MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental material
Right arrow Other Versions of this Article:
MCB.00814-06v1
27/1/368    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrowReprints and Permissions
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Cho, S.
Right arrow Articles by Jang, S. K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Cho, S.
Right arrow Articles by Jang, S. K.

 Previous Article  |  Next Article 

Molecular and Cellular Biology, January 2007, p. 368-383, Vol. 27, No. 1
0270-7306/07/$08.00+0     doi:10.1128/MCB.00814-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

BiP Internal Ribosomal Entry Site Activity Is Controlled by Heat-Induced Interaction of NSAP1{dagger} ,{triangledown}

Sungchan Cho, Sung Mi Park, Tae Don Kim, Jong Heon Kim, Kyong-Tai Kim, and Sung Key Jang*

NRL, POSTECH Biotech Center, Department of Life Science, Pohang University of Science and Technology, San 31, Hyoja Dong, Pohang, Kyungbuk 790-784, Republic of Korea

Received 8 May 2006/ Returned for modification 28 June 2006/ Accepted 4 October 2006


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
TheBiP protein, a stress response protein, plays an important role in the proper folding and assembly of nascent protein and in the scavenging of misfolded proteins in the endoplasmic reticulum lumen. Translation of BiP is directed by an internal ribosomal entry site (IRES) in the 5' nontranslated region of the BiP mRNA. BiP IRES activity increases when cells are heat stressed. Here we report that NSAP1 specifically enhances the IRES activity of BiP mRNA by interacting with the IRES element. Overexpression of NSAP1 in 293T cells increased the IRES activity of BiP mRNA, whereas knockdown of NSAP1 by small interfering RNA (siRNA) reduced the IRES activity of BiP mRNA. The amount of NSAP1 bound to the BiP IRES increased under heat stress conditions, and the IRES activity of BiP mRNA was increased. Moreover, the increase in BiP IRES activity with heat treatment was not observed in cells lacking NSAP1 after siRNA treatment. BiP mRNAs were redistributed from the heavy polysome to the light polysome in NSAP1 knockdown cells. Together, these data indicate that NSAP1 modulates IRES-dependent translation of BiP mRNA through an RNA-protein interaction under heat stress conditions.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Translation of eukaryotic mRNAs occurs either by cap-dependent scanning or by the binding of a ribosome to a specialized RNA element called an internal ribosomal entry site (IRES) (9). Since the discovery of IRESs in the 5' nontranslated regions (5'NTRs) of encephalomyocarditis virus (13) and poliovirus (28), many viral and cellular mRNAs have been shown to contain IRESs (http://www.rangueil.inserm.fr/iresdatabase).BiP was the first cellular mRNA shown to be translated by an IRES element in its 5'NTR (34).

IRES activities of many cellular mRNAs are modulated in diverse biological processes, including differentiation, cell cycle stages, and apoptosis, and under stress conditions such as hypoxia, UV irradiation, and heat. For instance, translation of a set of cellular mRNAs can be sustained or even induced through activation of the IRES element under stress conditions when translation of most mRNAs is repressed.

Heat stress is an important cellular stress from which most of our understanding of stress-induced translational control has emerged (3). Two main events are involved in the control of translation under heat stress conditions. First, a general repression of translation occurs owing to changes in the phosphorylation states of many eukaryotic initiation factors (32), and mRNAs and translation factors are sequestered to the subcellular stress granule (15). Second, synthesis of some important proteins required for cell survival under stress conditions, including heat shock proteins like the BiP protein, continues or even increases during heat treatment.

The BiP protein has been identified as an immunoglobulin heavy-chain-binding protein (2, 6) that also binds transiently to several nascent, wild-type secretory and transmembrane proteins and permanently to misfolded proteins that accumulate within the endoplasmic reticulum (ER). Proposed roles for the BiP protein include the mediation of proper folding, the assembly of nascent proteins, and the scavenging of misfolded proteins in the ER (2, 27). The BiP protein, also known as glucose-regulated protein 78 (GRP78), is a member of the heat shock protein 70 (HSP70) family (25). The expression of BiP is regulated at the transcriptional level. Its synthesis can be induced by several stress conditions, such as glucose starvation (23) and treatment with calcium ionophores, calcium-chelating agents such as EGTA (23, 39, 40), and tunicamycin or glucosamine, both of which block cellular glycosylation (26). Transcription of BiP mRNA is also induced after infection with paramyxovirus (29, 35). Furthermore, several reports have suggested that expression of the BiP protein is further regulated at the level of translation (31, 34, 37). For example, translation of BiP mRNA is increased in poliovirus-infected cells where translation of most host cellular mRNAs is inhibited (34). The 5'NTR of BiP mRNA contains an IRES element (42) to which polypyrimidine tract-binding protein (PTB) and autoantigen La proteins bind and modulate its activity (19, 20). Moreover, continuous heat treatment was shown to increase the IRES activity of BiP mRNA (21). However, the molecular basis of IRES activation by heat treatment remains to be determined.

In addition to the canonical translation initiation factors, which function in cap-dependent translation as well, many RNA-binding proteins play important roles in IRES-dependent translation (9, 12). For instance, host cell proteins such as PTB, La, poly(rC)-binding proteins, and the upstream of N-ras (Unr) protein bind directly to IRES elements and increase the translation of many viral and cellular mRNAs. Recently, NSAP1 was shown to function as an IRES-interacting trans-acting factor (ITAF) in hepatitis C virus (HCV) IRES-dependent translation through an interaction with a core-coding sequence downstream of the HCV initiation codon (18).

NSAP1 was originally identified as a protein capable of interacting with NS1, which is the main nonstructural protein of the mouse minute virus (7). NSAP1 is highly homologous (82.1% identity) to human heterogeneous nuclear ribonucleoprotein R (hnRNP R) (8). NSAP1 binds to RNA in vitro, preferentially to poly(A) or poly(U) regions, in a phosphorylation-dependent manner (10, 11, 24). NSAP1 was shown to modulate c-fos mRNA stabilization by forming a protein complex with Unr, PABP, and PAIP1 (5) and was found as a component of mRNA granules in neuronal cells (1a, 14).

Here we present a molecular mechanism for increased translation of BiP mRNA under the control of a heat-dependent IRES element. Using in vivo assay systems based on overexpression and/or knockdown of NSAP1, we show that NSAP1 binds to the IRES element of the BiP mRNA and increases its translation. We also show that the IRES activity of BiP mRNA and the binding of NSAP1 to the BiP IRES element increase under heat stress conditions. Moreover, NSAP1 knockdown reduces the activation of IRES-dependent translation of BiP mRNA with heat treatment. Redistribution of BiP mRNAs in ribosomal profiles of control and NSAP1 knockdown cells also supports the positive role of NSAP1 in BiP mRNA translation. Together, these results indicate that NSAP1 plays an important role in modulation of BiP mRNA translation via heat-induced binding to the IRES element.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Plasmid construction. Plasmid pR/BiP/F was constructed by inserting the 5'NTR of the BiP mRNA into the intercistronic site of the pRF vector as described previously (19). To determine the presence of the cis-acting IRES element in the BiP mRNA 5'NTR, serial deletions of the BiP 5'NTR were generated from pR/BiP/F (the A in the initiation codon is referred to as nucleotide +1 hereafter). To construct plasmids pR/BiP(–171 to –1)/F, pR/BiP({Delta}–95 to –8)/F, pR/BiP({Delta}–171 to –8)/F and pR/BiP(–95 to –1)/F, plasmid pR/BiP/F was digested with restriction enzymes SalI/NruI, XmaI/BstXI, NruI/BstXI, and SalI/XmaI, respectively, and the DNA fragments were then self-ligated. Plasmid pR/BiP(–171 to –1)/F was digested with restriction enzymes XmaI/BstXI, and the DNA fragment was self-ligated to construct pR/BiP(–171 to –95)/F. Construction of pSK/BiP(–225 to –1)-CAT has been described elsewhere (19). Plasmids pR/HCV/F, pR/Polio/F, and pGFP-NSAP1 were constructed as described previously (18). Plasmid pR/Apaf/F was constructed by ligating a PCR fragment (corresponding to the 5'NTR of the Apaf mRNA) that had been digested with restriction enzymes SacI/XhoI with the pRF vector linearized with restriction enzymes SalI/KpnI. To construct pGFP-hnRNP L, pGFP-hnRNP C1, and pGFP-PCBP1, each open reading frame was inserted into the linearized pEGFP DNA. The construction strategy used for pEBV-U6+27/NSAP1(734-752) has been described elsewhere (18).

UV cross-linking, immunoprecipitation of UV cross-linked proteins, and competition experiments. All of the UV cross-linking, immunoprecipitation, and competition experiments were performed as previously described (17), except that 32P-labeled RNA corresponding to the BiP IRES and purified NSAP1 were used as RNA and protein, respectively. To examine the ability of NSAP1 to bind to the BiP IRES under heat stress conditions, HeLa cells were grown at 42°C for 15 h and then lysed to obtain cytoplasmic extracts as described previously (21).

