| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
during hypoxia
Department of Genome Science, Genome Research Institute, College of Medicine, University of Cincinnati, Cincinnati, Ohio, USA
2Correspondence: Department of Genome Science, University of Cincinnati College of Medicine, 2180 E Galbraith Rd., Cincinnati, OH 45267-0505, USA. E-mail: maria.czyzykkrzeska{at}uc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunits. We have demonstrated that hypoxia stimulates translation of HIF-1 and -2 proteins by distributing HIF- mRNAs to larger polysome fractions. This requires influx of extracellular calcium, stimulation of classical protein kinase C- (cPKC-), and the activity of mammalian target of rapamycin, mTOR. The translational component contributes to 
40–50% of HIF- proteins accumulation after 3 h of 1% O2. Hypoxia also inhibits general protein synthesis and mTOR activity; however, cPKC- inhibitors or rapamycin reduce mTOR activity and total protein synthesis beyond the effects of hypoxia alone. These data show that during general inhibition of protein synthesis by hypoxia, cap-mediated translation of selected mRNAs is induced through the mTOR pathway. We propose that calcium-induced activation of cPKC- hypoxia partially protects an activity of mTOR from hypoxic inhibition. These results provide an important physiologic insight into the mechanism by which hypoxia-stimulated influx of calcium selectively induces the translation of mRNAs necessary for adaptation to hypoxia under conditions repressing general protein synthesis.—Hui, A. S., Bauer, A. L., Striet, J. B., Schnell, P. O., Czyzyk-Krzeska, M. F. Calcium signaling stimulates translation of HIF- during hypoxia.
Key Words: protein kinase C • oxygen-sensitive cells • protein translation • carotid body • PC12 cells
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. O2-sensing carotid body type I cells, respiratory neurons and muscles remain active for long periods despite O2 deprivation, sustaining adaptive hyperventilation. Thus, the expression of proteins necessary for their activation has to be maintained.
HIF is the primary transcription factor induced by low pO2, which in turn regulates the expression of many hypoxia-inducible genes including angiogenic factors, erythropoietin, glucose transporters, glycolytic enzymes, and tyrosine hydroxylase (TH) (2)
. These proteins augment O2 delivery and/or decrease O2 utilization at the cellular and organism levels under conditions of limited O2 availability. HIF is a heterodimer composed of a constitutively expressed ß subunit (HIF-ß) and hypoxia-inducible subunits (HIF-1, -2, -3) (3)
. Accumulation of the subunits during hypoxia results to a large extent from their stabilization due to the inhibition of proline hydroxylase/VHL degradation pathway under hypoxic conditions (2
, 3)
. However, the steady-state levels of the HIF- proteins during hypoxia depend on both the rates of protein degradation and translation. Relatively little is understood about translation of HIF-s during hypoxia. Lang et al. reported that translation of HIF- after 24 h of 1% O2 is maintained at the same level as during normoxia because of the initiation of translation from an internal ribosomal entry site (IRES) in the 5' untranslated region (UTR) (4)
. Such a regulatory mechanism is attractive because the IRES-mediated control of translation is less sensitive to the general inhibition imposed by cellular stressors, such as hypoxia, serum deprivation, heat shock, or apoptosis (5)
.
