RNA pull-down
Transiently transfected HEK293 cells or parathyroid extracts (300 µg) were precleared with 50 µl of streptavidin-coupled agarose beads for 1 h at 4°C. Full-length PTH mRNA was transcribed in vitro as described above, using biotin-tagged nucleotides (Roche, Mannheim, Germany). RNA (1 µg) was incubated with precleared extracts in binding buffer containing 10 nm HEPES, 3 mM MgCl2, 40 mM KCl, 5% glycerol, and 0.2 mg/ml heparin for 3 h at 4°C, after which streptavidin-agarose beads were added for an additional 2 h at 4°C. Eluted proteins were washed with RIPA buffer and separated on SDS-PAGE and analyzed by Western blots. Nonbiotinylated competitor PTH or GH RNA were used where indicated.
UV cross-linking
UV cross-linking was performed as described, using either full-length PTH mRNA or deletion mutants and HEK293 cell extracts (5)
. For immunoprecipitation of protein-RNA complexes, samples, after RNase A treatment, were diluted in RIPA buffer, incubated with anti-Flag or IgG protein A sepharose at 4°C overnight, and analyzed by SDS-PAGE and autoradiography.
IVDAs
Radiolabeled transcripts (200,000 counts/min) were incubated with 40 µg protein extract from either parathyroid glands or HEK293 cells in a volume of 50 µl and in a reaction buffer containing 3 mM Tris HCl (pH 7.5), 2 mM MgCl2, 3 mM NaCl, 10 mM ATP, and 36 U RNasin (Promega). At timed intervals, samples were removed; RNA was extracted, separated on agarose gels, and analyzed by autoradiography as described elsewhere (3)
.
PCR primers
Rat PTH sense and antisense primers (18)
were 5'-TTGTCTCCTTACCCAGGCAGAT-3' and 5'-TTTGCCCAGGTTGTGCATAA-3'. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) sense and antisense primers were 5'-GCAACTCCCATTCTTCCACC-3' and 5'-CATACCAGGAAATGAGCTTCACAA-3'. Primers for human GH mRNA were 5'-GGGAGGCTGGAAGATGGC-3' and 5'-CGTTGTGTGAGTTTGTGTCGAAC-3'. Primers for hypoxanthine-guanine phosphoribosyltransferase (HPRT) were 5'-CCCAGCGTCGTGATTAGTGA-3' and 5'-CCAAATCTTCAGCATAAGGTAT-3'.
siRNAs
Previously published siRNAs targeting the control CAT sequence, 5' r(GACGGUGAGCUGGUGAUAU)d(TT)-3' or TTCTCCGAACGAACGTGTCACGT-3'; KSRP: 5'-AAGATCAACCGGAGAGCAAGA-3' (19)
; and an additional set of commercial siRNAs for KSRP (no sequence available) were all synthesized by Qiagen (Hilden, Germany). siRNAs were used for transfection experiments according to the manufacturer’s instructions.
Plasmids
Rat PTH cDNA was cloned in either pcDNA3 (14)
for transfections or pBluescript II KS (3)
for in vitro transcription. Plasmids expressing different regions of PTH mRNA are described elsewhere (5)
. For IVDA, the pBluescript II KS plasmid containing the full-length rat PTH cDNA including a stretch of 150 dT nucleotides that by in vitro transcription produced a poly(A) tail was used. A BsaI-BclI fragment of the pBluescript II KS-PTH cDNA plasmid was removed by partial restriction enzyme digestion followed by ligation to produce a plasmid without the ARE. A construct containing the truncated rat PTH cDNA in which the deleted ARE was inserted in the reverse orientation was also used. The human PTH gene, including exons and introns, was in pcDNA3. KSRP in pcDNA3 contained either Flag-tagged full-length KSRP or different KSRP KH domains (9)
. The pGL2 Luciferase plasmid contained the cytomegalovirus (CMV) promoter (pGl2-CMV) (Promega). The growth hormone expression plasmid was kindly provided by O. Meyuhas (Hebrew University-Hadassah Medical School, Jerusalem, Israel) (20)
. The GH-PTH mRNA 63 nt plasmid was previously described (16)
and contained the 63 nt rat PTH mRNA ARE cloned between the 3' of the GH mRNA coding sequence and the GH mRNA 3'-UTR. A GH cDNA cloned into pBluescript II KS was used to in vitro transcribe GH RNA. Empty control vectors (PCDNA3 and pSG5) were used as indicated. Expression plasmids for myc-AUF1 isoforms were kindly provided by A.-B. Shyu (University of Texas Houston Health Science Center, Houston, TX, USA) (21)
.