RNA affinity chromatography. All of the experiments were performed as previously described (18), except that biotinylated RNAs corresponding to different regions of the BiP 5'NTR and cytoplasmic extracts of mock-treated and heat-stressed HeLa cells were used.

Purification of recombinant NSAP1. The procedure used for purification of recombinant NSAP1 from Escherichia coli cells has been described previously (18).

Establishment of cell clones expressing NSAP1 siRNAs. Plasmids (1 µg each) expressing small interfering RNAs (siRNAs) [pEBV-U6+27 and pEBV-U6+27/NSAP1(734-752)] were transfected into HeLa cells by electroporation. From 48 h posttransfection, cells were maintained in Dulbecco's modified Eagle's medium containing hygromycin (300 µg/ml; Calbiochem). After 1 month of selection, hygromycin-resistant cell colonies were pooled and cultivated for further analysis.

Cell culture and reporter assay. 293T and HeLa cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen) supplemented with 10% fetal bovine serum (Clontech). For the reporter assay, 293T cells transfected with dicistronic reporter plasmids and effector plasmids were lysed in passive lysis buffer (Promega) at 48 h posttransfection. Firefly luciferase (FLuc) and Renilla luciferase (RLuc) activities were measured with the Dual-Luciferase Reporter Assay System (Promega). ß-Galactosidase activity was determined from the same lysate with the ß-galactosidase Enzyme Assay System (Promega). To analyze the IRES activity of BiP mRNA under heat stress conditions, HeLa cells expressing NSAP1-specific siRNA were transfected with dicistronic reporter plasmid pR/BiP/F and at 24 h posttransfection were transferred to 42°C. After 15 h, cells were harvested and FLuc and RLuc activities were determined. For transfection experiments with dicistronic reporter RNA, 10 µg capped RNA generated in vitro was transfected into HeLa cells expressing NSAP1-specific siRNA. Cell lysates were prepared at 3 h posttransfection, and luciferase activities were measured.

Western blot analysis. Immunoblot analysis was performed with monoclonal anti-BiP (Stressgen), anti-green fluorescent protein (anti-GFP; Santa Cruz), anti-HuR (Santa Cruz), anti-hnRNP L (kindly provided by G. Dreyfuss, University of Pennsylvania), and anti-actin (ICN) as the primary antibodies and horseradish peroxidase-conjugated anti-mouse immunoglobulin G as the secondary antibody. For detection of NSAP1, an anti-NSAP1 polyclonal antibody (kindly provided by A. Mizutani, RIKEN, Japan) and horseradish peroxidase-conjugated anti-rabbit immunoglobulin G were used as the primary and secondary antibodies, respectively. The secondary antibodies were visualized by enhanced chemiluminescence (Amersham-Pharmacia Biotech).

mRNA purification and Northern blot analysis. The levels of dicistronic reporter mRNA (pR/BiP/F) in transfected cells were monitored by Northern blot analysis with a 32P-labeled probe corresponding to the FLuc gene. Three micrograms of poly(A)+ RNA purified from transfected cells was subjected to analysis. Levels of NSAP1 and BiP mRNA were also analyzed with 30 µg of total RNA prepared from HeLa cells expressing NSAP1-specific siRNA, as described previously (18). For detection of BiP mRNA, a sequence corresponding to nucleotides 220 to 741 of BiP mRNA was used as a probe.

Generation of adenoviruses expressing siRNAs. The backbone for the vector utilized in the generation of adenovirus expressing siRNA against NSAP1 was made by PCR with two primers specific for pShuttle-IRES-hrGFP vector (Stratagene) (5' primer 5'-AACTC GAGGG GTGGG AAAGA ATATA TA-3' and 3' primer 5'-AAAGG CCTTA CGCGC TATGA GTAAG TG-3'; restriction sites for subsequent cloning are underlined). Plasmid pShuttle-RNAi was constructed by ligation of the amplified DNA treated with XhoI and StuI and a DNA fragment of the pSuper RNAi system (Oligoengine) containing the H1 promoter generated by EcoR1-T4 DNA polymerase-XhoI treatments. To generate pShuttle-siNSAP1, two oligomers (sense, 5'-GATCT CCACT GGAAC GAGTG AAGAA GTTCA AGAGA CTTCT TCACT CGTTC CAGTT TTTTG GAAA-3'; antisense, 5'-AGGTG ACCTT GCTCA CTTCT TCAAG TTCTC TGAAG AAGTG AGCAA GGTCA AAAAA CCTTT TCGA-3' [the flanking sequences for cloning are underlined, and the sequences corresponding to an siRNA against NSAP1, siNSAP1, are italicized]) were annealed together and then cloned into the BglII/HindIII sites of the pShuttle-RNAi vector. A plasmid named pShuttle-siNSAP1(mt) expressing a negative control siRNA against NSAP1 with changes from 5'-ACT GGAAC GAGTG AAGAA G-3' to 5'-AGTCG AACGA GTCAA CAAC-3' (mutated sequences are in bold) was constructed by site-directed mutagenesis. Generation of recombinant adenoviruses from pShuttle-siNSAP1 and pShuttle-siNSAP1(mt) and amplification of adenoviral particles containing the recombinants were carried out according to the manufacturer's recommendations (AdEasy; Stratagene).

Ribosomal profiling. HeLa cells were treated with cycloheximide (100 µg/ml) for 5 min at 37°C and then harvested. Analyses of cell extracts by sucrose gradient fractionation were performed as described by Rousseau et al. (33). Fractionated samples were classified into six groups (unbound proteins, 40S, 60S, 80S, light polysome [LP], and heavy polysome [HP]). Total RNAs and proteins in each fraction were prepared by using TRI-reagent (Invitrogen) and the trichloroacetic acid precipitation method, respectively. The amounts of specific RNAs and proteins in each fraction were monitored by reverse transcription (RT)-PCR in the presence of [32P]dCTP (36) and Western blotting, respectively.

Immunocytochemistry. The subcellular localizations of endogenous NSAP1 in HeLa cells before and after heat treatment for 15 h were monitored by using a specific polyclonal antibody against NSAP1 as previously described (1).

Fractionation of cell extracts. HeLa cells grown at 37°C and 42°C for 5 and 15 h were fractionated into cytoplasmic and nuclear fractions as previously described (17) and then subjected to immunoblotting with a polyclonal anti-NSAP1 antibody.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Determination of heat-responsive element in the BiP IRES. The physiological importance of IRES-dependent translation of cellular mRNA has been increasingly recognized for numerous biological processes, including differentiation, cell cycle processes, and apoptosis, and under stress conditions such as hypoxia, UV irradiation, and heat stress. For instance, BiP IRES activity is increased under heat stress conditions (21).

When we observed the levels of endogenous BiP, HuR, and actin proteins by Western blotting, the amount of BiP protein markedly increased whereas the expression levels of HuR and actin did not show a noticeable change after 15 h of heat treatment (Fig. 1A, part i). To assess the extent of translational control in the accumulation of BiP protein under heat stress conditions, we monitored the level of BiP mRNA by the radiolabeling RT-PCR method. The amount of BiP mRNA was increased by about threefold after heat treatment for 5 h (Fig. 1A, parts ii and iii), suggesting that the increase in BiP protein after 5 h of heat treatment (2.2-fold increase shown in Fig. 1A, part i) is most likely attributable to the increased BiP mRNA level. However, a longer heat treatment (15 h) led to a slight decrease in the level of BiP mRNA compared with the untreated one (about 75% of the mock-treated level, shown in Fig. 1A, parts ii and iii). On the other hand, the level of BiP protein was increased by 4.2-fold after 15 h of heat treatment (Fig. 1A, parts i and iii). This indicates that the increase in BiP protein after 15 h of heat treatment is attributable to posttranscriptional regulation of gene expression. The level of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was not altered throughout heat treatment.