However, mechanisms of O2 sensing and signaling during hypoxia involve different pathways in various cell types (6)
. One of the crucial primary mechanisms activated in O2-sensitive excitable cells such as carotid body type I cells (1
, 6)
, and the pheochromocytoma-derived PC12 cell line used as a model system of the carotid body (7
, 8)
, is the inhibition of the activity of an O2-sensitive potassium channel (1
, 6
, 9
10
11)
. This leads to depolarization of the cell membrane, activation of the voltage-gated calcium channels and influx of extracellular calcium (6)
. The increase in intracellular calcium entering cells primarily from the external environment was previously demonstrated in both PC12 and carotid body type I cells (6
, 8
, 10
11
12)
. Calcium stimulates the release of specific neurotransmitters that trigger reflex hyperventilation by activating respiratory centers in the brainstem or participate in sustaining the activity of the O2-sensitive cells (6)
. It also stimulates the expression of enzymes synthesizing neurotransmitters, such as tyrosine hydroxylase, the rate limiting enzyme in catecholamine synthesis (7
, 13)
. Tyrosine hydroxylase belongs to the group of HIF target genes (14)
and its transcriptional induction by hypoxia requires an increase in the intracellular calcium (15)
. This study was undertaken to determine if and how calcium regulates HIF- protein expression during hypoxia in O2-sensitive PC12 cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, anti-HIF-2 and polyclonal anti-HIF-1ß antibodies were purchased from Novus Biologicals (Littleton, CO, USA). S6K1, rpS6, and 4E-BP1 antibodies were purchased from Cell Signaling (Beverly, MA, USA). The antibody against PKC- total was purchased from Upstate Cell Signaling Solutions (Charlottesville, VA, USA), and PKC- phosphorylated on Ser657 was purchased from U.S. Biologicals (Swampscott, MA, USA). Except where noted, all other chemicals were purchased from Sigma (St. Louis, MO, USA) or Fisher Scientific (Pittsburgh, PA, USA).
General techniques
Preparation of protein extracts, Western blot analysis, transient transfections and northern blot analysis were performed as described previously (7
, 14
, 16)
. The HRE element used in the transfection experiments consisted of five repeats of HRE element from the promoter of TH gene inserted into the pLuc vector (16)
. For PKC analysis, cells were lysed in buffer containing 10 mM Tris, pH 7.5, 300 mM NaCl, 3 mM MgCl2, 0.5% NP-40, 01% SDS, 0.5% sodium deoxycholate, and standard protease and phosphatase inhibitors. Lysates (100 µg) were subjected to SDS-PAGE in a 3% to 22.5% polyacrylamide gradient gels.
Cell culture and experimental paradigm
PC12 and HEK 293 cells were grown as described earlier (14
, 16)
. Experiments were performed when the cells were 70% to 90% confluent at the time of collection. Where indicated, the medium was changed to DMEM without added CaCl2 (Invitrogen, Carlsbad, CA, USA) prior to experimental treatment. Drugs were applied 30 min before exposure of cells to hypoxia unless otherwise indicated. Cells were exposed to 1% O2 (in 5% CO2 and balanced with N2) in a hypoxic workstation (Coy Laboratory Products, Grass Lake, MI, USA) for 3 h before collection. To avoid reoxygenation of the cells, all hypoxic samples were processed within the workstation until lysis had occurred.
Polysomal fractionation
Cultured cells were treated with 100 µg/mL cycloheximide 5 min prior to harvest. Cell pellets were washed with ice-cold PBS, pH 7.4, containing cycloheximide, then cells were lysed in polysome lysis buffer containing 100 mM KCl, 5 mM MgCl2, 10 mM HEPES, 1% Triton X-100, 0.5% sodium deoxycholate 100 U/mL of RNasin, 100 µg/mL of cycloheximide (Calbiochem), and standard protease inhibitors for 2 min on ice. Nuclei were pelleted by centrifugation at 12,000 x g at 4°C for 5 min. The total RNA in the samples was determined by measuring the optical density at 260 nm (OD260), and equivalent amounts of RNA were loaded onto a 15% to 50% (wt/vol) linear sucrose gradient in polysome lysis buffer containing 500 mM KCl and 20 U/mL RNasin without detergents. Gradients were centrifuged at 181,537 x g (for the average radius of the rotor) for 1 h 45 min at 4°C in a Sorvall SW41Ti rotor (Newtown, CT, USA). Fractions of equal volume were collected from the top using an ISCO fraction collector system (Isco, Inc., Lincoln, NE, USA). RNA in each fraction was extracted using TRI reagent LS (Molecular Research Center); its quality was analyzed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA).
