Expression of recombinant proteins

Escherichia coli BL21 (DE3) strain bacteria were transformed with plasmids, then expression was initiated by the addition of 0.2 mM isopropyl β-d-1-thiogalactopyranoside when the culture density reached 0.6 (OD₆₀₀). Proteins were expressed at 30°C for 3 h for recombinant RNA helicases or at 20°C overnight for recombinant PHAX and ALY proteins.

Purification of recombinant proteins

To purify glutathione S-transferase (GST)-PHAX (50), cells expressing GST-PHAX were harvested and lysed in a French press in buffer-500 (20 mM Tris–HCl [pH 8.0], 0.5 M NaCl, 1 mM DTT, 10% glycerol and 0.1 mM EDTA) containing 0.1% Nonidet P-40 and a proteinase inhibitor. The supernatant after centrifugation was bound to Glutathione Sepharose beads (Cytiva). Bound beads were washed five times with buffer-500, and bound proteins were subsequently eluted in buffer-500 containing 10 mM reduced glutathione. The eluate was dialyzed against buffer-100 (20 mM Tris–HCl [pH 8.0], 0.1 M NaCl, 1 mM DTT, 10% glycerol and 0.1 mM EDTA) and was then applied to HiTrap Heparin Sepharose (Cytiva). The bound protein was eluted with a linear gradient of NaCl (0.1–1 M). The eluate was dialyzed against buffer-100.

To purify FLAG-UAP56 wild-type (WT) protein, FLAG-UAP56 K95E mutant (51), FLAG-URH49 and FLAG-DBP5, cells expressing GST-FLAG-UAP56 WT, GST-FLAG-UAP56 K95E, GST-FLAG-URH49 and GST-FLAG-DBP5 were harvested and lysed in a French press in buffer-500 containing 1 mM EDTA, 0.1% Nonidet P-40, and a proteinase inhibitor. The supernatant after centrifugation was bound to Glutathione Sepharose beads. Bound beads were washed five times with buffer-500 containing 1 mM EDTA and twice with buffer-100. Bound proteins were subsequently treated with 2 U/μl of PreScission Protease (Cytiva) at 4°C overnight. The supernatant was applied to HiTrap Q Sepharose (Cytiva). The bound protein was eluted with a linear gradient of NaCl (0.1–1 M). The eluate was dialyzed against buffer-100.

To purify GST-ALY (51), cells expressing GST-ALY were harvested and lysed in a French press in buffer-500 containing 0.1% Nonidet P-40 and a proteinase inhibitor. The supernatant after centrifugation was bound to Glutathione Sepharose beads. Bound beads were washed five times with buffer-500, and bound proteins were subsequently eluted in buffer-500 containing 10 mM reduced glutathione. The eluate was dialyzed against buffer-500.

To purify FLAG-ALY, cells expressing GST-FLAG-ALY were harvested and lysed in a French press in buffer-500 containing 0.1% Nonidet P-40 and a proteinase inhibitor. The supernatant after centrifugation was bound to Glutathione Sepharose beads. Bound beads were washed five times with buffer-500. Bound proteins were subsequently treated with PreScission Protease at 4°C overnight. The supernatant was diluted with ten volumes of buffer-100 and was applied to HiTrap Heparin Sepharose. The bound protein was eluted with a linear gradient of NaCl (0.1–1 M). The eluate was dialyzed against buffer-500.

To purify His-T7-PHAX, cells expressing His-T7-PHAX were harvested and lysed in a French press in buffer-500b (20 mM Tris–HCl [pH 8.0], 0.5 M NaCl, 1 mM 2-mercaptoethanol, 10 mM Imidazole and 10% glycerol) containing 0.1% Nonidet P-40 and a proteinase inhibitor. The supernatant after centrifugation was bound to Ni Sepharose beads (Cytiva). Bound beads were washed five times with buffer-500b, and bound proteins were subsequently eluted in buffer-500b containing 0.5 M imidazole. The eluate was dialyzed against buffer-100 and was then applied to HiTrap Heparin Sepharose. The bound protein was eluted with a linear gradient of NaCl (0.1−1 M). The eluate was dialyzed against buffer-100. Coomassie staining of purified recombinant proteins is shown in Supplementary Figure S1.