Figure 1
Figure 1
View larger version (50K):
[in this window]
[in a new window]

  		FIG.1. Determination of heat-responsive elements in the BiP 5'NTR. (A, part i) Induction of BiP protein under heat stress conditions. Lysates of HeLa cells grown at 37°C (lane M) and 42°C for 5 and 15 h (lanes 5 and 15) were subjected to immunoblotting with anti-BiP, anti-HuR, and anti-actin antibodies. (ii) Level of BiP mRNA under heat stress conditions. Total RNAs were prepared from HeLa cells grown at 37°C (lane M) and 42°C for 5 and 15 h (lanes 5 and 15) and then subjected to radiolabeling RT-PCR with BiP-specific primers. GAPDH mRNA was used as a control message. (iii) Relative levels of BiP protein and BiP mRNA under heat stress condition. The intensities of BiP protein (i) and BiP mRNA bands (ii) were measured by the Sion Image Analysis Program, and the ratios of the band intensities before and after heat treatment (fold increases) are represented by filled squares and open circles, respectively. The intensities of BiP protein and BiP mRNA in untreated cells were set to 1. (iv) HeLa cells were transfected with dicistronic reporter plasmid pR/BiP/F or pR/Polio/F (5 µg each). These plasmids produce dicistronic mRNAs containing the RLuc and FLuc genes. Translation of RLuc occurs by cap-dependent scanning. On the other hand, translation of FLuc is directed by the BiP IRES or the polioviral IRES in the intercistronic region. The transfected HeLa cells were cultivated at 37°C for 24 h. Thereafter, cells were transferred to an incubator preheated to 42°C and maintained for 5 or 15 h. After heat treatment, the cells were harvested and FLuc and RLuc activities in the cell lysates were measured. The white and black columns depict FLuc and RLuc activities, respectively. The luciferase activities in cell lysate not subjected to heat treatment were set to 1. (B) Schematic diagram of the dicistronic mRNAs used for monitoring of the IRES activities of truncated BiP mRNAs. (C) HeLa cells were transfected with reporter plasmids expressing the dicistronic mRNAs shown in panel B. Twenty-four hours posttransfection, some cells were transferred to an incubator preheated to 42°C and some were left at 37°C. After 15 h of incubation, cells were harvested and luciferase activities were measured. The ratios of FLuc activity to RLuc activity were calculated and are shown in boxes below the graph. The ratio of luciferase activities in cells containing the dicistronic mRNA with the entire BiP IRES (nucleotides –225 to –1; RNA I in panel B) at 37°C was set to 1. The gray and black columns depict relative luciferase activities in mock-treated and heat-treated cells, respectively. The fold increases were calculated by the ratios of heat-stressed to mock-treated cells and are shown above the columns. Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars.

 
The BiP IRES activity was monitored by transfection of a reporter plasmid producing a dicistronic mRNA, R/BiP/F (Fig. 1A, part iv), containing a complete BiP 5'NTR (nucleotides –225 to –1) in the intercistronic region (RNA I in Fig. 1B) after heat treatment at 42°C for 5 and 15 h. With this construct, translation of RLuc from the first cistron is directed by ribosomal scanning but translation of FLuc is directed by the IRES element located in the intercistronic region. The synthesis of FLuc driven by the BiP IRES increased gradually up to about threefold after 15 h of heat treatment (white columns of the BiP portion of Fig. 1A, part iv), while the synthesis of RLuc, which reflects the level of general translation, was reduced by heat treatment for 5 h and then recovered with longer heat treatment (black columns of the BiP portion of Fig. 1A, part iv). In contrast to the BiP IRES, the synthesis of FLuc driven by polioviral IRES was severely decreased after 5 h of heat treatment and then recovered by a longer heat treatment (15 h). Therefore, the heat responsiveness of the polioviral IRES is similar to that of general translation (white columns of the Polio portion of Fig. 1A, part iv). Together, these data indicate that the activation of BiP IRES-dependent translation is, at least in part, responsible for the accumulation of BiP protein after heat treatment for 15 h.

To identify a cis-acting element responding to heat stress, we investigated the effects of various deletion mutant forms of the BiP 5'NTR (Fig. 1B) in HeLa cells on IRES-dependent translation before (gray columns in Fig. 1C) and after (black columns in Fig. 1C) 15 h of heat treatment. The full-length 5'NTR of the BiP mRNA (nucleotides –225 to –1) increased the translation of FLuc by about fourfold with heat treatment (value above column I in Fig. 1C). We arbitrarily divided the 5'NTR of the BiP mRNA into three regions, designated A, B, and C, which are composed of nucleotides –225 to –171, –171 to –95, and –95 to –1, respectively, considering the position of the first nucleotide of the translational initiation codon (A) number 1. Under normal conditions, region C showed the strongest IRES activity (gray columns of part VI of Fig. 1C) and regions A and B showed very weak IRES activity (gray columns of parts IV and V of Fig. 1C). Indeed, the IRES activity of region C was as strong as that of the full-length 5'NTR of BiP (compare the gray columns of parts I and VI of Fig. 1C). This is consistent with a previous report suggesting that this region contains the minimal IRES element of the BiP mRNA (42). Interestingly, the extent to which the IRES activity was increased with heat treatment, depicted by the values above the columns in Fig. 1C, varied among the regions tested. The IRES activities of regions B and C increased by factors of 2.2 and 2.8, respectively (parts V and VI of Fig. 1C); however, the IRES activity of region A increased only slightly, by a factor of 1.3 (part IV of Fig. 1C). Heat-induced IRES activity was shown to increase by about fourfold or more when regions B and C were both present (parts I and II of Fig. 1C). The heat-induced increase in activity in regions B and C was greater than that of the full-length 5'NTR of BiP mRNA because the basal IRES activity of the full-length NTR was higher. After heat treatment, the IRES activity of the full-length 5'NTR of BiP mRNA is greater than that of regions B and C (compare parts I and II of Fig. 1C). Only some enhancement of the heat-induced increase in IRES activity with heat was observed when region A was added to region B (compare parts III and V of Fig. 1C). These results indicate that heat-inducible IRES elements are mainly located in regions B and C.

NSAP1 interacts with the IRES of BiP mRNA. To understand the molecular basis of IRES-dependent translation of BiP mRNA, we sought to identify the cellular factors that interact with the BiP IRES element. We performed a UV cross-linking experiment with a cytoplasmic extract of HeLa S3 cells and a 32P-labeled RNA corresponding to the BiP IRES. Cellular proteins with apparent molecular masses of 50 kDa, 65 kDa, and 110 kDa were found to bind to the BiP IRES (Fig. 2A, lane 2). The 50-kDa protein had previously been identified as autoantigen La (19). We focused our investigation on the 65-kDa protein, one of the proteins cross-linked to BiP IRES RNA in the greatest concentrations, because this seemed to share several characteristics of the NSAP1 protein that functions as an ITAF for HCV IRES (18). First, the 65-kDa protein showed the same mobility as NSAP1 (Fig. 2A, indicated by an arrow in lane 1). Second, the 65-kDa protein showed high affinities to homopolymeric poly(A) and poly(U), the same as NSAP1. In competition experiments with homopolymeric RNA, the band intensity of the 65-kDa protein was decreased in the presence of poly(A) and poly(U) RNAs (lanes 3 and 5 in Fig. 2A). Third, the band intensity of the 65-kDa protein was reduced in the NSAP1 knockdown sample, as indicated by an arrow (compare lanes 1 and 2 in Fig. 2B). The identity of the 65-kDa protein was further confirmed by immunoprecipitation of the UV cross-linked proteins by 32P-labeled BiP IRES RNA. The 65-kDa protein was precipitated with a specific polyclonal antibody against NSAP1 (Fig. 2B, lane 3), and furthermore, the band intensity of the NSAP1 knockdown sample was dramatically decreased (compare lanes 3 and 4 in Fig. 2B), which is consistent with the band patterns of both the 65-kDa proteins revealed by UV cross-linking (compare lanes 3 and 4 with lanes 1 and 2 in Fig. 2B) and NSAP1 detected by Western blotting of input samples (compare lanes 3 and 4 in Fig. 2B with lanes 1 and 2 in Fig. 2C). Interestingly, a few other proteins also showed altered binding patterns between mock-treated and siRNA-treated samples (lanes 1 and 2 in Fig. 2B). This phenomenon presumably originated from the altered expression or altered affinity of RNA-binding proteins caused by the reduction of NSAP1. The possible role of these proteins in BiP IRES-dependent translation and their relationship with NSAP1 remain to be elucidated. By contrast, no specific band was detected when a negative control antibody (polyclonal antibody against GFP) was used (Fig. 2B, lanes 5 and 6).


Figure 2
View larger version (56K):
[in this window]
[in a new window]

  		FIG. 2. Identification of NSAP1 as a cellular protein interacting with the human BiP IRES. (A) UV cross-linking experiments were performed with S10 extracts of HeLa S3 cells (20 µg each) and 32P-labeled RNAs (3 x 105 cpm each) corresponding to nucleotides 18 to 402 of the HCV IRES (lane 1) and nucleotides –225 to +3 of the BiP IRES (lanes 2 to 7). The adenine in the initiation codon of the BiP mRNA was designated nucleotide 1, and the upstream sequences are denoted by minus signs. Competition experiments were carried out with various competitor RNAs, including 50 ng homopolymeric RNAs [poly(A), poly(C), and poly(U)] (lanes 3 to 5) and a 5- or 50-fold molar excess of unlabeled BiP IRES RNA (lanes 6 and 7). After cross-linking reactions, samples were treated with an RNase cocktail and then analyzed by 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. The arrow indicates a 65-kDa protein. (B) Immunoprecipitation (IP) of UV cross-linked proteins with the 32P-labeled BiP RNA probe. After UV cross-linking with the cytoplasmic extracts of normal HeLa cells and NSAP1 knockdown cells, samples were precleared with protein G-agarose resin and then reacted with 2 µg of polyclonal anti-NSAP1 antibody (lanes 3 and 4) or an anti-GFP antibody (lanes 5 and 6). Resin-bound proteins were resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Lanes 1 and 2 show the labeled proteins in normal HeLa cells and NSAP1 knockdown cells, respectively, prior to immunoprecipitation. Lanes 3 and 4 show immunoprecipitated NSAP1. (C) NSAP1 and actin in the lysates used for UV cross-linking were monitored by Western blotting with corresponding antibodies. (D) Determination of the NSAP1 binding site on BiP IRES. UV cross-linking experiments were carried out with purified NSAP1 (100 ng) and 32P-labeled RNAs (3 x 105 cpm each) corresponding to various regions of the BiP IRES shown in Fig. 1B (probes I, IV, V, and VI).