Reverse transcription PCR
Equal fractions of RNA were used for first-strand cDNA synthesis using a Cloned AMV First-Strand cDNA Synthesis Kit (Invitrogen) according to the manufacturer’s protocol. The following forward and reverse primers and annealing temperatures were used for cDNA amplification: for rat HIF-1, 5'ATGTCGCTTTCTTGGAAACGAG3' and 5'TGCCCCAGCAGTCTACATGC3', 57.1°C; for human HIF-1, 5'TGATGTAATGCTCCCCTCACCC3' and 5'GAAGTGGCTTTGGCGTTTCAG3', 64°C; for rat HIF-2, 5'TTCCCAGCCACCATCTACCAG3' and 5'GCCACTCCTGACCCCTTTTG3', 61°C; for human HIF-2, 5'ACAGCAAGAGCAGGTTCCCC3' and 5'GGCAGCAGGTAGGACTCAAATG3', 65°C; for rat TH, 5'CAGGGCTGCTGTCTTCCTACG3' and 5'TGTGTCTGGGTCAAAGGCTCG3', 61°C. The PCR standards for each gene target were dilutions of the purified cDNAs containing from 108 to 104 copies/µL. Standard curves specific to each gene were generated using the logarithm values of the copy number vs. values of the threshold cycle. The concentration of each specific mRNA was extrapolated from the standard curve using linear regression analysis.
RNA interference
Smartpool siRNA against rat or human PKC- were purchased from Dharmacon (Lafaettye, CO, USA). siRNA against a C-terminal region of exon 1 (that is conserved among all isoforms of cPKC from rat and human) (5'CAACCTTCTGCAGTCACTGTA3') was synthesized by Ambion (Austin, TX, USA). Nontarget siRNA with sequence showing no homology to rat or human sequences of cPKC was obtained from Z. Spicer (University of Cincinnati). Cells were seeded at low confluence in the antibiotic free medium 24 h before transfection. Transfections were performed using DramaFECT-1 and 75 nM of siRNA. The efficacy of the siRNA interference was determined by Western blot analysis with an antibody against PKC- 24, 48, and 72 h post-transfection. Inhibition of the PKC- protein was measured at 48 h.
Statistical evaluation of the data
Data was expressed as mean ± standard error of the mean. Analysis was performed with GraphPad Instat version 3.0 for Windows using 1-way analysis of variance (ANOVA), followed by Tukey-Kramer multiple comparison tests. Differences having a P value of <0.05 were considered to be statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
proteins during hypoxia and HIF-2 in PC12 cells, as described previously (14)
. However, when cells were treated with intra- or extracellular calcium chelators [BAPTAAM (Fig. 1
A) or EGTA (Fig. 1B
), respectively (16)
], the hypoxic accumulation of both HIF-s was substantially reduced. Low doses of BAPTAAM induced some accumulation of HIF-1 protein during normoxia (Fig. 1A
, line 5), consistent with the results found in neuroblastoma SH-SY5Y cells, as described by Berchner-Pfannschmidt et al. (17)
. In contrast, accumulation of HIF-1ß was not affected by calcium chelation (Fig. 1)
. The L-type calcium channel is activated by hypoxic depolarization of PC12 cells (12)
; treatment of cells with nifedipine, an inhibitor of L-type calcium channels, substantially attenuated HIF- protein accumulation without affecting HIF-1ß protein levels (Fig. 1C
). Ionomycin, the calcium ionophore, increased the hypoxic accumulation of HIF-1, but not of HIF-2 (Fig. 1D
), by increasing intracellular calcium concentration. Ionomycin did not affect the accumulation of HIF-1, HIF-2, or HIF-ß during normoxia (Fig. 1D
), and neither did depolarization of cells with KCl (data not shown). Ionomycin increased accumulation of HIF-1d but not HIF-2. These data demonstrate that influx of calcium from extracellular environments, through L-type channels, participates in full hypoxic accumulation of HIF- proteins.
|
To determine whether Ca2+ is necessary for HIF- activity, we transiently measured the activity of HRE-driven luciferase reporter construct containing five repeats of HRE element from tyrosine hydroxylase promoter (14)
in PC12 cells treated with EGTA or BAPTAAM (Fig. 1E
). These treatments did not significantly change the levels of luciferase activity during normoxia, but substantially attenuated stimulation of HRE-activity during hypoxia (Fig. 1E
). This demonstrates that the influx of extracellular calcium is necessary for activation of transcription by HIF- proteins.
