An internal ribosome entry site mediates the ... - Wiley Online Library

An internal ribosome entry site mediates the initiation of soluble guanylyl cyclase b2 mRNA translation Roberto I. Vazquez-Padron1, Si M. Pham1, Dania Mateu1, Sheik Khan1 and Abdelouahab Aitouche2 1 University of Miami Miller School of Medicine, FL, USA 2 National Institutes of Health, Bethesda, MD, USA

Keywords IRES; nitric oxide; soluble guanylyl cyclase; translation; untranslated region Correspondence R. I. Vazquez-Padron, Division of Cardiothoracic Surgery and Vascular Biology Institute, University of Miami Miller School of Medicine, 1600 NW 10th Avenue, RMSB 1063, Miami, FL, 33136, USA Fax: +1 305 243 5636 Tel: +1 305 243 1154 E-mail: [email protected] (Received 18 January 2008, revised 8 May 2008, accepted 13 May 2008) doi:10.1111/j.1742-4658.2008.06505.x

The soluble guanylyl cyclases (sGC), the receptor for nitric oxide, are heterodimers consisting of an a- and b-subunit. This study aimed to investigate the translational mechanism of the sGC b2-subunit. Two mRNA species for sGC b2 were isolated from human kidney. These transcripts had dissimilar 5¢-untranslated regions (5¢-UTRs). The most abundant sGC b2 mRNA showed numerous upstream open reading frames (ORFs) and stable secondary structures that inhibited in vivo and in vitro translation. To evaluate whether these 5¢-UTRs harbored an internal ribosome entry site (IRES) that allows translation by an alternative mechanism, we inserted these regions between the two luciferase genes of a bicistronic vector. Transfection of those genetic constructs into HeLa cells demonstrated that both sGC b2 leaders had IRES activity in a cell-type dependent manner. Finally, the secondary structural model of the sGC b2 5¢-UTR predicts a Y-type pseudoknot that characterizes the IRES of cellular mRNAs. In conclusion, our findings suggest that sGC b2 5¢-UTRs have IRES activity that may permit sGC b2 expression under conditions that are not optimal for scanning-dependent translation.

The soluble guanylyl cyclases (sGCs) are the only receptors for nitric oxide (NO). In the presence of NO, sGCs produce cGMP to initiate the intracellular signaling that leads to vascular smooth muscle cell relaxation and synaptic transmission [1,2]. sGCs are heterodimeric enzymes consisting of one a-subunit and one b-subunit. There are two types of a-subunit and bsubunit that may constitute four different isoforms (a1b1, a2b1, a1b2, and a2b2) [3]. sGC a1b1 is expressed in most mammalian tissues and has been extensively characterized [4–6]. Nonetheless, information about other sGCs is very limited. sGC a2b1 is specifically produced in the placenta [7,8], and mRNA coding for the b2-subunit has been detected in kidney [9,10] and brain [11]. A search of the Geo Profiles

database for sGC b2 gene expression has linked the activation of this gene to neoplastic growth and inflammatory processes. The function of sGC b2 remains elusive. The amino acid sequence of this subunit resembles that of a primitive sGC found in insects [12]. This subunit has an isoprenylation consensus site at the C-terminus that, in principle, could target this protein to the membrane. Interestingly, sGC a1b2 in transfected cells is 100-fold less active than the a1b1 heterodimer [11,13]. This finding suggests that sGC b2 could be a dominant negative isoform, desensitizing sGCs and thus inhibiting NO signaling. Recently, a post-transcriptional variant of the sGC b2 that forms active homodimers has been described [14].

Abbreviations CMV, cytomegalovirus; ECMV, encephalomyocarditis virus; IRES, internal ribosome entry site; NO, nitric oxide; sGC, soluble guanylyl cyclase.

3598

FEBS Journal 275 (2008) 3598–3607 ª 2008 The Authors Journal compilation ª 2008 FEBS

R. I. Vazquez-Padron et al.

The sGC b2 5¢-UTR contains an IRES

The human sGC b2 gene lies on chromosome 13q14.3 and expands for more than 73 kb. It consists of 17 exons, of which four correspond to the 5¢-UTR. This 5¢-leader has numerous upstream initiation codons (uAUG) and stable, predictable secondary structures that difficult gene translation by the conventional scanning mechanism [15–17]. The translation of genes bearing complex 5¢-UTRs occur through alternative mechanisms such as IRES and ribosome jumping. The IRES direct ribosome binding straight to the initiation codon, avoiding the ribosome migration along the mRNA leader sequence [18,19]. This mechanism was originally discovered in picornavirus RNAs. Recently, a number of IRES have been identified in eukaryotic mRNAs coding for proteins associated with stress, cell growth, and cell death [20–25]. This study aimed to investigate the translational mechanism of the sGC b2-subunit. We found that sGC b2 translation is mediated by an IRES. This is the first example of an IRES being identified in one of the sGC genes.

