Viral IRESes are denoted by being relatively long (> 200 nt), guanosine/cytosine (G/C) nt rich, and containing upstream open reading frames (uORFs). The FMR1 5' leader exhibits a subset of these characteristics being approximately 240 nt (depending upon the number of CGG repeats) and is > 80% G/C rich, but it does not contain any uORFs. To confirm a previous report that the FMR1 5' leader contained an IRES 28, the leader was inserted into the intercistronic region of a dicistronic luciferase construct. Two negative controls were used, the multiple cloning site (MCS) in the intercistronic region (pRF) and the β-globin 5' leader. The β-globin 5' leader was chosen since the β-globin mRNA is translated exclusively in a cap-dependent manner 29. The encephalomyocarditis (EMCV) virus IRES was chosen as the positive control 30. The constructs were transfected into the C6 glioma and N2a neuroblastoma cell lines. After 24 hrs the cells were harvested and assayed for Renilla and Photinus luciferase. The Photinus:Renilla luciferase (P:R) ratio obtained from the control dicistronic construct (pRF) was normalized to one. A ratio above one would indicate IRES activity. Both the EMCV IRES and the FMR1 5' leader exhibited P:R ratios significantly higher than that observed from the pRF (and β-globin) construct. Indeed, the P:R ratio obtained from the FMR1 5' leader was approximately 30 and 60 fold higher in C6 and N2a cells respectively (Fig. 1A). The level of FMR1 IRES activity was higher than that observed from the EMCV IRES. This initial result suggests that the FMR1 5' leader may contain an IRES.

In addition to internal initiation, increased levels of Photinus luciferase protein can be generated from the dicistronic luciferase constructs through cryptic splicing or cryptic promoter activity. For example, the presence of a cryptic promoter in the 5' leader will lead to the production of a monocistronic Photinus luciferase mRNA 3132 and artificially increase the P:R ratio. To determine if cryptic promoter activity was present in the FMR1 5' leader, the dicistronic luciferase constructs with or without the SV40 promoter and intron were transfected into C6 cells. Photinus luciferase activity from the promoterless constructs containing the pRF MCS, β-globin 5' leader, and EMCV IRES was very low, less than 1% of the Photinus luciferase activity obtained from the constructs with the intact promoter. This result confirms previous studies indicating that these leaders do not contain a cryptic promoter 2633. However, the promoterless construct containing the FMR1 5' leader generated approximately 15% of the 'total' Photinus luciferase activity. After subtracting the minor contribution of the cryptic promoter to the translation of the second cistron, the data still suggests that the FMR1 5' leader has an IRES, but it does temper this conclusion.

To determine more unambiguously whether the FMR1 5' leader contains an IRES, the dicistronic constructs were transcribed in vitro eliminating the possibility of alternative splicing and alternate promoters. Transfecting the mRNA into C6 cells did provide evidence of IRES activity (Fig. 2A). The P:R ratio obtained from the dicistronic luciferase mRNA containing the FMR1 5' leader was approximately 4.5 fold higher than the β-globin control. Although the P:R ratio generated from the FMR1 5' leader was substantially lower when comparing the RNA to DNA transfections, the results still indicate that the FMR1 5' leader contains an IRES.

The dicistronic luciferase assay is useful to identify sequences that can internally initiate translation, but it does not indicate the role of an IRES in a monocistronic mRNA, the context in which the IRES is normally found in cellular mRNA. Consequently, two approaches were utilized to determine whether the FMR1 IRES is a major contributor to the translation of a monocistronic capped mRNA. Initially, in vitro transcribed monocistronic mRNA containing the Photinus luciferase open reading frame (ORF) and the β-globin or FMR1 5' leader was translated in rabbit reticulocyte lysate. The overall level of Photinus luciferase synthesis was reduced by approximately 40% when the FMR1 5' leader was present. This result is not surprising since cap-dependent translation of a short unstructured 5' leader (β-globin) is very efficient. Increasing concentrations of cap analog were added to the lysate to compete with the cap structure for eIF-4E and inhibit cap-dependent translation. Translation of the mRNA containing the β-globin 5' leader decreased as the concentration of cap analog increased (Fig. 2B). This result demonstrates that the β-globin mRNA is being translated in a cap-dependent manner. On the other hand, translation of the mRNA containing the FMR1 5' leader was only moderately affected (Fig. 2B); translation of the Photinus luciferase cistron decreased by only 15% at the highest concentration of cap analog. This result not only indicates that the mRNA containing the FMR1 5' leader is being translated in a cap-independent manner, but that it may be the major mechanism for its translation.