Calcium is necessary for HIF- translation
To determine the molecular mechanism by which calcium affects the accumulation of HIF-s during hypoxia, we first determined the effects of hypoxia and calcium on HIF- mRNAs accumulation (Fig. 2
A). Hypoxia had no effect on the accumulation of HIF-1 mRNA, but it led to a clear small increase in the steady-state of HIF-2 mRNA. The steady-state of HIF- mRNAs during hypoxia did not differ in their response to EGTA and BAPTAAM treatments (Fig. 2A
). This finding demonstrates that the differences in the steady-state levels of HIF- mRNAs cannot account for the calcium-mediated accumulation in HIF- proteins.
|
Since the primary mechanism leading to HIF accumulation during hypoxia involves stabilization of HIF- proteins, we examined whether the absence of calcium affected their degradation. PC12 cells were pre-exposed to 1% O2 for 3 h to allow for accumulation of HIF- proteins prior to treatment with cycloheximide (with or without BAPTAAM or EGTA) under continued hypoxic conditions (Fig. 2B
). Under these conditions, treatment of cells with BAPTAAM or EGTA failed to induce degradation of HIF- proteins (Fig. 2B
; compare lanes 9–14 with 3–8 in the top panels and lanes 3, 5, 7, 9 with 4, 6, 8, 10 in the bottom panels). Thus, we hypothesized that calcium regulates translation of HIF- mRNAs.
To directly determine if calcium is necessary for HIF- mRNA translation during hypoxia, we investigated the effects of hypoxia and calcium chelators on the distribution of HIF- mRNAs on polysomes (Fig. 3
). We used linear 15% to 50% sucrose gradients containing 500 mM KCl. This very high salt concentration was chosen to prevent any nontranslational interactions of mRNAs with ribosomes. Hypoxia led to a general decrease in translation (Fig. 3A
, insert; compare lanes 1 and 2) and an increase in the number of 80S, 40S, and 60S ribosomes (Fig. 3A
, left). Treatment with BAPTAAM had comparable effects on the polysome profiles in normoxia and hypoxia (Fig. 3A
, right). BAPTAAM led to inhibition, although not abolishment, of protein synthesis during hypoxia (Fig. 3A
, insert).
|
Under normoxic conditions, distribution of HIF-1 and 2 mRNAs on polysomes showed a peak at fraction 9 (Fig. 3B
, left); an indication that under normoxic condition HIF- trasncripts are associated with a translating fraction of polysomes. While there was only one major peak in the distribution of HIF-2 mRNA, a pool of HIF-1 mRNA was also associated with smaller polysomes in fractions 4 and 5. Hypoxia led to a selective increase in the number of ribosomes associated with HIF- transcripts, distributing them to larger polysome fractions, 10 and 11. This effect was more pronounced in the case of HIF-1 mRNA. In contrast, mRNA for TH primarily remained distributed to fractions 8 and 9 of polysome under both normoxia and hypoxia. Treatment with BAPTAAM suppressed the hypoxic shift in distribution of HIF- transcripts to a larger polysome fraction (Fig. 3B
, right), while the effect on distribution of HIF- transcripts was only minimal during normoxia. BAPTAAM had a relatively minor effect on the association of TH mRNA with polysomes. The results were essentially the same for EGTA (data not shown). Thus, under hypoxic conditions an increase in the intracellular concentration of calcium, driven by influx from the extracellular environment, stimulates translation of HIF- proteins.