Antibodies
Anti-KSRP was described elsewhere (9)
. The anti-Flag, anti -tubulin were from Sigma-Aldrich. Anti-myc was from Cell Signaling (Boston, MA, USA) and anti-GAPDH from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies for exosome components were kindly provided by R. Raijmakers (Radboud University Nijmegen, Nijmegen, The Netherlands) (17)
.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. To determine whether the ARE-binding protein KSRP is involved in the regulation of PTH mRNA stability, we performed RIP experiments in intact rat parathyroid glands. Parathyroid glands were cross-linked, solubilized, and immunoprecipitated with either anti-KSRP antibody or IgG. RNA extracted from both input and immunoprecipitated samples was analyzed by real-time qRT-PCR. Anti-KSRP antibody specifically immunoprecipitated PTH mRNA but did not interact with either GAPDH (Fig. 1A
) or HPRT (data not shown) mRNAs, suggesting that KSRP is part of the PTH mRNA binding complex in the parathyroid.
|
PTH gene expression is regulated post-transcriptionally by diet-induced changes is serum calcium and phosphate concentrations (3)
. We therefore studied PTH mRNA-KSRP interactions in parathyroid glands of rats fed control or calcium- or phosphorus-restricted diets (see Materials and Methods). Calcium restriction markedly increased and phosphorus restriction decreased PTH mRNA levels compared to controls (Supplemental Fig. 1A). This effect is due to regulation of its t1/2 (3)
. Similarly, input PTH mRNA levels, analyzed by qRT-PCR, were significantly higher in low-calcium extracts and lower in low-phosphorus extracts compared to control (Fig. 1B
, left panel). However, anti-KSRP antibody immunoprecipitated 5-fold more PTH mRNA in parathyroid glands of rats fed the phosphorous-restricted diet compared to rats fed a control diet. In contrast, 4 times less PTH mRNA was immunoprecipitated in parathyroid glands of rats fed the calcium-restricted diet (Fig. 1B
, middle panel). The differences in KSRP-PTH mRNA interactions in low-phosphorus and low-calcium extracts inversely correlate with the stability and steady-state levels of PTH mRNA observed in these rats (ref. 3
; Supplemental Fig. 1A; Fig. 1B
, left panel). We have previously shown that AUF1 is a PTH mRNA-binding protein (13)
. Anti-AUF1 immunoprecipitated more PTH mRNA in parathyroid glands from rats fed a low-calcium diet than control parathyroids. In parathyroid glands from rats fed a low-phosphorus diet, AUF1 bound less PTH mRNA than both control and low-calcium parathyroids (Fig. 1B
, right panel), consistent with our earlier observations that AUF1 binding correlates with PTH mRNA stabilization (13)
.