In vitro transcription

³²P-labeled RNAs were transcribed in a 10 μl volume containing 20 U T7 RNA Polymerase (Promega), Transcription Buffer (Promega), 1 mM DTT (Promega), 12 U RNasin Plus (Promega), NTP mixture (0.5 mM ATP, 0.5 mM CTP, 0.1 mM UTP and 0.1 mM GTP), 1 μg DNA template, 1 mM m⁷G(5′)ppp(5′)G RNA Cap Structure Analog (New England Biolabs) and 0.74 MBq of [α-³²P] UTP (PerkinElmer). Cap Structure Analog was not added for U6Δss snRNA or tRNA^Phe. After incubation for 60 min at 37°C, RNA was recovered from the supernatants by phenol/chloroform extraction and purified using G-50 micro columns (Cytiva). RNA was then precipitated with ethanol and dissolved in H₂O.

In vitro RNA-protein binding assay

For RNA co-immunoprecipitation (IP), ³²P-labeled RNAs were mixed with 0.5 mM ATP, 20 mM creatine phosphate, 2 mM MgCl₂ and 30% HeLa nuclear extracts (HNEs; Cil Biotech) and incubated at 30°C for 20 min. After centrifugation, the reaction mixture was then rotated at 4°C for 1–3 h with Protein A Sepharose beads (Cytiva) that were pre-bound to an anti-PHAX antibody (53) or anti-mouse IgG antibody (Merck). The beads were then washed five times with RSB100N buffer (10 mM Tris–HCl [pH 7.5], 0.1 M NaCl, 2.5 mM MgCl₂ and 0.1% Nonidet P-40) and incubated in Homomix (50 mM Tris–HCl [pH 7.5], 5 mM EDTA, 1.5% SDS, 0.3 M NaCl, and 1.5 mg/ml proteinase K; Nacalai Tesque) at 50°C for 30 min. RNA was recovered from the supernatant by phenol/chloroform extraction and ethanol precipitation, then analyzed by denaturing polyacrylamide gel electrophoresis (PAGE) and autoradiography.

For GST pull-down, ³²P-labeled RNAs were mixed with 0.5 mM ATP, 20 mM creatine phosphate, 2 mM MgCl₂, 30% HNEs and either purified recombinant 50 nM GST or GST-PHAX, and the mixture was incubated at 30°C for 20 min. After centrifugation, the reaction mixture was then rotated at 4°C for 1–3 h with Glutathione Sepharose beads equilibrated with RSB100N buffer. The beads were then washed five times with RSB100N buffer and incubated in Homomix at 50°C for 30 min. RNA was recovered from the supernatant by phenol/chloroform extraction and ethanol precipitation. The samples were then analyzed by 8% denaturing PAGE and autoradiography.

To prepare nuclear lysates from Xenopus oocytes, 500 nuclei were dissected from Xenopus oocytes with sharp forceps and collected in a 1 ml volume containing 20 mM Hepes–NaOH [pH 7.9], 0.1 M KCl, 1 mM DTT, 0.2 mM EDTA and 1 mM PMSF. Nuclei were passed through a 27‐gauge needle 5 times and lysates were concentrated to 100 μl using >10 kDa Amicon Ultra (Merck). For GST pull-down, nuclear lysates from Xenopus oocytes were used instead of HNEs and the incubation was done at 19°C.

For RNA binding of UAP56, ³²P-labeled RNAs were mixed with or without 2 mM nucleotide (ATP, ADP, ATP-γS), 2 mM MgCl₂ and 2 μM FLAG-UAP56, and incubated at 30°C for 20 min. The reaction mixture was then rotated at 4°C for 1–3 h with Dynabeads Protein A (Thermo Fisher Scientific) that were pre-bound to an anti-FLAG M2 antibody (Merck). The beads were then washed five times with RSB100N and incubated in Homomix at 50°C for 30 min. RNA was recovered from the supernatant by phenol/chloroform extraction and ethanol precipitation, then analyzed by denaturing PAGE and autoradiography.