 
A direct interaction between NSAP1 and BiP IRES RNA was further confirmed by a UV cross-linking experiment with purified NSAP1 protein and 32P-labeled BiP RNA containing nucleotides –225 to +3 of the BiP mRNA (Fig. 2D, lane 1). Together, these data indicate that NSAP1 directly binds to the BiP IRES.

In order to map the site at which NSAP1 binds the BiP IRES, we performed a UV cross-linking experiment with purified NSAP1 protein and 32P-labeled RNAs corresponding to different portions of the BiP IRES (Fig. 1B). NSAP1 interacted strongly with RNAs V and VI and weakly with RNA IV (Fig. 2D), but each RNA (V or VI) did not show the full binding activity, which is well correlated with the finding that the heat-responsive element resides in regions B and C (Fig. 1C). These data indicate that NSAP1 strongly binds to regions B and C of the BiP IRES.

NSAP1 enhances IRES-dependent translation of the BiP mRNA. NSAP1 has been shown to function as an ITAF for the HCV IRES (18). Therefore, we investigated whether NSAP1 increases IRES-dependent translation of BiP mRNA in mammalian cells by conducting a series of cotransfection experiments with reporter plasmids producing the dicistronic mRNAs R/BiP/F, R/Apaf/F, R/Polio/F, and R/HCV/F (Fig. 3A) and an effector plasmid producing NSAP1 fused with GFP. For the transfection experiments, we used 293T cells with high transfection efficiency. Overexpression of GFP-NSAP1 in 293T cells increased the IRES activity of the BiP mRNA by about 3.5-fold (lanes 1 and 2 in Fig. 3B, part i), which was similar to the effect of NSAP1 on the HCV IRES (compare lanes 2 and 8 in Fig. 3B, part i); however, the IRES activities of the cellular Apaf mRNA and the polioviral mRNA were not affected by the expression of NSAP1 (lanes 3, 4, 5, and 6 in Fig. 3B, part i). To exclude the possibility that the increased FLuc expression seen is due to a putative alteration of the integrity and/or stability of the reporter mRNA, we monitored the dicistronic reporter mRNA R/BiP/F by Northern blot analysis with a 32P-labeled probe corresponding to the FLuc gene. Overexpression of GFP-NSAP1 had no effect on the amount or integrity of the dicistronic mRNA (Fig. 3B, part ii). No mRNA of a monocistronic mRNA size which might be produced by a putative cryptic promoter or cryptic splicing site was detected. Similar amounts of GFP and GFP-NSAP1 proteins were expressed in the transfected cells (data not shown). The effects of other well-known ITAFs (such as hnRNP L, hnRNP C1, HuR, and PCBP1) on the function of the BiP IRES were examined by cotransfection with dicistronic reporter plasmid pR/BiP/F and effector plasmids producing GFP-fused ITAFs. Only NSAP1 had a significant effect on BiP IRES activity (upper panel of Fig. 3C), even though the various effector proteins were shown to be expressed at detectable levels in the transfected cells (lower panel of Fig. 3C). Taken together, these results indicate that NSAP1 specifically enhances IRES-dependent translation of BiP mRNA in vivo.


Figure 3
View larger version (20K):
[in this window]
[in a new window]

  		FIG. 3. NSAP1 enhances the IRES activity of BiP mRNA in vivo. (A) Schematic diagram of dicistronic mRNAs used for monitoring the effect of NSAP1 on various IRES activities in vivo. The dicistronic mRNAs containing the BiP, Apaf, poliovirus, and HCV IRES regions are depicted as R/BiP/F, R/Apaf/F, R/Polio/F, and R/HCV/F, respectively. (B) (i) 293T cells were cotransfected with a reporter plasmid (pR/BiP/F, pR/Apaf/F, pR/Polio/F, or pR/HCV/F), an effector plasmid expressing GFP or GFP-NSAP1, and the control plasmid pCMV•SPORT-ßgal. Forty-eight hours posttransfection, cells were harvested and luciferase activities were measured. FLuc and RLuc activities were normalized with ß-galactosidase activity to adjust for transfection efficiency. Black and white columns depict RLuc and FLuc activities, respectively. Luciferase activities in cells expressing GFP were set to 1 (lanes 1, 3, 5, and 7). Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars. (ii) Northern blot analysis of the reporter dicistronic mRNA R/BiP/F produced in transfected cells. Three micrograms of poly(A)+ RNA purified from transfected cells was subjected to Northern blotting with a 32P-labeled probe corresponding to the FLuc gene. The positions of the 28S and 18S rRNAs are indicated. The human ribosomal protein large subunit 32 (hRPL32) blot was used as an internal control for poly(A)+ mRNAs. The arrow indicates the position of the reporter mRNA. (C) 293T cells were cotransfected with reporter plasmid pR/BiP/F and an effector plasmid expressing GFP (lane GFP) or expressing ITAFs fused with GFP. Forty-eight hours posttransfection, relative luciferase activities were determined. The ratio of FLuc activity to RLuc activity in cells transfected with effector GFP was set to 1 (upper panel). Lysates were analyzed by Western blotting with a monoclonal anti-GFP antibody (lower panel). The identities of the ITAFs are shown at the bottom. Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars.

 
NSAP1 activates BiP IRES-dependent translation mainly through the region proximal to the start codon (AUG). To further characterize the effect of NSAP1 on the IRES element of the BiP 5'NTR, we monitored the IRES activities of BiP NTR derivatives in 293T cells in which NSAP1 was overexpressed. 293T cells were cotransfected with the effector plasmid pGFP-NSAP1 and the dicistronic reporter plasmids with deletion mutations in the region corresponding to the 5'NTR of the BiP mRNA (Fig. 1B) in the intercistronic region. All of the deletion mutants containing region C (RNAs II and VI in Fig. 1B) showed about 80% or more IRES activity compared with RNA I (which contained a complete 5'NTR, nucleotides –225 to –1) (gray columns in Fig. 4). Deletion of this region (region C) reduced IRES activity by about 40% (gray column in part III). Further deletions (RNAs IV and V) reduced IRES activity by up to 60% (gray columns in parts IV and V). It is noteworthy that the basal IRES activity of region C was 14% less than that of the full-length BiP 5'NTR in 293T cells (gray columns in parts I and VI of Fig. 4). However, the same deletion mutant (RNA VI) had the same basal IRES activity as the full-length IRES in HeLa cells (Fig. 1C). This discrepancy is likely due to the difference in the constituents (ITAFs) in the cells tested. These data and the data from HeLa cells (Fig. 1) indicate that region C (nucleotides –95 to –1) plays a crucial role in determining the basal level of IRES activity and that regions A and B contribute partially to the basal IRES activity.


Figure 4
View larger version (21K):
[in this window]
[in a new window]

  		FIG. 4. Determination of NSAP1-responsive elements in the 5'NTR of the BiP mRNA. 293T cells were cotransfected with reporter plasmids expressing one of the dicistronic mRNAs shown in Fig. 1B and an effector plasmid expressing GFP or GFP-NSAP1. Forty-eight hours posttransfection, cells were harvested and luciferase activities were measured. The ratios of FLuc and RLuc activities were calculated, and the mean values are shown in boxes below the graph. Gray and black columns depict the relative luciferase activities in cells transfected with GFP and GFP-NSAP1, respectively. The fold increases were also calculated by the ratio of GFP-NSAP1 to GFP and shown by numbers above each column. The ratio of luciferase activities in cells producing the dicistronic mRNA (RNA I in Fig. 1B) with the entire BiP 5'NTR (nucleotides –225 to –1) and GFP was set to 1. Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars.

 
Different regions of the 5'NTR of BiP mRNA responded differently to the overexpression of NSAP1. Overexpression of NSAP1 increased the activity of the full-length BiP IRES (RNA I including nucleotides –225 to –1) by 3.4-fold (see the value above column I in Fig. 4). All of the constructs containing nucleotides –95 to –1 of BiP mRNA (RNAs I, II, and VI) showed an increase in IRES activity (at least threefold) with GFP-NSAP1 expression (columns I, II, and VI in Fig. 4). The presence of the middle region (nucleotides –171 to –95) and the 5' end (nucleotides –225 to –171) of the BiP 5'NTR also affected the response to NSAP1 (twofold or more, columns V and IV in Fig. 4). These data indicate that NSAP1 activates BiP IRES-dependent translation mainly through the region proximal to the start codon (AUG) and to a lesser extent through the other regions.

