Microtubule disruption targets HIF-1alpha mRNA to cytoplasmic P-bodies for translational repression

Abstract

The hypoxia inducible factor 1α (HIF-1α) is overexpressed in solid tumors, driving tumor angiogenesis and survival. However, the mechanisms regulating HIF-1α expression in solid tumors are not fully understood. In this study, we find that microtubule integrity and dynamics are intricately involved in orchestrating HIF-1α translation. HIF-1α messenger RNA (mRNA) traffics on dynamic microtubules when it is actively translated. Microtubule perturbation by taxol (TX) and other microtubule-targeting drugs stalls HIF-1α mRNA transport and releases it from polysomes, suppressing its translation. Immunoprecipitation of the P-body component Argonaute 2 (Ago2) after microtubule disruption shows significant enrichment of HIF-1α mRNAs and HIF-targeting microRNAs (miRNAs). Inhibition of HIF-repressing miRNAs or Ago2 knockdown abrogates TX's ability to suppress HIF-1α translation. Interestingly, microtubule repolymerization after nocodazole washout allows HIF-1α mRNA to reenter active translation, suggesting that microtubule dynamics exert tight yet reversible control over HIF-1α translation. Collectively, we provide evidence for a new mechanism of microtubule-dependent HIF-1α translation with important implications for cell biology.

Figure 1.

MTDs inhibit HIF-1α translation. (A) Polysome association profile of MCF7 cells after overnight treatment with 25 nM TX, 25 nM vinblastine (VBL), or 25 µM 2ME2 visualized after sucrose gradient centrifugation. The 40, 60, and 80S ribosomal subunits and polysomes were fractionated and monitored with continuous A₂₅₄ measurements. Representative profiles and microtubule images after each drug treatment are shown. Bar, 10 µm. (B) RNA was extracted from each fraction, and HIF-1α, GAPDH, and p53 expression was quantified by qRT-PCR. The distribution of each mRNA between nontranslating (1_6) and translating (7_12) fractions is plotted (mean ± SEM; n = 3–10; ***, P < 0.001). (C) Immunoblot of HIF-1α and actin from MCF7 cells treated as in A and exposed to normoxia (N) or 4-h hypoxia to visualize HIF-1α protein. Black lines indicate that intervening lanes have been spliced out.

Figure 2.

Microtubule disruption is required for drug-induced inhibition of HIF-1α translation. (A and B) Parental 1A9 (A) and 1A9/PTX10 (B) cells treated with 10 nM TX overnight and subjected to sucrose gradient fractionation. The percentage of HIF-1α or GAPDH mRNA/fraction is plotted for each cell line and drug treatment. (bottom) HIF-1α and GAPDH protein levels in 1A9 (A) and 1A9/PTX10 (B) cells treated with 10 nM TX overnight assessed by immunoblotting after 4-h hypoxia. (C and D) Similar experiments performed using 1A9 (C) or 1A9/2MRC (D) cells to assess the HIF-1α polysome profile after 10 µM 2ME2 treatment. (bottom) Immunoblot of HIF-1α and GAPDH after 4-h hypoxia (Table I).

Figure 3.

HIF-1α mRNA associates with cellular microtubules. (A) HIF-1α MBs incubated at 37°C for 1 h with a complementary (C) or sense (S) DNA ODN or total RNA from MCF7 cells. Relative fluorescence units (RFU) are represented as mean ± SEM. (B) MCF7–GFP-tub (white) cells fixed and processed for MB (red) hybridization. (zoom) Higher magnification views of the boxed area are shown. Two representative untreated cells are shown. (C) MCF7–GFP-tub (white) cells transfected with 100 nM HIF-1α MB (red) overnight, treated with 100 µM 2ME2 or 100 nM TX (1 h) and imaged continuously for 60 s. Arrows show inhibition of HIF-1α movement. (max) Stack arithmetic shows trajectory of MB movement. (max-zoom) Higher magnification view of max from boxed areas in the first column (Videos 1–4). (D) Anti-GFP was used to immunoprecipitate lysates from MCF7–GFP-tub cells treated overnight with 25 µM 2ME2 or 25 nM TX. HIF-1α, p53, and GAPDH mRNA expression are shown as mean fold change ± SEM (n = 3). Bars, 10 µm.

Figure 4.

MTDs induce P-body formation. (A) Live cell imaging of HeLa–GFP-Ago2 cells treated with 100 µM 2ME2, 1 µM TX, or 30 µM Noc for 1 h. (B) HeLa–GFP-Ago2 (green) cells treated with 25 µM 2ME2 or 50 nM TX overnight, fixed, and stained for GE-1 (red) and α-tubulin (white). Arrows show colocalization of Ago2 and GE-1. (C) The number of Ago2 foci per cell was quantified (100 cells/condition) and plotted as percentage of cells with either three or less or more than three P-bodies (mean ± SEM; n = 3; **, P < 0.01). (D) Immunoblot of HeLa–GFP-Ago2 cells treated as in B and followed by 4-h hypoxia. Bars, 10 µm.

Figure 5.