Searching for factors that bind PHAX and ATP

The mixture was incubated in a 250 μl volume containing 200 μl HNEs, 5 μg GST-PHAX, 50 μg RNase A, 0.5 mM ATP, 20 mM creatine phosphate, and 2 mM MgCl₂ at 30°C for 20 min. After centrifugation, the reaction mixture was added to Glutathione Sepharose beads equilibrated with RSB100N buffer and incubated on a rotating platform at 4°C for 1 h. Beads were then washed five times with buffer-150 (50 mM Tris–HCl [pH 7.5], 150 mM NaCl, 1 mM DTT, 0.2 mM EDTA, 10% glycerol). Bound proteins were subsequently treated with PreScission Protease (Cytiva) at 4°C overnight, and the supernatant was dialyzed against buffer-100. The mixture was incubated in a 20 μl volume containing the 14 μl dialysate, 0.37 MBq of [α-³²P] ATP (PerkinElmer), and 2 mM MgCl₂ at 30°C for 15 min. The reaction mixture was irradiated with 254 nm ultraviolet (UV) light (FUNA-UV-LINKER FS-800; Funakoshi) at 200 mJ/cm² on ice. The irradiated sample was analyzed by SDS-PAGE and autoradiography. To identify the cross-linked protein, IP was performed followed by SDS-PAGE and autoradiography.

Western blotting

Primary antibodies were used in TBS-T buffer (20 mM Tris–HCl [pH 7.5], 150 mM NaCl and 0.1% Tween 20) containing 5% skim milk. Incubations were generally performed at 4°C overnight. For secondary antibodies, horseradish peroxidase (HRP)-labeled anti-mouse, anti-rabbit, or anti-goat antibodies (Jackson ImmunoResearch) were used in TBS-T buffer containing 5% skim milk. Incubations were conducted as recommended by the manufacturers. The antibodies used in this study are described in Supplementary Table S1.

In vitro protein–protein binding assays

Purified recombinant proteins were mixed with 2 mM ATP, 2 mM MgCl₂, 1 mg/ml BSA and 0.5 mg/ml RNase A. In the presence of HNEs, purified recombinant proteins and HNEs were mixed with 0.5 mM ATP, 20 mM creatine phosphate, 2 mM MgCl₂ and 0.5 mg/ml RNase A. The mixture was incubated at 30°C for 20 min. After centrifugation, the reaction mixture was added to Glutathione Sepharose beads equilibrated with RSB100N buffer and incubated on a rotating platform at 4°C for 1 hour. The beads were then washed five times with RSB100N buffer, then the bound material was recovered and analyzed by SDS-PAGE and western blotting.

Identification of bridging factors between PHAX and UAP56

The mixture was incubated in a 250 μl volume containing 200 μl HNEs, 5 μg GST-PHAX, 5 μg FLAG-UAP56, 50 μg RNase A, 0.5 mM ATP, 20 mM creatine phosphate, and 2 mM MgCl₂ at 30°C for 20 min. After centrifugation, the reaction mixture was added to an anti-FLAG M2 affinity gel (Merck) equilibrated with RSB100N buffer and incubated on a rotating platform at 4°C for 1 h. The beads were then washed five times with RSB100N buffer and bound proteins were subsequently eluted with 3xFLAG peptide. The eluate was applied to Glutathione Sepharose. The bound proteins were treated with HRV 3C protease, and the supernatant was fractionated by SDS-PAGE followed by Coomassie brilliant blue (CBB) staining. Mass spectrometry analyses of cut-out bands were entrusted to Japan Proteomics.

Xenopus oocyte microinjection

Xenopus oocyte microinjections were performed as previously described (54). Briefly, ³²P-labeled RNAs were injected with antibodies (200 ng IgG/oocyte) into the nucleus. After incubation at 19°C, oocytes were dissected into nuclear and cytoplasmic fractions with sharp forceps. These fractions were incubated in Homomix buffer at 50°C for 30 min. RNA was recovered from the supernatants by phenol/chloroform extraction and ethanol precipitation, then analyzed by 8% denaturing PAGE and autoradiography.