Knockdown of NSAP1 by siRNA inhibits IRES-dependent translation of the BiP mRNA. We further analyzed the effect of NSAP1 on BiP IRES activity by applying an NSAP1-specific siRNA to the cell to assess the effects of reduced levels of cellular NSAP1. We constructed HeLa cell clones that constitutively expressed siRNA against NSAP1 corresponding to nucleotides 734 to 752 of NSAP1 mRNA (Fig. 5A). To exclude the experimental bias that can result from the use of a single clone, we pooled hygromycin-resistant cell colonies from selection plates, and the mixed cell lines were used for further analyses. Cells expressing NSAP1-specific siRNA had markedly reduced NSAP1 levels compared with cell clones containing the negative control plasmid (NSAP1 in Fig. 5B). The levels of other proteins, such as the well-known ITAFs HuR and hnRNP L and the negative-control protein actin, were not affected by the presence of the NSAP1-specific siRNA (HuR, hnRNP L, and actin in Fig. 5B). The effects of the NSAP1-specific siRNAs on the NSAP1 and BiP mRNAs were monitored by Northern blot analysis. The level of NSAP1 mRNA was markedly reduced by expression of the NSAP1-specific siRNAs (NSAP1 mRNA in Fig. 5C). By contrast, the levels of BiP mRNA and hRPL32 mRNA (a negative control) were not affected by the siRNAs.


Figure 5
View larger version (29K):
[in this window]
[in a new window]

  		FIG. 5. Effect of NSAP1-specific siRNA on BiP IRES activity. (A) Schematic diagram of siRNA-expressing plasmid pEBV-U6+27 (top) and the predicted secondary structure of siRNAs generated from pEBV-U6+27/NSAP1(734-752) (bottom). (B) Western blot analysis of HeLa cells expressing the siRNA siNSAP1. Stably transformed HeLa cells were generated by transfection of the control (Con) vector and vectors encoding siNSAP1. Cells were harvested, and protein levels were analyzed by immunoblotting with anti-NSAP1, anti-HuR, anti-hnRNP L, anti-actin, and anti-BiP antibodies. Arrows indicate two isoforms of NSAP1, and the uppermost band represents hnRNP R (NSAP1 blot) (C) Northern blot analysis of HeLa cells expressing siRNA against NSAP1. Thirty micrograms of total RNA was resolved on a denaturing gel and immobilized on a nylon membrane. The filter was hybridized consecutively with 32P-labeled probes specific for the BiP, NSAP1, and hRPL32 mRNAs. The arrowheads indicate isoforms of NSAP1 transcripts. (D) The effect of siRNA on BiP IRES function was monitored by transfection of the R/BiP/F dicistronic RNA (shown in Fig. 3A) into the established cell lines. Relative IRES activities were measured 3 h posttransfection. The value of siRNA-lacking control cells was set to 1. Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars. (E, part i) The effect of siRNA on BiP IRES activity was also monitored by transfection of DNA expressing dicistronic mRNA R/BiP/F (shown in Fig. 3A). Cells with or without siRNA were transfected with plasmid pR/BiP/F, and IRES activity was measured by FLuc activities 48 h posttransfection. DNA transfection efficiency was normalized by RLuc activity directed by cap-dependent translation, and ratios of FLuc activity to RLuc activity were calculated. The value of the control cells was set to 1. Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars. (ii) Northern blot analysis of dicistronic mRNAs produced from HeLa cells expressing siRNA against NSAP1. Three micrograms of poly(A)+ RNAs prepared from transfected cells was subjected to Northern blotting with a 32P-labeled probe corresponding to the FLuc gene. The positions of the 28S and 18S rRNAs are indicated. The hRPL32 blot was used as an internal control for poly(A)+ mRNAs. The arrow indicates the position of the reporter mRNA.

 
The effect of siRNA-induced NSAP1 protein knockdown on BiP IRES activity was assessed by transfection of an in vitro-transcribed dicistronic mRNA (Fig. 5D) or the plasmid pR/BiP/F (Fig. 5E). The RNA transfection method was used to exclude the possibility of a putative monocistronic mRNA being produced from a putative cryptic promoter or a cryptic splicing acceptor in the 5'NTR of the BiP mRNA. Both the DNA and RNA transfection experiments showed that a reduction in the level of NSAP1 by siRNA resulted in reduced IRES activity of the BiP 5'NTR. The amount and integrity of the dicistronic reporter mRNA in the cells containing siRNAs were monitored by Northern blot analysis with a 32P-labeled probe corresponding to the FLuc gene. NSAP1 knockdown affected neither the amount nor the integrity of reporter mRNAs (compare lanes 2 and 3 in Fig. 5E, part ii). Consistent with the effect of NSAP1 on activation of translation from the BiP IRES, the level of endogenous BiP protein decreased with the level of NSAP1 (BiP and NSAP1 in Fig. 5B). The reduced level of endogenous BiP protein is most likely due to the reduced BiP IRES activity. Taken together, these data indicate that NSAP1 increases the translation of endogenous BiP via an IRES element in the BiP 5'NTR.

NSAP1 plays an important role in the activation of BiP IRES-dependent translation under heat stress conditions. To determine whether NSAP1 plays an important role in the activation of BiP IRES function under heat stress conditions, we investigated BiP IRES activation in cells expressing NSAP1-specific siRNA. Twenty-four hours after transfection by electroporation of dicistronic plasmid pR/BiP/F into HeLa cells expressing the NSAP1-specific siRNA, some cells were transferred to a 42°C incubator while others were maintained at 37°C. After exposure to continuous heat stress for 15 h, cells were harvested and RLuc and FLuc activities were measured. The BiP IRES activity in control cells increased by 2.5-fold after heat treatment (lanes 1 and 2 in Fig. 6A, part i). By contrast, the BiP IRES activity in NSAP1 knockdown cells was not increased by heat stress (lanes 3 and 4 in Fig. 6A, part i). During the same experiment, NSAP1 and BiP protein levels were also monitored by immunoblotting (Fig. 6A, part ii). The level of BiP protein in control cells increased markedly with heat stress (compare lane 1 and lane 2 in Fig. 6A, part ii), but the level in NSAP1 knockdown cells did not show any significant change under the same conditions (compare lanes 3 and 4 in Fig. 6A, part ii). This strongly suggests that NSAP1 plays an important role in heat-dependent activation of BiP IRES function. To rule out the possibility that the change in the BiP protein level is due to a change in the BiP mRNA level, we monitored the BiP mRNA level in the same experimental set by using radiolabeling RT-PCR. The BiP mRNA level was significantly reduced by heat treatment for 15 h (compare lane 1 with lane 2 in Fig. 6A, part iii), as previously shown in Fig. 1A, part ii. The same change in the BiP mRNA level was observed in the cells treated with heat and siRNA against NSAP1 (compare lanes 3 and 4 with lanes 1 and 2 in Fig. 6A, part iii), which indicates that the BiP mRNA level is reduced by heat treatment but not by the amount of NSAP1. The level of GAPDH mRNA, a negative control, was affected by neither heat nor the siRNA against NSAP1. Considering the reduced amount of BiP mRNA after heat treatment, the increased amount of BiP protein must be attributable to augmentation of gene expression at the posttranscriptional level(s), including translation.


Figure 6
Figure 6
View larger version (47K):
[in this window]
[in a new window]

  		FIG. 6. The role of NSAP1 in heat-dependent translational activation of BiP mRNA. (A) The IRES activity of BiP mRNA was monitored in cells expressing NSAP1-specific siRNA under heat stress conditions. HeLa cells expressing NSAP1-specific siRNA (siNSAP1) and control (Con) cells were transfected with plasmid pR/BiP/F (3 µg each). Twenty-four hours posttransfection, some of the transfected cells were transferred to an incubator preheated to 42°C (black columns, lanes 2 and 4), while the rest of the transfected cells were maintained at 37°C (gray columns, lanes 1 and 3). After cultivation for 15 h, cells were harvested and FLuc and RLuc activities were measured. The relative luciferase activity in the cell lysate of control cells without siRNA that were incubated at 37°C was set to 1. Experiments were performed at least three times for each experimental set, and standard deviations are shown as error bars. (ii) BiP, NSAP1, and actin protein levels in each cell lysate were monitored by Western blotting with anti-BiP, anti-NSAP1, and anti-actin antibodies. (iii) The levels of BiP mRNAs under heat stress and/or NSAP1 knockdown conditions. Total RNAs were prepared from HeLa cells and then subjected to radiolabeling RT-PCR with BiP-specific primers. GAPDH mRNA was used as a negative control message. (B) Distributions of mRNAs in heat-stressed and/or NSAP1 knockdown cells. HeLa cells were infected with recombinant adenoviruses expressing mutant siRNA (parts iii and iv) or siRNA against NSAP1 (parts v and vi) and then cultivated at 37°C for 24 h. Thereafter, some cells were transferred to an incubator preheated to 42°C and maintained for 15 h (parts ii, iv, and vi) or cultivated at 37°C continuously (parts i, iii, and v). After heat treatment, the cells were treated with cycloheximide (100 µg/ml) for 5 min at 37°C and then harvested. Extracts from heat-treated and/or NSAP1 knockdown HeLa cells were subjected to sucrose gradient centrifugation and then divided into the following six fractions: proteins not associated with ribosomes (fraction 1, unbound proteins), RNAs and proteins at the 40S ribosomal fraction (fraction 2, 40S), RNAs and proteins at the 60S ribosomal fraction (fraction 3, 60S), RNAs and proteins at the monosomal fraction (fraction 4, 80S), RNAs and proteins at the LPs (fraction 5, LP), and RNAs and proteins at the HPs (fraction 6, HP). The distribution of BiP and GAPDH mRNAs across the gradients was analyzed by radiolabeling RT-PCR. In addition, the NSAP1 and S6 proteins were also monitored by Western blotting with corresponding antibodies. Distinct strong peaks appeared right after the heavy polysomal fractions in the heat-treated samples (denoted by asterisks in parts ii, iv, and vi). These rapidly sedimenting entities might be a component of stress granules formed in heat-treated cells. (C) The band intensities of BiP and GAPDH mRNAs shown in panel B were measured with the L process and Image Gauge programs (Fuji Photo Film Co., Ltd.), and the relative proportion of each band was calculated and is depicted in the graphs. The proportions of fractions HP, LP, monosome (80S), and the sum of fractions 40S plus 60S plus unbound proteins are depicted by white, black, light gray, and dark gray bars, respectively.