Experimental kidney failure is an additional model for secondary hyperparathyroidism characterized by a marked increase in PTH gene expression dependent on post-transcriptional control (4)
. We also studied PTH mRNA-KSPR interactions in the parathyroid glands of these rats. Mature rats were fed a control or an adenine-high phosphorous diet for 7 days to induce kidney failure. Kidney failure led to a marked increase in PTH mRNA levels by Northern blot (Supplemental Fig. 1B), as in our previous reports (4)
. Despite the fact that PTH mRNA levels were more than 3-fold higher in input extracts of the adenine-fed rats by qRT-PCR (Fig. 1C
, left panel), anti-KSRP antibody precipitated 4-fold less PTH mRNA in these extracts compared to parathyroid extracts from control rats. Therefore, KSRP specifically interacts with PTH mRNA in parathyroid glands and this interaction correlates with decreased steady state levels and stability of PTH mRNA.
The exosome is required for PTH mRNA degradation by parathyroid extracts
We previously reported (9)
that KSRP interacts with the exosome and recruits this enzymatic complex to ARE containing mRNAs in HeLa cells. We reproduced this interaction in parathyroid extracts (Fig. 2A
). IVDAs using parathyroid extracts recapitulate the differences in PTH mRNA stability induced by calcium and phosphate in vivo (3)
. We next performed IVDA with exosome-immunodepleted parathyroid extracts. Depletion was achieved using antibodies directed to three exosome components, PM/Scl100, Rrp40, and Rrp41, simultaneously. Immunoblot analysis with antibody for an additional exosome component, Rrp46, confirmed exosome depletion (Fig. 2B
). Uniformly labeled polyadenylated full-length rat PTH mRNA or PTH mRNA lacking the ARE was subjected to IVDA with either exosome-depleted or mock-depleted parathyroid extracts (Fig. 2C
). The transcript was stable throughout the experiment when no extract was added (Fig. 2C
, right panel). The full-length PTH mRNA transcript was degraded in a time-dependent manner by mock depleted parathyroid extracts (Fig. 2C
, top panel) with t1/2 = 45 min. Exosome immunodepletion prevented the decay of the full-length PTH mRNA transcript (t1/2>120 min) (Fig. 2C
). In contrast, exosome depletion did not affect the stability of a polyadenylated transcript lacking the PTH ARE (t1/2>120 min) (Fig. 2C
, bottom panel). In addition, exosome depletion had no effect on the decay of a PTH transcript with the ARE in the reverse orientation as a size-matched RNA control (Supplemental Fig. 2). These results demonstrate that KSRP interacts with the exosome in the parathyroid and that the exosome is required for PTH mRNA decay by parathyroid extracts.
|
KSRP binds to the 63 nt ARE-containing region of PTH mRNA 3'-UTR
RIP analysis showed that KSRP specifically interacts with PTH mRNA in parathyroid glands (Fig. 1)
. We also studied KSRP-PTH interactions in parathyroid extracts and in transfected HEK293 cells by RNA pull-down experiments using biotinylated full-length rat PTH mRNA and streptavidin agarose beads. KSRP was specifically recovered by PTH RNA (Fig. 3A bound, lanes 2 and 5
) in parathyroid extracts. Increasing concentrations of a nonbiotinylated PTH RNA used as competitor prevented the binding of KSRP to PTH RNA (Fig. 3A
, left panel, lanes 3 and 4). A nonspecific GH RNA had no effect on binding (Fig. 3A
, right panel, lanes 6 and 7). To characterize KSRP-PTH mRNA interactions in greater detail, we studied KSRP-PTH mRNA binding in HEK293 cells transiently transfected with a Flag-KSRP expression plasmid or a control plasmid. RNA pull-down as above was performed with Flag-KSRP transfected cell extracts. Immunoblot analysis with anti-Flag antibody showed that KSRP was specifically recovered by PTH RNA (Fig. 3B
, left panel, bound) and this was prevented by increasing concentrations of a nonbiotinylated PTH RNA used as competitor (Fig. 3B
, right panel), but not by a nonspecific GH RNA (not shown).