Identification of PHAX-interacting proteins as potential RNA helicases involved in PHAX loading

One of the biochemical features of RNA helicases is that they can be cross-linked to ATP by ultraviolet (UV) irradiation. For example, when purified recombinant wild-type (WT) UAP56 protein or the ATP-binding deficient mutant protein (K95E) was irradiated with UV light in the presence of ³²P-labeled ATP, the WT protein, but not K95E, was labeled with ³²P (Supplementary Figure S3) (51). If RNA helicase loaded PHAX onto RNA, then the helicase should interact with PHAX. Taking advantage of these two features, we developed an in vitro strategy to identify RNA helicases involved in PHAX loading (Figure ​(Figure2A).2A). Purified recombinant GST-PHAX was incubated with HNEs, ATP and RNase A, followed by GST pull-down. The pulled down proteins were incubated with ³²P-labeled ATP and irradiated with UV light. Two bands corresponding to 55 kDa (p55) and 100 kDa (p100) proteins were successfully detected (Figure ​(Figure2B).2B). We focused on the stronger band, p55. From the observed expression in HeLa cells, nuclear localization, molecular weight on SDS-PAGE, and binding to anion exchange resin at pH 7.5 (Figure ​(Figure2C,2C, and Supplementary Figure S4), we presumed that p55 was UAP56 or DBP5. Therefore, we performed an IP assay to identify p55 using the cross-linked fraction. As a result, p55 was specifically precipitated using an antibody against UAP56, but not with one against DBP5 (Figure ​(Figure2D),2D), indicating that p55 was UAP56. GST pull-down and western blot analysis clearly showed the interaction between PHAX and UAP56 in HNEs (Figure ​(Figure2E).2E). We also observed an interaction between PHAX and ALY. The candidate RNA helicase for p100 was GEMIN3 since its molecular weight is 92.2 kDa and the theoretical isoelectric point is 6.49. However, the interaction of PHAX with GEMIN3 was not observed. Similarly, the interaction of PHAX with GEMIN2 or Y14 was not observed.

Open in a separate window

Figure 2.

Identification of candidate helicases for ATP-dependent loading of PHAX (A) Schematic diagram of the process for identifying candidate helicases. HNEs were incubated with GST-PHAX, His-CBP80 and His-CBP20 in the presence of ATP and RNase A, followed by GST pull-down and HRV 3C protease treatment. PHAX-interacting proteins were irradiated by ultraviolet (UV) light and detected by SDS-PAGE and autoradiography. (B) The two bands corresponding to 55 kDa and 100 kDa were named p55 and p100, respectively. (C) The molecular weight (Mw) and theoretical isoelectric point (pI) of RNA helicases of ~55 kDa are shown. Mw and theoretical pI were obtained from the ExPASy Compute pI/Mw tool. (D) The irradiated sample in (B) was immunoprecipitated using antibodies against UAP56 and DBP5, then detected by SDS-PAGE and autoradiography. (E) HNEs were subjected to GST pull-down assay using GST or GST-PHAX in the presence of RNase A followed by western blotting.

UAP56 and URH49 stimulate RNA binding of PHAX

To investigate if UAP56 could load PHAX onto U snRNA, we performed the in vitro RNA-protein binding assay (Figure ​(Figure3A).3A). A mixture of ³²P-labeled in vitro transcribed RNAs containing intronless dihydrofolate reductase (DHFR) mRNA, U1ΔSm, U5ΔSm, U6Δss snRNAs, and tRNA^Phe was incubated with purified recombinant GST-PHAX and either buffer or FLAG-UAP56 in the presence of HNEs and ATP, and a GST pull-down assay was performed. As shown in Supplementary Figure S2, GST-PHAX efficiently bound to U1ΔSm and U5ΔSm snRNAs (Figure ​(Figure3B,3B, lane 4). PHAX-binding to U1ΔSm and U5ΔSm snRNAs was stimulated by the addition of purified FLAG-UAP56 WT compared with the K95E mutant (Figure ​(Figure3B3B and C). Pull-down efficiency of GST-PHAX was not affected by FLAG-UAP56 WT or the K95E mutant (Figure ​(Figure3D).3D). These results indicate that UAP56 loaded PHAX onto U snRNAs in an ATP-dependent manner. PHAX interacted with FLAG-UAP56 WT but not the K95E mutant (Figure ​(Figure3D),3D), suggesting that PHAX preferentially interacted with the ATP-bound form of UAP56.

Open in a separate window

Figure 3.