Regulation of HIF- translation by calcium involves cPKC-
The influx of extracellular calcium in response to hypoxia activates specific signaling pathways involving calcium-dependent kinases or phosphatases. We found that inhibition of the classical, calcium-regulated cPKC- mimicked the effects of calcium chelation on HIF- protein accumulation in PC12 cells during hypoxia. Treatment of cells with a specific cPKC inhibitor, Gö6976, resulted in a substantial attenuation of accumulation of HIF- proteins during hypoxia not seen when inhibitors of protein kinase A (H89), MAPK (PD98059, U0126), or CaM kinase II (KN93 and its inactive form KN92) were used (Fig. 4
A). In addition, Gö6976 inhibited the effect on HIF-mediated trans-activation of an HRE promoter during hypoxia (Fig. 4B
). Similar to calcium chelators, cPKC inhibition did not affect the rate of HIF-1 or -2 protein degradation in cells pre-exposed to hypoxia (Fig. 4C
). To confirm results obtained from pharmacologic studies using molecular approaches, we performed siRNA-mediated knockdowns of cPKC using either a commercially available cocktail of four siRNAs specific for cPKC-, or a single custom-designed siRNA against a region within the first exon that is highly conserved among all cPKCs in humans and rodents. Nonspecific siRNAs were used in control experiments. Both types of siRNA decreased the expression of cPKC- by 80% to 90% 48 h after transfection. It was significant that both inhibited hypoxic accumulation of HIF-1 and -2 by 50% (Fig. 4D
). These data are consistent with the results using Gö6976 (Fig. 4A
). The fact that siRNA specific for cPKC- and siRNA that target all isoforms of cPKC have the same effect on the HIF- accumulation supports the conclusion that cPKC- is the isoform directly involved in regulation of HIF-.
|
Next, we directly evaluated the effects of cPKC- inhibitor on HIF- mRNAs translation by analyzing distribution of HIF- mRNAs on polysomes. We found that Gö6976 reversed the effects of hypoxia. HIF-1 and -2 mRNAs were redistributed to smaller polysomes in a manner similar to their distribution during normoxia (i.e., to fraction 9; Fig. 5
A). In contrast, the compound had no effects on the polysomal distribution of HIF- transcripts during normoxia (data not shown). This indicates that inhibition of cPKC- inhibits hypoxia-stimulated translation of HIF- mRNAs. Because HIF-1 mRNA translation had previously been proposed to be initiated during hypoxia from IRES (4)
we attempted to measure HIF-1 IRES-mediated translation using a bicistronic construct containing a fragment of the HIF-1 5' UTR located between renilla and firefly luciferase as described by Lang et al. (4)
. Using this construct, we failed to detect any IRES-mediated translation activity (data not shown). Instead, we hypothesized that translation of HIF- mRNAs during hypoxia may be regulated in a cap-mediated manner. We used rapamycin, a specific inhibitor of cap-mediated translation and of mTOR (mammalian target of rapamycin) pathway crucial for translational regulation (18
, 19)
. We found that the hypoxia-dependent distribution of HIF-1 and HIF -2 mRNAs on larger polysomes was sensitive to rapamycin (Fig. 5A
). Rapamycin resulted in the redistribution of the pool of HIF- mRNAs to a smaller polysome fraction (fraction 3-5), or the soluble fraction 1. The observed effects were selective for HIF- transcripts. Hypoxia, cPKC- inhibition, or rapamycin had minimal effects on the distribution of TH mRNA on polysomes; TH mRNA remained associated with fractions 8 and 9. We further determined that rapamycin and Gö6976 each caused an 40% to 50% reduction in the accumulation of the HIF- proteins, as measured by Western blot (Fig. 5B
). The use of both drugs in the same experiment did not cause further decrease in HIF-1 protein levels, and only caused a small additional decrease in HIF-2 protein (Fig. 5B
). Thus, the effects of rapamycin and Gö6976 on accumulation of HIF- proteins converge on the same major pathway. When either Gö6976 or rapamycin were applied together with hypoxia, they inhibited protein synthesis in PC12 cells to a greater extent than did hypoxia alone (Fig. 5
, insert). This demonstrates that under hypoxic conditions, in spite of the general inhibition of mTOR pathway (20
21
22
23)
, some activity of the mTOR pathway remains functional in supporting translation of selective proteins in a rapamycin-sensitive manner. These data indicate that translation of HIF- mRNAs is mediated during hypoxia in a cap-dependent manner.