To determine the role of the FMR1 IRES within a cell, cap-dependent translation was inhibited ex-vivo. The 4E-Binding Protein 1 (4E-BP1) binds and sequesters eIF-4E preventing cap-dependent translation 34, but phosphorylation of 4E-BP1 decreases its affinity to eIF-4E. Consequently, C6 cells were transfected with a construct coding for a mutant of 4E-BP1 (4E-BPmut) with the two key phosphorylation sites mutated (Thr – 37 – Ala/Thr – 46 – Ala) or a control plasmid 20. Monocistronic constructs containing the β-globin, EMCV, or FMR1 5' leader were co-transfected. In the presence of over-expressed 4E-BPmut, the level of Photinus luciferase activity derived from the mRNA containing the β-globin 5' leader decreased by 72% (Fig. 2C). However, translation from the mRNAs containing the FMR1 or EMCV 5' leader only decreased by 19% and 27%, respectively (Fig. 2C). Since 81% of the luciferase activity remains when cap-dependent translation is inhibited, it indicates that IRES-dependent translation may be the primary mechanism for translation of the FMR1 mRNA in vivo.

To examine whether IRES-dependent translation is utilized for the synthesis of endogenous FMRP, cap-dependent translation was inhibited by treating C6 cells with rapamycin. The mammalian target of rapamycin (mTOR) is a kinase that is a key link in the promotion of cap-dependent translation by phosphorylating 4E-BP and p70 S6 kinase. Exposure to 20 nM rapamycin for 24 hrs actually led to a modest increase in FMRP expression (Fig. 3A). Phosphorylation of p70 S6 kinase was abolished demonstrating that the mTOR pathway was inhibited and the expression level of eIF-4E was decreased indicating cap-dependent translation was repressed (Fig. 3A). These results indicate that FMRP can be expressed under conditions in which cap-dependent translation is impeded.

In addition to inhibiting cap-dependent translation, blocking mTOR activity can alter other cellular pathways and indirectly alter translation. To more directly examine whether IRES-dependent translation is utilized for the synthesis of endogenous FMRP, cap-dependent translation was inhibited by knocking-down expression of eIF-4E. C6 cells were exposed to a pool of siRNA (Dharmacon) directed against the eIF-4E mRNA or a nonsense siRNA for 72 hr. Quantitation of Western blots showed that eIF-4E expression was reduced by > 70% (Fig. 3B, C). The level of the transcription factor JunB whose mRNA is translated in a cap-dependent manner was also reduced by > 75%. However, FMRP expression was decreased by only 12% (Fig. 3B, C). The observations that FMRP expression is maintained despite reduced levels of eIF-4E or mTOR activity (detailed above) indicates that IRES-dependent translation is likely a important contributor to the synthesis of FMRP.

A deletional analysis was performed to identify regions in the FMR1 5' leader important for IRES activity. Serial 5' truncations ranging from 11 to 53 nt in length were created. The abridged 5' leaders were inserted into the dicistronic luciferase construct and in vitro transcribed mRNA was transfected into C6 cells. The initial truncations resulted in a loss of IRES activity. Truncating the nt from ^-235 to ^-208 resulted in the largest decrease in IRES activity with the resulting P:R ratio being only 1.5 fold over that obtained from the β-globin control (Fig. 4A). Deleting an additional 42 nt (^-208 – ^-167) abolished all IRES activity. Additional truncations of 46 nt (^-166 – ^-121) actually led to an increase in the P:R ratio of approximately 1.5 fold over the control value. Further truncations of 53 and 13 nt (^-120 – ^-54) that encompassed the CGG repeats resulted in a decrease and complete loss of IRES activity, respectively. These results indicate that 1) the 5' end of the 5' leader is important for wild-type IRES activity, 2) an internal region may be inhibitory, and 3) the region including ^-120 – ^-54 exhibits a basal level of IRES activity. Interestingly, this latter region encompasses the CGG repeats of which nine are contained in the present 5' leader.