ATP-bound UAP56 stimulates RNA binding of PHAX (A) Schematic diagram of the in vitro PHAX-RNA binding assay. (B) A mixture of in vitro-transcribed ³²P-labeled RNAs containing dihydrofolate reductase (DHFR) mRNA, U1ΔSm small nuclear RNA (snRNA), U5ΔSm snRNA, U6Δss snRNA, and tRNA^Phe was incubated with HeLa nuclear extracts (HNEs), ATP, either GST or GST-PHAX, and either buffer, FLAG-UAP56 WT, or K95E mutant. Then, GST pull-down assays were performed. U6Δss snRNA and tRNA^Phe were uncapped, while the other RNAs were m⁷G-capped. Pulled down RNAs were analyzed by denaturing PAGE and autoradiography. (C) The quantification of pull-down efficiency of DHFR mRNA, U1ΔSm snRNA, U5ΔSm snRNA and tRNA^Phe from four independent experiments performed as in (B). Pull-down efficiencies were determined by dividing the amount of pulled down RNAs by the amount of each RNA input. Averages and standard deviations are noted. P-values were calculated by a two-tailed t-test. (D) Pulled down proteins were analyzed by SDS-PAGE and western blotting. ³²P-labeled RNAs were not included in this assay. TBP: TATA-binding protein. (E) A mixture of in vitro-transcribed ³²P-labeled RNAs containing U1ΔSm, U5ΔSm, and U6Δss snRNAs was incubated with HNEs, ATP, GST-PHAX and either buffer, FLAG-UAP56 (39B), FLAG-URH49 (39A), or FLAG-DBP5 (19). Then, GST pull-down assays were performed. Pulled down RNAs were analyzed by denaturing PAGE and autoradiography. Pulled down proteins were analyzed by SDS-PAGE and western blotting. ³²P-labeled RNAs were not included in this protein–protein binding assay. (F) Quantification of pull-down efficiency of U1ΔSm, U5ΔSm and U6Δss snRNAs from three independent experiments performed as in (E). Pull-down efficiencies were determined by dividing the amount of pulled down RNAs by the amount of each RNA input. Averages and standard deviations are noted. P-values were calculated by a two-tailed t-test. *P< 0.05, ns: not significant.

UAP56/DDX39B and URH49/DDX39A are closely related RNA helicases (56) (Supplementary Figure S3B). The in vitro RNA-protein binding assay shows that URH49/DDX39A, but not DBP5/DDX19, also stimulated RNA binding of PHAX (Figure ​(Figure3E3E and F). UAP56 and URH49, but not DBP5, interacted with PHAX (Figure ​(Figure3E).3E). These results suggest that the loading of PHAX onto RNA is the unique activity of the two DDX39 helicases, UAP56 and URH49. Similar results were obtained when the GST pull-down assay was performed using nuclear lysates from Xenopus oocytes instead of HNEs (Supplementary Figure S5A), indicating that in Xenopus oocytes as well as in HeLa cells, UAP56 loaded PHAX onto U snRNAs.

To test whether UAP56 could directly load PHAX onto RNA, we performed an in vitro RNA-protein binding assay using purified recombinant proteins in the absence of HNEs. A mixture of ³²P-labeled in vitro transcribed RNAs containing U1ΔSm, U5ΔSm, and U6Δss snRNAs was incubated with purified recombinant CBC, PHAX, and UAP56 in the presence or absence of ATP or a slowly hydrolyzable ATP analog (ATP-γS), and an RNA co-IP assay was performed (Supplementary Figure S5B). UAP56 did not stimulate RNA binding of PHAX under these experimental conditions. This suggests that some factors in HNEs could be involved in the loading of PHAX onto RNA in cooperation with UAP56.

ATP stimulates the interaction between PHAX and UAP56 in HNEs

To identify the additional factors involved, we analyzed the interaction between PHAX and UAP56 in the presence of HNEs. The interaction was stimulated by the addition of ATP (Figures ​(Figures4A,4A, lanes 7 and 8, 4B, lanes 5 and 6, and Supplementary Figures S6A and B). Consistently, when the K95E mutant was used, the ATP-dependent stimulation was notably weak (Figure ​(Figure4B,4B, lanes 7 and 8, and Supplementary Figure S6B). These results suggest that PHAX preferentially interacted with the ATP-bound form of UAP56 in the presence of HNEs. To investigate whether the interaction between PHAX and UAP56 was direct, a GST pull-down assay was performed in the absence of HNEs. The interaction between PHAX and UAP56, if any, was notably weak, and its ATP dependency was not observed (Figure ​(Figure4C,4C, and Supplementary Figure S6C), indicating that the observed PHAX-UAP56 interaction in HNEs was indirect. The interaction was seen following the addition of HNEs (Supplementary Figure S6D), suggesting that some factors in HNEs mediated the interaction between PHAX and UAP56.