HIF-1α mRNA is sequestered to P-bodies upon microtubule disruption. (A) HeLa–GFP-Ago2 cells treated with 25 µM 2ME2 or 25 nM TX overnight, lysed, and immunoprecipitated (IP) using an anti-GFP antibody. Ago2 protein was detected by immunoblotting. Ago2-bound RNA was extracted, and HIF-1α, p53, and GAPDH expression is displayed as mean fold change ± SEM (n = 3; *, P < 0.05). (B) RNA from Ago2 unbound lysates was processed as in A (n = 3; *, P < 0.05). (C) Bar graph showing the percentage of HIF-1α mRNA localized to P-bodies after the indicated drug treatments. (D) HeLa–GFP-Ago2 cells transfected with 100 nM HIF-1α MB (red) and treated with 100 nM TX or 100 µM 2ME2 for 1 h. Bar, 10 µm. Right column shows the percentage of overlap (mean ± SEM; n = 3) between Ago2 foci and HIF-1α MB (***, P < 0.001; Videos 6 and 7).

Figure 6.

HIF-1α translation inhibition is reversed after microtubule repolymerization. (A) HeLa–GFP-Ago2 (green) cells treated with 10 µM Noc for 6 h followed by drug washout (W/O). Cells were fixed at the indicated times and immunostained for α-tubulin (white). The percentage of cells with either three or less or more than three P-bodies is plotted (mean ± SEM; n = 10 fields of view/condition). (B) HIF-1α and actin protein levels in HeLa–GFP-Ago2 cells treated with 1 µM Noc overnight and lysed after 6-h washout (6) or continued drug treatment (−) with 4-h hypoxia. (C) The percentage of HIF-1α mRNA associated with untranslating fractions (1_6) or actively translating polysomes (7_12) after 1 µM Noc treatment overnight and 6 h drug washout is plotted (mean ± SEM; n = 3; *, P < 0.05). (D) HeLa–GFP-Ago2 cells were incubated in glucose-free media for 1 h and stained for α-tubulin (white). The percentage of cells with three or less or more than three P-bodies (mean ± SEM; n = 3; *, P < 0.05) is plotted. (E) Immunoblot of HIF-1α and actin after 1-h glucose starvation and 4-h hypoxia. (F) Ago2-bound RNA extracted from control (Ctrl), glucose-starved (1 h) cells, or cells treated overnight with 1 µM Noc (Noc) and followed by 6-h washout. HIF-1α expression was quantified by qRT-PCR (mean ± SEM; n = 2). Bars, 10 µm.

Figure 7.

Role of HIF-1α–targeting miRNAs in the microtubule-dependent translation of HIF-1α. (A) HIF-1α immunoblot of MCF7 cells transfected with each pre-miR, a combination of all four, or a Cy3-labeled negative control (Cy3 Ctrl) and treated with 10 µM proteasome inhibitor MG-132 for 4 h. (B) HIF-1α mRNA expression quantified by qRT-PCR after transfection of all HIF-targeting miRNAs (miRs) in combination (mean ± SEM; n = 2). (C) Relative expression of Ago2-bound miRNAs quantified by qRT-PCR from HeLa–GFP-Ago2 cells treated overnight with 25 nM TX, glucose starvation (left; mean ± SEM; n = 3), or 1 µM Noc followed by a 6-h drug washout (right; mean ± SEM; n = 2, *, P < 0.05; **, P < 0.01). (D) Endogenous miRNA expression was assessed by qRT-PCR in HeLa cells after overnight treatment with 25 nM TX or 25 µM 2ME2.

Figure 8.

TX requires the 3′ UTR of HIF-1α to regulate its translation. (A) HeLa–GFP-Ago2 cells transfected (3:1 or 3:2 Fugene/DNA ratio) with a GFP-tagged HIF-1α construct lacking both the 5′ and 3′ UTR (HIF-1α–GFP) and treated with 25 nM TX overnight. (B) HeLa–GFP-Ago2 cells transfected with a HIF-1α 3′ UTR luciferase expression vector either alone or cotransfected with four HIF-targeting miRNAs (miR), four HIF-specific anti-miRs (anti), or a scrambled miRNA-like sequence (Scram). Luciferase activity was measured in untreated or TX-treated cells as indicated. *, P < 0.05. Error bars indicate mean ± SEM. (C) HIF-1α immunoblot of HeLa–GFP-Ago2 cells transfected with HIF-specific anti-miRs (622, 338, 411, and 519b), a combination of all four, or Cy3 control for 24 h followed by overnight 25 nM TX treatment. Transfection of two non-HIF–related anti-miRs (let7a and 545) was used as a control.

Figure 9.

Ago2 is necessary for MTD-induced inhibition of HIF-1α translation. (A) Ago2 mRNA expression by qRT-PCR after stable expression of Ago2 (shRNA; mean ± SEM; n = 2; ***, P < 0.001) or control shRNA (Neg Ctrl). (B) Immunoblot of HeLa cells transfected as in A and treated with 25 nM TX overnight. (right) A more intense scan of lanes 5 and 6 is depicted as a result of sample underloading. (C) Ago2-GFP is visualized in fixed untreated or TX-treated HeLa–GFP-Ago2 cells transfected with Ago2 siRNA (top) and stained for GE-1 (bottom). Bar, 10 µm.