|
Next we used UV cross-linking assays to further define the interaction of KSRP with PTH mRNA. HEK293 cells were transiently transfected with expression plasmids for either full-length Flag-tagged KSRP or for the deletion mutant containing all 4 KH domains of KSRP (KSRP1–4) or just two KH domains, either KH12 (KSRP1–2) or KH34 (KSRP3–4) (9)
. Immunoblotting confirmed the expression of Flag-tagged KSRP polypeptides in the input extracts (Fig. 3C; D
, left panel). Cell extracts were analyzed by UV cross-linking with radiolabeled rat PTH full-length transcript. RNA-protein complexes were detected in extracts expressing Flag-KSRP, KSRP1–4, and KSRP3–4, but not in extracts expressing KSRP1–2 (Fig. 3C
, right panel, input; D, right panel). To characterize the PTH mRNA-protein complexes, UV cross-linking reactions were immunoprecipitated with either anti-Flag antibody or control mouse IgG and then analyzed by gel electrophoresis. Both full-length KSRP and KSRP1–4 protein-RNA complexes were specifically recovered by anti-Flag antibody (Fig. 3C
, right panel, lanes 3 and 6). These results show that transfected Flag-KSRP interacts with PTH mRNA and that its KH domains 3–4 are sufficient for this association.
To determine the PTH mRNA-KSRP interacting site, UV cross-linking experiments were performed using mRNA fragments encompassing different regions of rat PTH mRNA (Fig. 3E
) and total extracts from HEK293 cell transiently transfected with either Flag-KSRP or Flag-KSRP1–4 plasmid. KSRP1–4 contains the four KH domains, and as expected (9)
, was fully able to bind PTH mRNA (Fig. 3C, D
). Therefore, we used KSRP1–4 for further studies. Both KSRP1–4 and KSRP interacted with the PTH mRNA 3'-UTR (Fig. 3F
, lane 2, and data not shown) as well as with transcripts comprising the PTH mRNA AREs (Fig. 3F
, lanes 3 and 4, and data not shown). Conversely, neither KSRP1–4 nor KSRP bound to a PTH transcript lacking the 3'-UTR (Fig. 3F
, lane 1, and data not shown). These results indicate that KSRP binds to the previously defined 63 nt PTH mRNA ARE (5)
. A molar excess of the 3'-UTR 100 nt transcript, but not the transcript lacking the 3'-UTR, competed with the full-length PTH mRNA for the interaction with KSRP, indicating specificity of the binding (Fig. 3G
). We further confirmed the specificity of KSRP interaction with the PTH mRNA 63 nt ARE by using a chimeric construct expressing GH mRNA fused to the rat PTH 63 nt ARE (GH63) (5)
. Lysates from HEK293 cells transiently transfected with Flag-KSRP1–4 and either a construct expressing wild type human GH mRNA or the GH63 mRNA were subjected to RIP analysis. Immunoprecipitation with anti-Flag antibody followed by RT-PCR for GH mRNA showed that KSRP specifically interacts with the reporter GH63 mRNA (Fig. 3H
, lane 6) and not with wild type GH mRNA (Fig. 3H
, lane 3). These experiments indicate that KSRP-PTH mRNA interaction is mediated by the PTH mRNA 3'-UTR 63 nt ARE.
Transfected KSRP decreases PTH mRNA levels and promotes the decay of PTH mRNA in HEK293 cells by targeting the PTH mRNA ARE
Because a parathyroid cell line is not available, the effect of KSRP on PTH mRNA levels was studied in HEK293 cells cotransfected with expression plasmids for PTH and Flag-tagged KSRP or Flag-KSRP1–4. KSRP expression decreased both human and rat PTH mRNA levels with no effect on cotransfected luciferase or endogenous L32 mRNAs, and 18S ribosomal RNA (Fig. 4A
). Both PTH and luciferase expression were driven by a CMV promoter, but only PTH was affected by KSRP, suggesting a post-transcriptional effect of KSRP on PTH mRNA levels. PTH mRNA stability was measured by IVDA with transfected cell extracts. IVDA was performed using full-length radiolabeled rat PTH mRNA (as in Fig. 2
) and extracts from cells transfected with either control or expression plasmids for KSRP or KSRP1–4. Flag-KSRP1–4 or Flag-KSRP overexpression accelerated PTH mRNA decay (Fig. 4B
, top panel; C). Flag-KSRP overexpression had no effect on the decay of PTH mRNA lacking the ARE (Fig. 4B
, bottom panel).