Open in a separate window

Figure 4.

UAP56 interacts with PHAX in an ATP-dependent manner (A) HeLa nuclear extracts (HNEs) and RNase A were incubated with GST or GST-PHAX in the presence or absence of ATP. Pulled down proteins were detected by western blotting using anti-UAP56 and anti-CBP80 antibodies. The gel after blotting was stained with Coomassie brilliant blue (CBB). (B) HNEs, GST-PHAX, and RNase A were incubated with FLAG-UAP56 WT or K95E mutant in the presence or absence of ATP. Pulled down proteins were detected by western blotting using an anti-FLAG antibody. The gel after blotting was stained with CBB. (C) HNEs and RNase A were incubated with either GST or GST-PHAX and either FLAG-UAP56 WT or K95E mutant in the presence or absence of ATP. Pulled down proteins were detected by western blotting using an anti-FLAG antibody. The gel after blotting was stained with CBB. Uncropped CBB staining is shown in Supplementary Figure S6.

Identification of bridging factors between PHAX and UAP56

To search for the bridging factors between PHAX and UAP56, we sequentially isolated proteins that interacted with both PHAX and UAP56 from HNEs (Figure ​(Figure5A).5A). GST-PHAX and FLAG-UAP56 were incubated with HNEs, ATP, and RNase A. The mixture was applied to anti-FLAG M2 agarose to isolate FLAG-UAP56-interacting proteins. The eluate with FLAG peptide was applied to Glutathione Sepharose to isolate GST-PHAX-interacting proteins. The bound proteins were released from the beads by digesting the GST tag with HRV 3C protease, and the supernatant was fractionated by SDS-PAGE followed by CBB staining. Several specific protein bands were detected only when both GST-PHAX and FLAG-UAP56 were added to the isolation system (Figure ​(Figure5B),5B), and they were analyzed by mass spectrometry (Figure ​(Figure5C,5C, and Supplementary Table S2). Many of them were components of the TREX complex: HPR1, CIP29 and ALY (Figure ​(Figure5C).5C). Among them, ALY was a good candidate for the bridging factor because an ATP-dependent interaction between UAP56 and ALY was previously reported (57). The interaction between PHAX and ALY in HNEs was observed (Figures ​(Figures2E and2E and 5B, lower panel).

Open in a separate window

Figure 5.

Identification of bridging factors between PHAX and UAP56 (A) Schematic diagram of the process for identifying bridging factors. (B) HeLa nuclear extracts (HNEs) and RNase A were incubated with either buffer or GST-PHAX and either buffer or FLAG-UAP56, in the presence of ATP. Immunoprecipitation (IP) experiments were then performed. The bound proteins were eluted with a FLAG peptide. The eluate was subjected to GST pull-down. The bound proteins were treated with HRV 3C protease. The supernatant was analyzed by sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and mass spectrometry or western blotting using an anti-ALY antibody. (C) Identified proteins specific to lane 4 of (B) are listed. Components of the TREX complex are shown in red text.

ALY acts as a bridge between PHAX and UAP56

When the GST pull-down assay was performed using ALY-depleted HNEs, the interaction between PHAX and UAP56 was abolished (Figure ​(Figure6A,6A, and Supplementary Figure S7A), indicating that ALY was required for the PHAX-UAP56 interaction. When the GST pull-down assay was performed in the absence of HNEs, the efficient interaction between PHAX and UAP56 was observed only when GST-PHAX, FLAG-UAP56, FLAG-ALY and ATP were all included (Figure ​(Figure6B,6B, lane 8, and Supplementary Figure S7B), indicating that ALY was sufficient for the interaction between PHAX and the ATP-bound form of UAP56. The interaction between PHAX and ALY was also observed (Figure ​(Figure6B,6B, lanes 7 and 8, and Supplementary Figure S7B). UAP56 and ALY were slightly pulled down by GST in the presence of ALY (Figure ​(Figure6B,6B, lanes 3 and 4), probably because ALY was a sticky protein and bound to GST or beads to some extent. The direct interaction between PHAX and UAP56, if any, was weak and not stimulated by ATP (Figure ​(Figure6B,6B, lanes 5 and 6). To investigate the direct interaction between PHAX and ALY, the GST pull-down assay was performed using purified GST-ALY and T7-PHAX. As expected, the interaction between PHAX and ALY was direct (Figure ​(Figure6C,6C, and Supplementary Figure S7C). These results showed that ALY could act as a bridge between PHAX and the ATP-bound form of UAP56 (Figure ​(Figure6D6D).