 
The role of NSAP1 in BiP mRNA translation was also investigated by analyzing the distribution patterns of BiP mRNA and a control GAPDH mRNA in ribosome profiles with or without knockdown of NSAP1 by siRNA before and after heat treatment. HeLa cell extracts were subjected to sucrose gradient centrifugation and then divided into six fractions, i.e., unbound proteins, 40S, 60S, 80S, LP, and HP (fractions 1 to 6, respectively, in Fig. 6B). The identity of each fraction was confirmed by Western blotting of a component of 40S ribosomal subunit S6 protein that is abundant in the 40S fraction (S6 in Fig. 6B, part i). The amount of this protein increased at the 80S fraction in the heat-treated cells (S6 in Fig. 6B, part ii) because of the increased amount of 80S ribosomal complex composed of 40S and 60S ribosomal subunits, which is reflected in the increased optical density at 254 nm of the 80S peak (optical absorption profile in Fig. 6B, part ii). Most of the NSAP1 resided in the unbound fraction, as shown by Western blotting (lane 1 of the NSAP1 portions of Fig. 6B, parts i and ii). Some NSAP1 was observed in the 40S ribosomal subunit fraction (lane 2 of the NSAP1 portions of Fig. 6B, parts i and ii).

BiP and GAPDH mRNAs were semiquantified by the radiolabeling RT-PCR method with RNAs purified from each fraction (BiP mRNA and GAPDH mRNA portions of Fig. 6B). Under normal conditions, both the BiP and GAPDH mRNAs were enriched in the heavy polysomal fraction (lane 6 of the BiP mRNA and GAPDH mRNA portions of Fig. 6B, part i). As expected, a dramatic shift in GAPDH mRNA, which is not a heat shock mRNA, from the HP to the monosome fraction reflecting a reduction in translation was observed upon heat treatment (compare lane 4 of the GAPDH mRNA portion of Fig. 6B, part ii, with that in part i). The reduction of translation of general mRNAs during heat stress is also reflected in the ribosomal profile showing a reduction of the polysomal peaks and an increase in the monosomal peak (optical absorption profiles in Fig. 6B, parts i and ii). On the other hand, a shift of BiP mRNA, which is a heat shock mRNA, from the monosome to the HP fraction, reflecting an increase in translation, was observed upon heat treatment (compare lane 4 of the BiP mRNA portion of Fig. 6B, part ii, with that in part i).

In order to investigate the role of NSAP1 in BiP mRNA translation, we infected HeLa cells with a recombinant adenovirus expressing an siNSAP1 or a control adenovirus expressing an siRNA containing mutations in the siNSAP1 (control siRNA). Upon infection of the control adenovirus, no obvious change in the ribosome profile or the distribution patterns of BiP and GAPDH mRNAs was observed before and after heat treatment compared with uninfected cells (compare parts i and ii with parts iii and iv of Fig. 6B). The distribution pattern of GAPDH mRNA was not changed by knockdown of NSAP1 before and after heat treatment (compare the GAPDH mRNA portions of Fig. 6B, parts iii and v, with those of parts v and vi). On the other hand, a dramatic shift of BiP mRNAs from the HP fraction to the LP fraction reflecting a reduction in BiP mRNA translation was observed in the NSAP1 knockdown cells (compare lane 5 of the BiP mRNA portion of Fig. 6B, part v, with that in parts i and iii). The distribution pattern of BiP mRNA in the NSAP1 knockdown cells was not changed by heat treatment. The relative distributions of the BiP and GAPDH mRNAs in different fractions under various cellular conditions are depicted in Fig. 6C. The distribution patterns of BiP mRNA strongly suggest that NSAP1 plays an important role in the heat-dependent translation of BiP mRNA.

Binding of NSAP1 to the BiP IRES increases under heat stress conditions. To examine whether the amount of NSAP1 bound to the BiP IRES increases under heat stress conditions, we performed UV cross-linking experiments with 32P-labeled BiP IRES RNA and cytoplasmic extracts from mock-treated or heat-stressed HeLa cells. The binding of some proteins (35-kDa, 40-kDa, and 65-kDa proteins, indicated by arrows in Fig. 7A, part i) greatly increased with heat treatment (compare lanes 1 and 2 in Fig. 7A, part i), whereas the binding of other proteins (110 kDa and 50 kDa) did not change with heat treatment. The 65-kDa protein was confirmed to be NSAP1 by immunoprecipitation of the UV cross-linked proteins with a polyclonal antibody against NSAP1 (lanes 3 and 4 in Fig. 7A, part i). The intensity of the immunoprecipitated band from heat-treated cell extract was much greater than that from the mock-treated cell extract, which is consistent with the band pattern of the 65-kDa protein revealed by UV cross-linking (compare lanes 3 and 4 with lanes 1 and 2). By contrast, no band was detected when the negative control antibody (polyclonal antibody against GFP) was used (lanes 5 and 6). To determine whether NSAP1 accumulated during heat treatment, the level of NSAP1 was monitored by Western blotting. No significant change, unless there was a decrease, in the NSAP1 protein level was observed in heat-treated cells (lanes M and HS in Fig. 7A, part ii). Moreover, enrichment of NSAP1 in the cytoplasm of heat-treated cells was not observed when we investigated the distribution of NSAP1 in cells by both fractionating cellular components into nuclear and cytoplasmic parts and immunostaining endogenous NSAP1 (see Fig. S1 in the supplemental material). Taken together, these data indicate that increased binding of NSAP1 to the BiP IRES plays an important role in the activation of the BiP IRES under heat stress conditions.


Figure 7
View larger version (24K):
[in this window]
[in a new window]

  		FIG. 7. Affinity of NSAP1 binding to the BiP IRES and BiP IRES activity. (A) Binding of NSAP1 protein to the BiP IRES before and after heat treatment. (i) UV cross-linking experiments were performed with 32P-labeled BiP IRES RNA and cytoplasmic extract of HeLa cells with (lanes 2, 4, and 6) and without (lanes 1, 3, and 5) heat treatment for 15 h. After UV cross-linking and RNase treatment, samples were precleared with protein G-agarose resin and then reacted with 2 µg of polyclonal anti-NSAP1 antibody (lanes 3 and 4) or anti-GFP antibody (lanes 5 and 6). Resin-bound proteins were resolved by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Lanes 1 and 2 show the labeled proteins prior to immunoprecipitation (IP). The arrows depict the NSAP1, 40-kDa, and 35-kDa proteins. (ii) NSAP1 and actin in the lysates were identified by Western blotting with corresponding antibodies. M and HS represent cell lysates without and with heat treatment, respectively. (B) Heat-induced enhancement of NSAP1 binding to various regions of the 5'NTR of the BiP mRNA. Extracts of HeLa cells were prepared without (–; lanes 1, 3, 5, and 7) or with (+; lanes 2, 4, 6, and 8) heat treatment for 15 h and subjected to an in vitro RNA-binding assay with biotinylated RNA probes corresponding to different regions of the BiP IRES (II, V, and VI in Fig. 1B). NSAP1 protein bound to each RNA probe was monitored by Western blotting with a polyclonal NSAP1-specific antibody. Negative-control experiments were performed without an RNA probe (lanes 1 and 2). (C) The increase in IRES activity with heat treatment correlates well with the increased binding of NSAP1 to the IRES element. The intensities of NSAP1 bands in panel B were measured by the Sion Image Analysis Program, and the ratios of the band intensities before and after heat treatment were calculated and are depicted by black columns. Fold increases in the IRES activities of the RNAs caused by heat treatment are depicted by gray columns. Names of RNAs (II, V, and VI in Fig. 1C) are shown at the bottom.