|
We then confirmed that the regulation of PTH mRNA levels by KSRP is exerted via the PTH mRNA ARE, with the GH reporter gene containing the rat PTH 63 nt ARE (GH63) used in Fig. 3H
. Overexpression of KSRP decreased GH63 mRNA levels but had no effect on wild-type GH mRNA levels (Fig. 4D, E
). Our results indicate that KSRP specifically decreases steady-state PTH mRNA levels through the PTH mRNA ARE.
Next, we decreased KSRP protein levels by 2 different sets of siRNAs targeting KSRP and assessed the resulting effects on cotransfected PTH mRNA levels (Fig. 5
and data not shown). A 60% reduction in KSRP expression levels (Fig. 5A
) led to an approximately 2-fold increase in PTH mRNA levels (Fig. 5B
). Similarly, KSRP depletion prolonged the t1/2 of PTH mRNA compared to control CAT depletion in IVDA (Fig. 5C, D
). Together, our data indicate that KSRP interacts with PTH mRNA ARE and promotes PTH mRNA decay.
|
AUF1p45 competes with KSRP for PTH mRNA binding and partially blocks the KSRP-induced decrease in PTH mRNA in transfected cells
We have previously shown that AUF1 is a PTH mRNA binding and stabilizing protein (13)
. We over-expressed each of the 4 AUF1 isoforms together with Flag-KSRP in HEK293 cell. KSRP-PTH mRNA interaction was then studied by UV cross-linking analysis and RNA pull-down experiments. Overexpression of myc tagged AUF1 isoforms p42 and p45, but not p37 and p40, dose dependently decreased KSRP-PTH mRNA interaction by UV cross-linking (Fig. 6A
). Extracts from HEK293 cells transfected with each of the AUF1 isoforms were also analyzed by RNA pull-down experiments with biotinylated full-length rat PTH mRNA and streptavidin agarose beads. Interestingly, PTH mRNA specifically pulled down AUF1p37, p40, and p42, but not AUF1p45 (Fig. 6B
). We then studied the effect of AUF1 isoforms overexpression on KSRP-dependent reduction of PTH mRNA levels. Cells were cotransfected with expression plasmids for PTH, KSRP, and each of the AUF1 isoforms. As expected, KSRP overexpression decreased PTH mRNA levels (Fig. 6C, D
; also see Fig. 4A
). Overexpression of AUF1p45 isoform prevented in part the KSRP mediated decrease in PTH mRNA steady-state levels (Fig. 6C, D
). AUF1p37 and p40 had no effect, while AUF1p42 produced limited effects. AUF1p45 markedly increased PTH mRNA levels when transfected in the absence of KSRP, although the effect of the other isoforms was not significant (Fig. 6D
and data not shown). These results indicate that AUF1p45 affects both KSRP-PTH mRNA binding and KSRP-mediated decrease in PTH mRNA steady state levels in transfected cells. Together our data suggest that AUF1p45, although unable to bind PTH mRNA, indirectly affects PTH mRNA levels.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. We show that KSRP specifically interacts with PTH mRNA in parathyroid glands by both RIP analysis and RNA pull-down experiments. The increase in PTH mRNA stability by calcium depletion and kidney failure correlates with decreased KSRP-PTH mRNA binding, compared to controls by RIP. Phosphorus depletion produced opposite effects. These results indicate that KSRP directly or indirectly responds to changes in serum calcium and phosphate concentrations and renal failure by altering its association with PTH mRNA. In addition, using parathyroid glands we confirmed by RIP analysis that AUF1 interacts with PTH mRNA and that this association is increased in parathyroids from rats fed a low-calcium diet and decreased in parathyroids from rats fed a low-phosphorus diet compared to rats fed a control diet, as in our previous study (13)
. The pattern of interactions of KSRP and AUF1 with PTH mRNA suggest that these proteins have opposing roles in the regulation of PTH gene expression in vivo.