|
To determine whether the effects of rapamycin on HIF- mRNAs translation can result from the effects of rapamycin on cPKC- activity or accumulation, we measured phosphorylation of cPKC- on Ser657, a marker of PKC- activation (24)
(Fig. 5C
). Hypoxia induced Ser657 phosphorylation of cPKC-, while Gö6976 and EGTA inhibited cPKC- phosphorylation during hypoxia without affecting its total accumulation (Fig. 5C
). In contrast, rapamycin had no effect on either cPKC- total accumulation or Ser657 phosphorylation. These data show that rapamycin does not inhibit translation of HIF- mRNAs by inhibiting the activity of cPKC-.
cPKC- participates in maintaining the activity of mTOR pathway during hypoxia
To begin investigating cross-talk between cPKC- and mTOR pathways, we determined the effects of Gö6976 and rapamycin on the activity of mTOR measured as the mTOR-dependent phosphorylation of its two targets, S6 kinase, S6K1, and translation initiation factor eIF4E binding protein, 4E-BP1, in PC12 cells during normoxia and hypoxia. Phosphorylation of S6K results in phosphorylation of ribosomal S6 protein (rpS6), while phosphorylation of 4E-BP1 prevents its binding and inhibition of eIF4E translation factor (18)
. Phosphorylation of both is needed to maintain active translation (18)
. Hypoxia inhibited mTOR activity as measured by decrease in phosphorylation of S6K and 4E-BP (Fig. 5D
; 19, 20–22). Both Gö6976 and rapamycin inhibited the mTOR-mediated phosphorylation of S6K1 on Thr389; of rpS6 on Ser235/236 and Ser240/244; and of 4E-BP1 on Thr37 and Ser65 above and beyond the inhibition of these phosphorylation events caused by hypoxia alone (Fig. 5D
). While rapamycin was effective under both normoxic and hypoxic conditions, Gö6976 inhibited the phosphorylation only during hypoxia, when intracellular calcium increased, but not during normoxia (Fig. 5D
, compare lanes 2 and 5). The effects of hypoxia on 4E-BP1 phosphorylation were substantially less dramatic than those on S6K1 and rpS6. This indicates that despite hypoxic inhibition of mTOR, an activity of mTOR pathway remained active in a manner dependent on the hypoxia-induced activity of cPKC-. The data show that mTOR activity toward 4E-BP is less sensitive to hypoxic inhibition, than mTOR activity toward the S6K1. These findings imply that hypoxic stimulation of HIF- cap-dependent translation might involve regulation of eIF4E-4E-BP interaction and availability of free eIF4E.