 
The relationship between the level of activity of the BiP IRES and the amount of NSAP1 bound was further analyzed by using deletion mutant forms of the BiP IRES. Biotinylated RNAs corresponding to various regions of the BiP IRES that respond to heat treatment (RNAs II, V, and VI, corresponding to regions B-C, B, and C, respectively), were used to analyze NSAP1-binding affinity with mock-treated (lanes 1, 3, 5, and 7 in Fig. 7B) and heat-treated (lanes 2, 4, 6, and 8 in Fig. 7B) HeLa cell extracts. The fold increases in the IRES activities of these IRES elements caused by heat treatment (Fig. 1C) and the fold increases in the affinity of NSAP1 for these RNAs after heat treatment (Fig. 7B) are shown in Fig. 7C. The increases in IRES activity (gray columns in Fig. 7C) and the increases in NSAP1 binding affinity (black columns in Fig. 7C) are well correlated, indicating that NSAP1 plays an important role in heat-induced IRES activity under heat stress conditions. The molecular basis of the heat-induced binding of NSAP1 to the BiP IRES is under investigation.


   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
It has been shown that an IRES element in the 5'NTR of the BiP mRNA directs translation of the mRNA (42) and that the IRES activity of the BiP mRNA is increased by heat treatment (21). Therefore, it is likely that the IRES activity of the BiP mRNA varies with intracellular and extracellular changes that modulate the availability and/or activity of ITAFs. Although the PTB and La proteins have been shown to interact with the IRES element of the BiP mRNA (19, 20), no ITAF has been reported to modulate BiP IRES activity under stress conditions.

Here, we report that NSAP1, which has been shown to activate the HCV IRES element, increases BiP IRES activity through an RNA-protein interaction. Several lines of evidence indicate that NSAP1 increases BiP IRES function. First, NSAP1 specifically interacts with the BiP IRES element. This was shown by UV cross-linking of HeLa cell extracts (normal and with NSAP1 knockdown) with 32P-labeled BiP IRES RNA followed by immunoprecipitation with an anti-NSAP1 antibody (Fig. 2A and B) and is further supported by UV cross-linking of purified NSAP1 proteins with 32P-labeled BiP IRES RNA (Fig. 2D). Second, overexpression of NSAP1 increased the activity of a reporter under the control of the BiP IRES element in an artificial dicistronic mRNA in vivo (Fig. 3B). Third, siRNA-mediated knockdown of NSAP1 inhibited the IRES activity of BiP mRNA, as shown by transfection of RNA (Fig. 5D) and DNA (Fig. 5E) reporters. Fourth, heat-induced activation of the BiP IRES did not occur in cells with an NSAP1-specific siRNA (Fig. 6A). Fifth, a considerable amount of BiP mRNA was shifted from the HP to the LP in NSAP1 knockdown cells, indicating that translation of BiP mRNA is reduced in cells lacking NSAP1 (parts v and vi of Fig. 6B and C). Sixth, the level of BiP protein was well correlated with the level of NSAP1 in cells expressing NSAP1-specific siRNA (Fig. 5B).

How does NSAP1 enhance BiP protein expression through IRES-dependent translation under heat stress conditions? Some experimental data indicate that increased binding of NSAP1 to the BiP mRNA IRES element under heat stress conditions is crucial for an increase in BiP IRES activity. First, in UV cross-linking experiments with heat-stressed cell extracts and 32P-labeled BiP IRES RNA, binding of NSAP1 to IRES elements increased markedly even though the level of NSAP1 was not affected by heat treatment (Fig. 7A). Second, the binding of NSAP1 to the BiP IRES (Fig. 2D and 7B), heat-dependent activation of the BiP IRES (Fig. 1C), and NSAP1-dependent activation (Fig. 4) involve the same regions (B and C) of the BiP IRES. Third, increased binding of NSAP1 to biotinylated RNAs that correspond to different portions of the BiP IRES element is highly correlated with the changes in IRES activity (Fig. 7C).

How does the binding of NSAP1 to the BiP IRES increase under heat stress conditions? One possibility is that the conformation of NSAP1 and/or BiP RNA might change under heat stress conditions in a manner that facilitates binding. The putative conformational changes in NSAP1 and the BiP IRES are under investigation. A second possibility is that as-yet-unidentified cellular factors may facilitate the binding of NSAP1 to the BiP IRES. Interestingly, in addition to NSAP1, the binding of two other proteins to the BiP IRES was increased by heat treatment (lane 2 in Fig. 7A). The HuR (35 kDa) and hnRNP C (40 kDa) proteins, which are known to modulate other IRES elements (17, 22), were included in the group of proteins the binding of which to the BiP IRES was strengthened by heat (see Fig. S2 in the supplemental material). The effects of HuR and hnRNP C on the activity of the BiP IRES were investigated by cotransfection of plasmids expressing hnRNP C and/or HuR with the dicistronic reporter plasmid pR/BiP/F, with or without overexpression of NSAP1. Neither hnRNP C nor HuR affected BiP IRES activity (data not shown). Moreover, binding of NSAP1 to the BiP IRES was not affected by the addition of purified hnRNP C1 (data not shown). These data do not support a positive role for HuR or hnRNP C in NSAP1-dependent translation of BiP mRNA. However, we cannot exclude the possibilities that cellular factors such as hnRNP C and HuR indirectly facilitate the interaction of NSAP1 with the BiP IRES under heat stress conditions and that unidentified cellular factors with the same molecular masses as hnRNP C and HuR are involved in this process. A third possibility is that a putative posttranslational modification such as phosphorylation may modulate NSAP1 activity. It has recently been shown that NSAP1 is phosphorylated at tyrosine residues by insulin and osmotic stress and that this phosphorylation inhibits its RNA-binding ability, suggesting a possible role for NSAP1 phosphorylation in translational regulation (11). Many cellular proteins, including translational initiation factors, have been shown to be phosphorylated by a stress-induced kinase such as the c-Jun N-terminal kinase (JNK) under stress conditions (4, 32). Some ITAFs, such as PTB, hnRNP A, and hnRNP C, have been shown to be phosphorylated by kinases (30, 38, 41). NSAP1 may also be phosphorylated by a stress-activated kinase under heat stress conditions, and its binding to BiP IRES might be increased by heat-induced phosphorylation. However, a physiological role for the stress-induced phosphorylation of these proteins in translational regulation remains to be determined.

It is not clear how NSAP1 that is bound to the BiP IRES increases translation of the BiP mRNA; however, possible modes of action are outlined below. First, a putative NSAP1-ribosome interaction might facilitate the binding of ribosomes to the IRES. By sucrose gradient fractionation, we observed that NSAP1 cofractionated with the 40S ribosome rather than the 60S, 80S, and polysome fractions (NSAP1 in Fig. 6B). The cofractionation of NSAP1 with the 40S ribosomal subunit might indicate that NSAP1 forms a complex with the ribosomal subunit. Alternatively, NSAP1 binding could induce a conformational change in the IRES, resulting in a structure that is better recognized by the translational machinery including the 40S ribosomal subunit. Second, it is possible that NSAP1 recruits a canonical translation factor(s) through a putative protein-protein interaction. Third, NSAP1 may activate IRES-dependent translation through a putative interaction with another ITAF. In this respect, it is noteworthy that many ITAFs are known to interact with each other and form protein complexes (16).

Here we report the molecular basis of the heat-induced expression of the BiP protein by the binding of NSAP1 to the IRES element. This is a good example of translational regulation of protein expression by a trans-acting factor responding to an environmental change. Further investigation into the mechanism of heat-induced binding of NSAP1 to the BiP IRES and the molecular basis of the mechanism by which ribosomes are recruited to the IRES element by NSAP1 will provide a detailed understanding of the translational regulation of eukaryotic protein expression.


   ACKNOWLEDGMENTS
 
We thank A. Mizutani (RIKEN, Japan) and G. Dreyfuss (University of Pennsylvania) for providing the anti-NSAP1 antibody and anti-hnRNP L antibody, respectively.

The present study was supported in part by grant MCBRG (M10501000022-05-N0100-02200) from MOST, grants 02-PJ2-PG1-CH16-0002 and A050291 from KHIDI, grant SBD-NCRC (R15-2004-033-01001-0) from KRF, grant FPR05B 1-310 of the 21C Frontier Functional Proteomics Project from KMST, and a grant from POSCO.


   FOOTNOTES
 
* Corresponding author. Mailing address: PBC, Department of Life Science, Pohang University of Science and Technology, San 31, Hyoja Dong, Pohang, Kyungbuk 790-784, Republic of Korea. Phone: 82-54-279-2298. Fax: 82-54-279-8009. E-mail: sungkey{at}postech.ac.kr. Back

{dagger} Supplemental material for this article may be found at http://mcb.asm.org/. Back

{triangledown} Published ahead of print on 30 October 2006. Back


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
1. Back, S. H., Y. K. Kim, W. J. Kim, S. Cho, H. R. Oh, J.-E. Kim, and S. K. Jang.2002 . Translation of polioviral mRNA is inhibited by cleavage of polypyrimidine tract-binding proteins executed by polioviral 3C9pro. J. Virol. 76:2529-2542.[Abstract/Free Full Text]

1. Bannai, H., K. Fukatsu, A. Mizutani, T. Natsume, S. I. Iemura, T. Ikegami, T. Inoue, and K. Mikoshiba. 2004. An RNA-interacting protein, SYNCRIP (hnRNP Q1/NSAP1) is a component of mRNA granule transported with inositol 1,4,5-trisphosphate receptor type 1 mRNA in neuronal dendrites. J. Biol. Chem. 279:53427-53434.[Abstract/Free Full Text]

2. Bole, D. G., L. M. Hendershot, and J. F. Kearney. 1986. Posttranslational association of immunoglobulin heavy chain binding protein with nascent heavy chains in nonsecreting and secreting hybridomas. J. Cell Biol. 102:1558-1566.[Abstract/Free Full Text]

3. Duncan, R. F. 1996. Translational control during heat shock, p. 271-293. In J. W. B. Hershey, M. B. Mathews, and N. Sonenberg (ed.), Translational control. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.