Open in a separate window

Figure 6.

ALY bridges between PHAX and ATP-bound UAP56 (A) HNEs and RNase A were subjected to protein A Sepharose that was pre-bound to rabbit antibody against mouse IgG (control) or ALY (ΔALY). The depleted HNEs were incubated with GST-PHAX in the presence or absence of ATP, and GST pull-down was performed. The bound proteins were analyzed by SDS-PAGE and western blotting using antibodies against UAP56, ALY or CBP80. The gel after blotting was stained with CBB. (B) Purified recombinant GST or GST-PHAX was incubated with FLAG-UAP56 and RNase A in the presence or absence of FLAG-ALY and ATP, and GST pull-down was performed. The pulled down proteins were analyzed by SDS-PAGE and western blotting using an anti-FLAG antibody. The relative pull-down efficiency of FLAG-UAP56 and FLAG-ALY is shown. The gel after blotting was stained with CBB. (C) Purified recombinant GST or GST-ALY was incubated with T7-PHAX and RNase A, and GST pull-down was performed. Pulled down proteins were analyzed by SDS-PAGE and western blotting using an anti-T7 antibody. The gel after blotting was stained with CBB. (D) Schematic representation of the interaction between PHAX and UAP56 via ALY. Uncropped CBB staining is shown in Supplementary Figure S7.

UAP56 and ALY are involved in U snRNA export

UAP56 interacted with PHAX and stimulated PHAX-binding to U snRNAs in HeLa nuclear extracts and Xenopus oocyte nuclear lysates (Figure ​(Figure33 and Supplementary Figure S5), and that ALY mediated the interaction between UAP56 and PHAX (Figure ​(Figure6).6). To investigate if UAP56 and ALY are involved in U snRNA export, we performed the well-established Xenopus oocyte microinjection assay, in which RNA export is separated from other steps of gene expression and its kinetics is evaluated (e.g. (27,55,58,59) (Figure ​(Figure7A).7A). In this assay, a mixture of ³²P-labeled RNAs containing intronless DHFR mRNA, U1ΔSm, U5ΔSm, U6Δss snRNAs, and tRNA^Phe was injected with an anti-mouse IgG antibody control into the nuclei of Xenopus oocytes. After incubation, the oocyte was dissected to separate the nucleus and cytoplasm. RNAs were then extracted from each fraction, separated using PAGE, and detected by autoradiography. All RNAs were nuclear immediately after nuclear injection (Figure ​(Figure7B,7B, lanes 1 and 2). After 3.5 h of incubation, DHFR mRNA, U1ΔSm snRNA, U5ΔSm snRNA and tRNA^Phe were exported to the cytoplasm, whereas the non-exported U6Δss snRNA control remained in the nucleus (Figure ​(Figure7B,7B, lanes 3 and 4). We used two antibodies against UAP56 and ALY. In vitro experiments show that the anti-UAP56 antibody delayed ATP hydrolysis by UAP56, and that the anti-ALY antibody inhibited RNA binding of ALY (Supplementary Figure S8). As the antibodies are polyclonal, they may have additional effects. When RNAs were injected with the anti-UAP56 antibody, DHFR mRNA export was strongly inhibited (Figure ​(Figure7B,7B, lanes 5 and 6, and Figure ​Figure7C).7C). Notably, U1ΔSm snRNA export was slightly inhibited and U5ΔSm snRNA export was efficiently inhibited (Figure ​(Figure7B,7B, lanes 5 and 6, and Figure ​Figure7C).7C). Similar results were obtained when using the anti-ALY antibody (Figure ​(Figure7B,7B, lanes 7 and 8, and Figure ​Figure7C).7C). These results indicate that, in addition to mRNA export, UAP56 and ALY are involved in U snRNA export.

Open in a separate window

Figure 7.