Hypoxia stimulates translation of HIF- in a manner dependent on cPKC- and mTOR pathway in HEK 293 cells
The effects of hypoxia on HIF- translation were not unique to PC12 cells, but also occurred in the HEK293 cell line (Fig. 6
). Similar to PC12 cells, hypoxia caused distribution of HIF-1 and -2 mRNAs to larger polysome fractions 9, 10, and 11 (Fig. 6A
). In contrast to PC12 cells, the constitutive distribution of HIF-1 mRNA during normoxia showed a major peak in fraction 8 whereas HIF-2 mRNA had two peaks: one associated with fraction 5 and another with fraction 7. Hypoxia caused a larger shift in the case of HIF-2 mRNA. Inhibition of cPKC- or mTOR using Gö6976 or rapamycin, respectively, substantially inhibited hypoxic accumulation of HIF-1 and -2 during hypoxia (Fig. 6B
). RNA interference with siRNA against cPKC- showed a comparable reduction in accumulation of both HIF- proteins during hypoxia (Fig. 6C
). These data indicate that stimulation of HIF- mRNAs translation by hypoxia in a manner dependent on cPKC- and the mTOR pathway is a more general phenomenon, not limited to PC12 cells.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
subunits of HIF is regulated by an influx of extracellular calcium, activation of classical PKC-, and of the mTOR pathway. We measured translation of HIF-s directly, by determining association of the transcripts with individual polysomal fractions. This translational stimulation is clearly necessary for the accumulation of HIF- proteins during hypoxia, because the protein levels of HIF-1 and -2 are substantially attenuated when either calcium influx and cPKC- are inhibited or when mTOR is inhibited with rapamycin beyond its inhibition by hypoxia. The effects of hypoxia, calcium, and cPKC- inhibitor on distribution of HIF- transcripts on polysomes appear to be selective for HIF- mRNAs. Polysome distribution of mRNA for TH, another hypoxia-induced gene, was only minimally affected by any of these treatments. The effect of hypoxia on HIF- mRNAs translation appears to be fairly general, as it was measured in rat PC12 and human HEK293 cells. In that respect, HEK293 cells have potassium (25
26
27
28)
and L-type like calcium (29)
channels, and recent unexpected evidence indicates that they have some neuronal features as well (30)
.
Results were essentially similar for both HIF-1 and -2. However, some intriguing differences were observed. For example, constitutive distribution of HIF-1 mRNA on polysomes in nonhypoxic PC12 cells had two peaks, with the smaller bound to 2 or 3 ribosomes, and the larger bound to 8 or 9 ribosomes. In contrast, most of the HIF-2 mRNA was constitutively distributed to larger polysome fractions. This coincided with a stronger hypoxic effect on distribution of HIF-1 mRNA, as compared with HIF-2 mRNA. A different result was measured in HEK293 cells. In those cells, HIF-2 mRNA had a bimodal distribution on polysomes in normoxia and it was shifted to larger polysome fraction much stronger than in the case of HIF-1 mRNA. Although the mechanisms causing these differences are not clear, they may correlate with relative tissue-specific abundance and functional variability of both HIF-s.
An important conclusion from this study is that there is a clear selectivity of the effects of hypoxia on protein translation (Fig. 7
). Hypoxia inhibits general protein synthesis and activity of the mTOR pathway. However, this inhibition is incomplete, and translation of selected mRNAs, such as HIF- mRNAs, is induced in the cap-dependent manner through the mTOR pathway. The signaling and biochemical mechanisms of such regulation remain to be elucidated. Regulatory role of mTOR was originally discovered for translation of mRNAs with the specific polypyrimidine tracks in the 5'UTR or 5'TOP sequences (19
, 31)
, but HIF-s do not contain classic 5'TOPs. Hypoxic inhibition of the mTOR activity is not uniform; it results in a profound decrease in S6K1 activity, but a less pronounced dephosphorylation of 4E-BP. This suggests that at the indicated level and duration of hypoxia in PC12 cells, the 4E-BP component of mTOR pathway is more protected from hypoxic inhibition. As a result, sequestration of eIF4E translation initiation factor by dephosphorylated 4E-BP would be potentially mild, allowing a pool of eIF4E to participate in initiation of cap-dependent translation under hypoxic conditions (Fig. 7
, left). These results also suggest that hypoxia may actually increase the pool of free eIF4E, due to the repression of S6K and inhibition of translation of those mRNAs, which require S6K activity for ribosome recruitment (Fig. 7
, left). This repression may free translation factors such as eIF4E, and, paradoxically, stimulate translation of other mRNAs, such as HIF- transcripts (Fig. 7
, right); as was previously suggested with Cdc25A phosphatase (32
, 33)
. eIF4E translation factor is indispensable for cap-mediated translation, but is present in several cell lines at limited concentrations (34
, 35)
. Transcripts with complex 5'UTRs must compete for eIF4E availability and their translation depends on the amount of free eIF4E (34
, 35)
. It is possible that HIF- translation is sensitive to the levels of free eIF4E.