To determine the regulatory elements in the FMR1 IRES, C6 cells were exposed to 500 μM ameloride for 24 hrs to block the Na+/H+ antiporter. Intracellular acidification inhibits neural activity and is a model of an inactive neuron 35. The P:R ratio from the dicistronic mRNA containing the full-length 5' leader dramatically decreased in the presence of ameloride (Fig. 4A). However, only subtle differences in the P:R ratio were observed from truncations of the 3' 208 nt. The minimal level of IRES activity remained in the shorter leaders. This result implies that a decrease in intracellular pH inhibits FMR1 IRES activity and this effect is mediated by repressing the IRES-promoting region located at the 5' end of the 5' leader.

The basal level of IRES activity exhibited in the 5' leader containing the 3' 120 nt was not affected by changes in intracellular pH. An additional truncation deleting the CGG repeats abolished all IRES activity. To further characterize the role of the CGG repeats in the FMR1 IRES, the repeats were internally deleted within the full-length 5' leader. Transfection of the dicistronic mRNA containing the FMR1 5' leader with the CGG repeats deleted exhibited an approximately 50% decrease in the P:R ratio compared to the full-length 5' leader (Fig. 4B). This result demonstrates the importance of the CGG repeats for internal initiation mediated by the FMR1 5' leader.

IRES-dependent translation is affected by multiple cellular stimuli 36. We sought to determine whether the FMR1 IRES is regulated by environmental stimuli which regulate the processes in which FMRP participates. Neural activity leads to translation of FMR1 mRNA 237 and we modeled this phenomenon by treating cells with 50 mM KCl for 30 or 150 min. The short KCl exposure led to a 32% increase in the P:R ratio from the dicistronic mRNA containing the FMR1 5' leader (Fig. 5A). However, the longer KCl treatment resulted in a 28% decrease in the P:R ratio. Moreover, the FMR1 P:R ratio after the 150 min treatment was only one-third higher than that obtained from the control dicistronic mRNA. This result indicates that cellular depolarization differentially affects FMR1 IRES activity depending upon its duration.

To examine whether KCl also affects FMRP expression, Western blots were performed from lysates obtained from the treated and untreated cells. Changes in endogenous FMRP levels mirrored that observed from FMRP IRES (Fig. 5B, C). An increase of 58% was seen after a 30 min KCl treatment, whereas a 150 min treatment led to a 40% decrease in FMRP expression.

FMRP has been localized to the RNA-induced silencing complex (RISC), a nuclease complex that mediates RNA interference (RNAi). Deletion of FMRP leads to a loss of RNAi 38. To determine if the presence of double stranded RNA affects FMR1 IRES activity, C6 cells were exposed to 500 μg/ml of polyinosinic:polycytidylic acid (poly I:C), a double stranded polyribonucleotide. After a 7 hr exposure to poly I:C., the FMR1 P:R ratio increased by 41% (Fig. 6A). This result indicates a possible positive feedback mechanism to stimulate FMRP synthesis in response to double stranded RNA and it predicts that RNAi activity will increase FMR1 translation in an IRES-dependent manner. Tempering this conclusion was that expression of endogenous FMRP did not change after a 7 hr exposure to poly I:C (Fig. 6B, C). As was observed from the KCl experiments, it is possible that FMRP expression may be upregulated at different timepoints following poly I:C exposure.

The FMR1 5' leader was PCR amplified from a human brain cDNA library (Clontech) and inserted into the dual luciferase vector – pRF 5455 (a generous gift from Dr. Anne Willis, University of Leicester) with EcoRI and NcoI restriction endonuclease sites. The promoterless construct was created by digesting the dicistronic construct with SmaI and EcoRV and religating the construct. The monocistronic vector for the ex vivo experiments was created by digesting the RP vector with EcoRI and BamHI. The digest released the 5' leader, the Photinus luciferase gene, and the SV40 3' UTR, which were cloned into the pGL3 vector (Promega). The monocistronic vector for the in vitro experiments was created by inserting the above Photinus gene into the SK+ Bluescript vector (Stratagene) downstream of a T7 promoter.