UAP56 and ALY are involved in U small nuclear RNA (snRNA) export (A) Schematic diagram of Xenopus oocyte microinjection. (B) The same RNA mixture as in Figure ​Figure33 was injected with a rabbit antibody against mouse IgG (control), UAP56, or ALY into the nucleus. U6Δss snRNA and tRNA^Phe were uncapped, while the other RNAs were m⁷G-capped. U6Δss does not leave the nucleus and served as an internal control for nuclear integrity. RNA was immediately extracted from nuclear (N) and cytoplasmic (C) fractions (lanes 1 and 2) or 3.5 h (lanes 3–8) after the injection, and then analyzed by denaturing polyacrylamide gel electrophoresis (PAGE) and autoradiography. (C) Quantification of the export of dihydrofolate reductase mRNA (DHFR), U1ΔSm snRNA (U1), U5ΔSm snRNA (U5), and tRNA^Phe from three independent experiments performed in (B). The export efficiency of the control was set to 100%. Averages and standard deviations are noted. P-values were calculated by comparisons with the control: *P< 0.05, ***P< 0.001, ns: not significant. (D) The same RNA mixture was injected with BSA-NES (NES) or BSA-NES M10 mutant (NES mut). BSA-NES, but not NES M10 mutant, saturates CRM1 export (61). RNA export was analyzed as in (B). (E) Quantification of the export of DHFR mRNA, U1ΔSm snRNA, U5ΔSm snRNA, and tRNA^Phe from three independent experiments performed in (D). The export efficiency of the control was set to 100%. Averages and standard deviations are noted. P-values were calculated by comparisons with the NES mut: ***P< 0.001, ns: not significant. (F) The same RNA mixture was injected with CTE RNA (CTE) or CTE M36 mutant RNA (CTE mut). CTE, but not CTE M36 mutant, binds specifically to TAP and inhibits the TAP-dependent export (62). RNA export was analyzed as in (B). (G) Quantification of the export of DHFR mRNA, U1ΔSm snRNA, U5ΔSm snRNA, and tRNA^Phe from three independent experiments performed in (F). Averages and standard deviations are noted. P-values were calculated by comparisons with the CTE mut: ***P< 0.001, ns: not significant.

It has been reported that bulk mRNAs and U snRNAs are exported by different pathways, the TAP-dependent and CRM1-dependent pathways, respectively (20,27,55,60,61). However, it was thought that a fraction of U snRNAs might be exported exclusively by the TAP-dependent pathway. To examine this possibility, we used two inhibitors, BSA-NES and CTE. BSA-NES is a conjugate of NES peptides coupled to BSA that binds specifically to CRM1 (61), whereas CTE is a constitutive transport RNA element from type D retroviruses that binds specifically to TAP (62). Injection of BSA-NES and CTE inhibits the CRM1- and TAP-dependent pathways by saturating CRM1 and TAP, respectively. When RNAs were injected with BSA-NES, U1ΔSm and U5ΔSm snRNA export, but not DHFR mRNA or tRNA^Phe export, was strongly inhibited (Figure ​(Figure7D7D and E). This indicates that U snRNAs, but not DHFR mRNA or tRNA^Phe, were exported by the CRM1-dependent pathway. When CTE was injected, DHFR mRNA export, but not U snRNA or tRNA^Phe export, was strongly inhibited, indicating that DHFR mRNA was exported by the TAP-dependent pathway, but U snRNAs were not (Figure ​(Figure7D7D and E). These results suggested that UAP56 and ALY participated in U snRNA export through a mechanism distinct from that of mRNA export.

We investigated the importance of UAP56 ATP binding activity in U snRNA export (Supplementary Figure S9). Consistent with previous reports (13,51), injection of WT UAP56 inhibited mRNA export in a dose-dependent manner (Supplementary Figure S9B, lanes 5–10). The UAP56 K95E mutant protein inhibited DHFR mRNA export more strongly than the WT protein (Supplementary Figure S9B, lanes 11–16) (51). Notably, the K95E mutant protein also inhibited the export of U1 and U5 snRNAs (Supplementary Figure S9). These results suggest that UAP56 and its ATP binding activity are important for export of not only mRNAs, but also U snRNAs.

We thank Makoto Kitabatake (Kyoto University) for his continuous support and critical comments on this manuscript, Shintaro Akiyama and Naoyuki Kataoka for sharing unpublished data, and the members of the Ohno and Hirose laboratories for valuable discussions. We thank Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.