KSRP is a RNA-binding protein implicated in a variety of cellular processes, including transcription, alternative pre-mRNA splicing, and editing as well as mRNA localization and stability (9
, 22
23
24)
. On binding to ARE, KSRP promotes rapid mRNA decay of several inherently labile mRNAs in vivo in cultured cells and in vitro, recruiting the 3' to 5' exoribonucleolitic complex exosome to the RNAs (9)
. In this study, we show that KSRP interacts with the exosome in parathyroid extracts and that the exosome is necessary for PTH mRNA decay, as indicated by in vitro degradation experiments. PTH mRNA contains a 63 nt ARE-like region in its 3'-UTR that determines PTH mRNA stability (3
, 16)
. There is no parathyroid cell line, but in IVDA, parathyroid extracts from calcium-depleted rats lead to PTH mRNA stabilization, whereas parathyroid extracts from phosphorus-depleted rats lead to destabilization of PTH mRNA when compared to parathyroid extracts from control rats (3)
. The differences in PTH mRNA stability in the IVDA with different parathyroid extracts correlate with PTH mRNA levels in vivo (6)
and are dependent on the PTH mRNA 3'-UTR ARE (5)
. Addition of recombinant AUF1 stabilizes the PTH transcript in the IVDA, whereas addition of BSA or another PTH mRNA binding protein, LC8 (25)
, have no effect on PTH mRNA decay in this assay (13)
.
We report that in transfected cells, Flag-KSRP interacts with the PTH mRNA ARE and that its RNA-binding KH domains 3–4 are sufficient for this association, as reported previously for other mRNAs (9)
. The 4 adjacent KH domains of KSRP recognize AREs and interact with the mRNA degradation machinery to promote rapid decay of several target mRNAs. All 4 KH domains are necessary to ensure its interaction with the decay-promoting machinery (9
, 26)
. Overexpression of KSRP or KSRP1–4 specifically decreased both rat and human PTH mRNA levels in cotransfected HEK293 cells and accelerated PTH mRNA decay. The effect of KSRP was dependent on the PTH mRNA ARE. Conversely, KSRP knock-down increased both PTH mRNA stability and PTH mRNA steady-state levels. Together these results point to KSRP as a regulator of PTH mRNA levels in HEK293 cells. The role of KSRP in the regulation of PTH gene expression in response to changes in extracellular calcium and phosphate levels in parathyroid cells can not be studied directly due to the lack of a functional parathyroid cell line.
KSRP is a phosphoprotein; we recently demonstrated that stimuli that alter its phosphorylation status affect its decay promoting function (27
, 28)
. In addition, it has recently been reported that KSRP is arginine-methylated, and this may affect stabilization of specific KSRP mRNA targets in the etiology of spinal muscular atrophy (29)
. Changes in calcium and phosphate levels may induce post-translational modifications of KSRP in the parathyroid, thus affecting its ability to promote PTH mRNA degradation. Interestingly, calcium and phosphorus depletion as well as kidney failure lead to post-translational modifications of the PTH mRNA stabilizing protein AUF1, with no change in AUF1 protein levels (4
, 15)
. Preliminary results suggest that KSRP protein levels are also unchanged in parathyroids from either calcium- or phosphorous-depleted rats. However, calcium and phosphorus depletion lead to post-translational modification of KSRP (unpublished data). It is possible that a single signal-transduction pathway regulates PTH mRNA turnover targeting both AUF1 and KSRP.