|
Understanding the cross-talk mechanism between the cPKC- and mTOR pathways during hypoxia requires further investigation. Hypoxic inhibition of mTOR activity appears to require the presence of the intact tuberous sclerosis TSC1/TSC2 tumor suppressor complex (23)
. Hypoxia stimulates classical PKC- activity through an increase in intracellular calcium (36)
. We propose that cPKC- directly or indirectly phosphorylates a fraction of TSC2, thereby preventing TSC-dependent inhibition of mTOR-mediated translation of certain mRNAs, such as HIF-s (Fig. 7)
. Our hypothesis is supported by published evidence that stimulation of PKC with phorbol esters activates mTOR activity as measured by phosphorylation of mTOR substrates, S6K1 (37
38
39)
and 4E-BP (39
, 40)
, in a manner independent from, but analogous to, the classic PI3K/AKT pathway (41
, 42)
.
There is previous evidence regarding involvement of the mTOR pathway in the regulation of HIF-1 translation; however this evidence comes from experiments conducted under normoxic conditions. Activation of the tyrosine receptor HER2 pathway induces HIF- protein synthesis under nonhypoxic conditions through stimulation of PI3K/AKT-mTOR, a mechanism important to the progression of breast cancer (43)
. A similar pathway is likely to be involved in the translational induction of HIF-1 in normoxia by vasoactive hormones, such as angiotensin (44)
. This effect appears to be mediated through the 5' UTR of HIF-1. mTOR pathway was also implicated in the degradation of HIF- proteins (45)
.
The biologic effects of calcium and PKC- on HIF- are significant. Clearly, calcium is necessary for transcriptional induction of TH (15)
, a gene stimulated during hypoxia by HIF (14)
and AP1 site (46)
. Our work further extends these data showing that indeed calcium and PKC- are necessary for the activation of TH HRE during hypoxia. Others have reported that calcium is necessary for HIF activity acting through an ERK-mediated pathway (47)
. Another important biological implication is that activation of classical PKC has neuroprotective effects against a number of stressors, including ischemia (48
49
50)
. It is possible that activation of PKC- through stimulation of HIF- translation participates in that process.
Our data are potentially controversial with two previously published studies. First, calcium chelation, particularly by BAPTAAM, stimulates HIF- accumulation by inhibiting prolyl-hydroxylase activity and causing stabilization of HIF (17)
. In our work, we observed that both EGTA and BAPTAAM had some stabilizing effects on HIF- proteins; however, this stabilization was minor and did not measure any trans-activation of HRE-reporter. These differences potentially result from different pools of calcium that are activated in different cell lines during hypoxia. Second, previous investigations concluded that translation of HIF-1 during hypoxia was initiated from an IRES, and there were no changes in the association of HIF- mRNAs with polysomal fractions relative to nonhypoxic conditions (4)
. However, the experiments in those studies were performed differently from our experiments. First, the authors did not analyze HIF- mRNA in individual polysomal fractions, but rather in the total pool of translated and nontranslated mRNA. Thus, any specific differences among individual fractions could not be detected. Second, the IRES-mediated translation was identified in cells exposed to 1% O2 for 24 h. This is more severe than the 3 h of 1% O2 used in our experiments. Therefore, it is possible that cap-mediated translation of HIF- mRNAs plays a role during short-lasting hypoxia, while an IRES is used during long-lasting O2 deprivation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication September 16, 2005. Accepted for publication October 25, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  R. M. Young, S.-J. Wang, J. D. Gordan, X. Ji, S. A. Liebhaber, and M. C. Simon Hypoxia-mediated Selective mRNA Translation by an Internal Ribosome Entry Site-independent Mechanism J. Biol. Chem., June 13, 2008; 283(24): 16309 - 16319. [Abstract] [Full Text] [PDF]
|
|