Serial truncations were produced by PCR amplification with 5' and 3' primers containing EcoRI and NcoI endonuclease restriction sites, respectively. Deletion of the CGG repeats was accomplished by amplifying the region 3' to the repeats in the FMR1 5' leader and inserting it into the pRF construct with EcoRI and NcoI restriction sites on the 5' and 3' end, respectively. The 5' leader upstream of the CGG repeats was amplified and inserted upstream of the 3' FMR1 5' leader using EcoRI restriction sites. In experiments using a hypophosphorylated form of 4E-BP1 (containing Thr-37-Ala/Thr-46-Ala mutations), plasmids expressing this protein or the parent vector (both based on pACTAG-2) were co-transfected with the monocistronic constructs described above, using a 8-fold molar excess of the 4E-BP1 or control expression constructs 20. The 4E-BP1 double mutant and control expression plasmids were generously provided by Dr. Nahum Sonenberg (McGill University, Montreal).

The Bluescript SK+ vector containing the 5' leaders upstream of the Photinus luciferase gene was linearized with BamHI and in vitro transcribed using mMessage Machine (Ambion) producing capped mRNA. The mRNA was extracted with phenol/chloroform and a sample was run on an agarose gel to ensure RNA integrity. 0.5 μg of the mRNA and 1.6 nM methionine was added to rabbit reticulocyte lysate (Speed Read, Novagen) and incubated for 1 hr at 30°C in the presence or absence of cap analog (Ambion). The sample was subsequently assayed for Photinus luciferase activity.

C6 cells were obtained from ATCC and cultured in DMEM, 10% fetal bovine serum and 2 mM L-glutamine. Polyinosinic:polycytidylic acid, ameloride, and KCl were obtained from Sigma. Cells were transfected with 2 μg of DNA using Fugene transfection reagent (Roche) or 4 μg of mRNA using the TransMessenger RNA transfection reagent (Qiagen) according to the manufacturer's directions. The RNA was in vitro transcribed using the mMessage Machine (Ambion) and purified as detailed above. After 24 hr (DNA transfections) or 7 hours (RNA transfections), the cells were lysed with 500 μl of lysis buffer (Promega). Forty μl of the supernatant were used for the luciferase assays using the Dual-Luciferase Reporter Assay System and analyzed in a Luminoskan luminometer. For the siRNA experiments, C6 cells were transfected with 10 pmol of rat siRNA targeted against eIF-4E (On-Target plus SMARTpool, cat# L-0088826-01, Dharmacon) or nonsense siRNA (Dharmacon) using INTERFERin siRNA transfection reagent (Polyplus-Transfection) as directed by the manufacturer. Cells were lysed with 200 μl of lysis buffer (Promega) including phosphatase (Pierce) and protease (Roche) inhibitors after 72 hours. The cell lysates were analyzed by Western blot.

Cells were harvested in cell lysis buffer (Promega) with protease (Roche) and phosphatase inhibitors (Pierce). The cell lysate was analyzed by Western blot by separating the proteins on a 12% SDS-polyacrylamide gel and transferred onto nitrocellulose. The membranes were blocked (Invitrogen) and probed with a monoclonal antibody (7G-1, 1:1000 dilution; Developmental Hybridoma Bank) or a polyclonal antibody (sc-28739, 1:200 dilution; Santa Cruz) directed against FMRP. Both antibodies mainly recognized a single band at approximately 80 kD in lysates from C6 cells and rat hippocampus (see Additional file 1). In addition monoclonal antibodies directed against phosphorylated p70 S6 kinase (9206,1:1000 dilution; Cell Signaling) and eIF-4E (610269; 1:500 dilution, BD Biosciences) or polyclonal antibodies directed against JunB (Ab31421, 1:500 dilution; Abcam), and GAPDH (sc-25778, 1:200 dilution; Santa Cruz) were used in 5% nonfat dried milk in a solution of PBS containing 0.1% Tween 20. The blots were then incubated with alkaline phosphatase-conjugated secondary antibodies (1:10,000 – 1:50,000 dilution; Invitrogen) for 1 hour. Immunoreactive bands were detected using chemiluminescence (Invitrogen) as per manufacturer's directions. Westerns blots were either from individual PAGE gels (Fig. 3A) or the nitrocellulose was restripped and reprobed (Figs. 3B, 5B, 6B). The Western blots were quantitated using ImageQuant software (Applied Biosystems).