KSRP (this study) and AUF1 (13)
bind to the same PTH mRNA 3'-UTR 63 ARE. AUF1 stabilizes, whereas KSRP destabilizes PTH mRNA. The AUF1 gene is expressed as 4 isoforms: p37, p40, p42, and p45 (30)
. Knock down of all AUF1 isoforms by siRNA decreased cotransfected PTH gene expression (15)
. Addition of recombinant AUF1 to rat PT extracts in IVDA experiments specifically stabilized the PTH mRNA transcript (13)
. Herein we show that transfected AUF1p45 did not bind PTH mRNA in RNA pull-down experiments performed in HEK293 cells. However, overexpression of this isoform competed for KSRP-PTH mRNA binding and partially prevented the KSRP-mediated decrease in PTH mRNA levels in transfected cells. We have recently demonstrated, by yeast 2-hybrid screening and coimmunoprecipitation, that AUF1p45 isoform interacts with KSRP but does not bind to KSRP target transcripts identified by microarray screening (27)
. AUFp45 may therefore affect PTH mRNA stability without directly interacting with the PTH transcript. It is noteworthy that, although all AUF1 isoforms bind PTH mRNA in parathyroid glands and recombinant AUF1p40 and p37 stabilize PTH mRNA in IVDAs (13)
, only AUF1p45 interferes with KSRP-PTH ARE binding and KSRP-mediated decrease in PTH mRNA levels in transfected HEK293 cells. The regulated interaction of AUF1 with PTH mRNA in parathyroids from rats fed calcium- or phosphorus-depleted diets by RIP analysis (Fig. 1B
) does not distinguish between the interactions of the different AUF1 isoforms.
In summary, we have identified KSRP as a PTH mRNA binding protein and regulator of PTH mRNA levels. In parathyroid glands, KSRP interaction with PTH mRNA is regulated by changes in serum calcium and phosphate and by adenine-induced kidney failure. Phosphorus restriction increases the association of KSRP with PTH mRNA. On interaction with PTH mRNA ARE, KSRP may recruit the exosome to degrade PTH mRNA. Calcium depletion induces the formation of a stabilizing complex, consisting of AUF1 and Unr, which may block the interaction of KSRP with PTH mRNA, leading to increased PTH mRNA stability and levels (Fig. 7
). Similar interactions occur in parathyroids of rats with kidney failure compared to control rats. In the absence of a parathyroid cell line, the role of the PTH mRNA interacting proteins cannot be studied directly in the parathyroid. Targeted deletion of the genes coding for these proteins specifically in the parathyroid would provide a further opportunity to support this model in vivo. On the basis of the regulated association of AUF1 and KSRP with the PTH ARE, we propose a role for these proteins in the regulation of PTH mRNA decay and mRNA levels in response to either dietary-induced changes in calcium and phosphate levels or chronic kidney disease.
|
|
ACKNOWLEDGMENTS |
|---|
Received for publication February 10, 2008. Accepted for publication May 22, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Pollock and S. Huang The Perinucleolar Compartment Cold Spring Harb Perspect Biol, February 1, 2010; 2(2): a000679 - a000679. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Lavi-Moshayoff, J. Silver, and T. Naveh-Many Human PTH gene regulation in vivo using transgenic mice Am J Physiol Renal Physiol, September 1, 2009; 297(3): F713 - F719. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Nechama, I. Z. Ben-Dov, J. Silver, and T. Naveh-Many Regulation of PTH mRNA stability by the calcimimetic R568 and the phosphorus binder lanthanum carbonate in CKD Am J Physiol Renal Physiol, April 1, 2009; 296(4): F795 - F800. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Xu, J. R. McCarrey, and N. B. Hecht A cytoplasmic variant of the KH-type splicing regulatory protein serves as a decay-promoting factor for phosphoglycerate kinase 2 mRNA in murine male germ cells Nucleic Acids Res., December 1, 2008; 36(22): 7157 - 7167. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