4. Gabai, V. L., and M. Y. Sherman. 2002. Invited review: interplay between molecular chaperones and signaling pathways in survival of heat shock. J. Appl. Physiol. 92:1743-1748.[Abstract/Free Full Text]

5. Grosset, C., C. Y. Chen, N. Xu, N. Sonenberg, H. Jacquemin-Sablon, and A. B. Shyu. 2000. A mechanism for translationally coupled mRNA turnover: interaction between the poly(A) tail and a c-fos RNA coding determinant via a protein complex.Cell 103:29-40.[CrossRef][Medline]

6. Haas, I. G., and M. Wabl. 1983. Immunoglobulin heavy chain binding protein. Nature 306:387-389.[CrossRef][Medline]

7. Harris, C. E., R. A. Boden, and C. R. Astell.1999 . A novel heterogeneous nuclear ribonucleoprotein-like protein interacts with NS1 of the minute virus of mice.J. Virol. 73:72-80.[Abstract/Free Full Text]

8. Hassfeld, W., E. K. Chan, D. A. Mathison, D. Portman, G. Dreyfuss, G. Steiner, and E. M. Tan. 1998. Molecular definition of heterogeneous nuclear ribonucleoprotein R (hnRNP R) using autoimmune antibody: immunological relationship with hnRNP P. Nucleic Acids Res. 26:439-445.[Abstract/Free Full Text]

9. Hellen, C. U., and P. Sarnow. 2001. Internal ribosome entry sites in eukaryotic mRNA molecules. Genes Dev. 15:1593-1612.[Free Full Text]

10. Hresko, R. C., and M. Mueckler. 2000. A novel 68-kDa adipocyte protein phosphorylated on tyrosine in response to insulin and osmotic shock. J. Biol. Chem. 275:18114-18120.[Abstract/Free Full Text]

11. Hresko, R. C., and M. Mueckler. 2002. Identification of pp68 as the tyrosine-phosphorylated form of SYNCRIP/NSAP1. A cytoplasmic RNA-binding protein. J. Biol. Chem. 277:25233-25238.[Abstract/Free Full Text]

12. Jang, S. K. 2006. Internal initiation: IRES elements of picornaviruses and hepatitis C virus. Virus Res. 119:2-15.[CrossRef][Medline]

13. Jang, S. K., H. G. Kräusslich, M. J. Nicklin, G. M. Duke, A. C. Palmenberg, and E. Wimmer. 1988. A segment of the 5' nontranslated region of encephalomyocarditis virus RNA directs internal entry of ribosomes during in vitro translation. J. Virol. 62:2636-2643.[Abstract/Free Full Text]

14. Kanai, Y., N. Dohmae, and N. Hirokawa. 2004. Kinesin transports RNA: isolation and characterization of an RNA-transporting granule. Neuron 43:513-525.[CrossRef][Medline]

15. Kedersha, N., and P. Anderson. 2002. Stress granules: sites of mRNA triage that regulate mRNA stability and translatability.Biochem. Soc. Trans. 30:963-969.[CrossRef][Medline]

16. Kim, J. H., B. Hahm, Y. K. Kim, M. Choi, and S. K. Jang. 2000. Protein-protein interaction among hnRNPs shuttling between nucleus and cytoplasm.J. Mol. Biol. 298:395-405.[CrossRef][Medline]

17. Kim, J. H., K. Y. Paek, K. Choi, T. D. Kim, B. Hahm, K. T. Kim, and S. K. Jang.2003 . Heterogeneous nuclear ribonucleoprotein C modulates translation of c-myc mRNA in a cell cycle phase-dependent manner. Mol. Cell. Biol. 23:708-720.[Abstract/Free Full Text]

18. Kim, J. H., K. Y. Paek, S. H. Ha, S. Cho, K. Choi, C. S. Kim, S. H. Ryu, and S. K. Jang. 2004. A cellular RNA-binding protein enhances internal ribosomal entry site-dependent translation through an interaction downstream of the hepatitis C virus polyprotein initiation codon. Mol. Cell. Biol. 24:7878-7890.[Abstract/Free Full Text]

19. Kim, Y. K., S. H. Back, J. Rho, S. H. Lee, and S. K. Jang. 2001. La autoantigen enhances translation of BiP mRNA. Nucleic Acids Res. 29:5009-5016.[Abstract/Free Full Text]

20. Kim, Y. K., B. Hahm, and S. K. Jang.2000 . Polypyrimidine tract-binding protein inhibits translation of bip mRNA. J. Mol. Biol. 304:119-133.[CrossRef][Medline]

21. Kim, Y. K., and S. K. Jang. 2002. Continuous heat shock enhances translational initiation directed by internal ribosomal entry site. Biochem. Biophys. Res. Commun. 297:224-231.[CrossRef][Medline]

22. Kullmann, M., U. Gopfert, B. Siewe, and L. Hengst. 2002. ELAV/Hu proteins inhibit p27 translation via an IRES element in the p27 5'UTR. Genes Dev. 16:3087-3099.[Abstract/Free Full Text]

23. Lee, A. S., J. Bell, and J. Ting. 1984. Biochemical characterization of the 94- and 78-kilodalton glucose-regulated proteins in hamster fibroblasts. J. Biol. Chem. 259:4616-4621.[Abstract/Free Full Text]

24. Mizutani, A., M. Fukuda, K. Ibata, Y. Shiraishi, and K. Mikoshiba.2000 . SYNCRIP, a cytoplasmic counterpart of heterogeneous nuclear ribonucleoprotein R, interacts with ubiquitous synaptotagmin isoforms. J. Biol. Chem. 275:9823-9831.[Abstract/Free Full Text]

25. Munro, S., and H. R. Pelham. 1986. An Hsp70-like protein in the ER: identity with the 78 kd glucose-regulated protein and immunoglobulin heavy chain binding protein. Cell 46:291-300.[CrossRef][Medline]

26. Olden, K., R. M. Pratt, C. Jaworski, and K. M. Yamada.1979 . Evidence for role of glycoprotein carbohydrates in membrane transport: specific inhibition by tunicamycin. Proc. Natl. Acad. Sci. USA 76:791-795.[Abstract/Free Full Text]

27. Pelham, H. R. 1986. Speculations on the functions of the major heat shock and glucose-regulated proteins.Cell 46:959-961.[CrossRef][Medline]

28. Pelletier, J., and N. Sonenberg. 1988. Internal initiation of translation of eukaryotic mRNA directed by a sequence derived from poliovirus RNA. Nature 334:320-325.[CrossRef][Medline]

29. Peluso, R. W., R. A. Lamb, and P. W. Choppin.1978 . Infection with paramyxoviruses stimulates synthesis of cellular polypeptides that are also stimulated in cells transformed by Rous sarcoma virus or deprived of glucose. Proc. Natl. Acad. Sci. USA 75:6120-6124.[Abstract/Free Full Text]

30. Piñol-Roma, S., and G. Dreyfuss. 1993. Cell cycle-regulated phosphorylation of the pre-mRNA-binding (heterogeneous nuclear ribonucleoprotein) C proteins. Mol. Cell. Biol. 13:5762-5770.[Abstract/Free Full Text]

31. Prostko, C. R., M. A. Brostrom, E. M. Galuska-Malara, and C. O. Brostrom. 1991. Stimulation of GRP78 gene transcription by phorbol ester and cAMP in GH3 pituitary cells. The accommodation of protein synthesis to chronic deprivation of intracellular sequestered calcium. J. Biol. Chem. 266:19790-19795.[Abstract/Free Full Text]

32. Rhoads, R. E., and B. J. Lamphear. 1995. Cap-independent translation of heat shock messenger RNAs. Curr. Top. Microbiol. Immunol. 203:131-153.[Medline]

33. Rousseau, D., R. Kaspar, I. Rosenwald, L. Gehrke, and N. Sonenberg.1996 . Translation initiation of ornithine decarboxylase and nucleocytoplasmic transport of cyclin D1 mRNA are increased in cells overexpressing eukaryotic initiation factor 4E. Proc. Natl. Acad. Sci. USA 93:1065-1070.[Abstract/Free Full Text]

34. Sarnow, P. 1989. Translation of glucose-reg