Results Cloning of the sGC b2 5¢-UTRs The 5¢-UTR of sGC b2 was isolated from human kidney RNA by RT-PCR. Two cDNA fragments of 298 bp and 195 bp were amplified using specific primers (Fig. 1A). The 5¢-UTR of 298 bp was the most abundant in kidney, and several tumors and their normal adjacent tissues (Fig. S1 Supplementary materials). The 298 bp fragment corresponded to the full-length

5¢-UTR with a perfect match to sGC b2 GeneBank sequences NP004120 and AF038499 (Fig. 1A). The less abundant cDNA fragment of 195 bp showed a new and rare sGC b2 5¢-UTR that lacked exon III. This alternative sGC b2 mRNA has a new start codon that would generate a polypeptide with an N-terminus longer than the one encoded by the full-length 5¢-UTR (Fig. 1C). The 5¢-UTR of sGC b2 mRNA contains numerous uAUGs and stable predictable secondary structures The full-length sGC b2 5¢-UTR was characterized of having nine uAUGs that could potentially initiate translation (Table 1) [26]. Only the one at +116 bp was in frame with the sGC b2 ORF though it did not overlap the sGC b2 primary sequence. This uORF could potentially block access of ribosomes to the sGC b2 start codon to inhibit the scanning translation of this gene. The sGC b2 5¢-UTR had a calculated folding energy values up to )72.2 kcalÆmol)1 and predicted Y-type stem–loop structures (Fig. 2). This kind of structural motif is often detected in IRES of cellular mRNAs [27–30]. The sGC b2 5¢-UTR inhibits in vitro translation of a downstream reporter gene The sGC b2 5¢-UTR was inserted upstream to the FLuc gene under the control of the T7 promoter and the SV40 polyadenylation signal (Fig. 3A). Equimolar amounts of genetic constructs with and without the

A B

Fig. 1. Cloning and sequencing of sGC b2 5¢-UTRs. (A) Isolation of sGC b2 5¢-UTRs by RT-PCR. The arrows indicate amplified DNA fragments. (B) Nucleotide sequence of the 298 bp 5¢-UTR. (C) Nucleotide sequence of the 195 bp 5¢-UTR. Upstream AUGs (uAUG) are show in bold and underlined. The authentic initiation codon is identified by the amino acid sequence. The exons are limited by arrows.


C

3599



Table 1. Characteristics of the initiation codons of the sGC b2 5¢UTR.

AUG

Position

Context sequence

Consensusa

uORF lengthb

In framec

uAUG uAUG uAUG uAUG uAUG uAUG uAUG uAUG uAUG iAUG

54 58 116 151 159 171 175 178 234 282

AACaugU UGUaugG AACaugG AAGaugC UUCaugA ACAaugU UGUaugA AUGaugA UCCaugG AAGaugU

+ + ++ ) ) + ) ) + +

75 11 20 2 39 35 24 23 15 NA

No No Yes No No No No No No Yes

a

Degree of agreement with the Kozac consensus sequence for the initiation of protein translation. b Number of amino acids encoding for the upstream ORF (uORF) of the sGC b2 5¢-UTR. c Whether this ATG is in frame with the initiation codon.

sGC b2 leader were transcribed ⁄ translated in the presence of 7-methyl-G(5¢)ppp(5¢) guanosine cap analog in reactions supplemented with [35S]Met. In vitro translation produced luciferase and background proteins probably generated by proteolysis and protein aggregation. The presence of the sGC b2 leader decreased luciferase protein three-fold as compared to the reaction performed with the control plasmid pGEM-FLuc (Fig. 3B,C). On the other hand, the absence of cap analog slightly decreased the in vitro translation of sGC b2-FLuc. Similar results were obtained by inserting the alternative sGC b2 between T7 promoter and the reporter gene. This leader also decreased luciferase activity three-fold with respect to the control vector. These findings suggested that a region of stable secondary structure was obstructing gene translation and ⁄ or that factors that were not present in the rabbit reticulocyte lysates were required for in vitro protein production.

The sGC b2 5¢-UTR inhibits in vivo transcription and translation of a downstream reporter gene Next, the sGC b2 5¢-UTR was inserted between the cytomegalovirus (CMV) promoter and the FLuc gene to study its effects on downstream gene expression (Fig. 4A). Genetic constructs were cotransfected with the transfection control plasmid (pRL-TK) in HeLa cells. The levels of firefly and renilla mRNAs were measured by TaqMan RT-PCR. Total RNAs were treated with DNase to avoid contamination. HeLa cells transfected with the control plasmid, the pFLuc, produced two times (2.1 ± 1.32, n = 4) more FLuc mRNA than those transfected with the pFLuc b2. The ratio of renilla mRNA between the two groups was 0.97 ± 0.13, which indicated no differences in transfection. This result suggested that the sGC b2 5¢-UTR may have a negative effect on the gene transcription or mRNA stability. The latter seemed less likely, as northern blot analysis showed no signals of RNA degradation (Fig. 4B). However, after adjusting the luciferase activity on the basis of mRNA levels, the luciferase activity in cells transfected with pFLuc-b2 was 1.8 times lower than in those transfected with the control plasmid (Fig. 4C). This agrees with the in vitro results and suggests the presence of stable secondary structure in the sGC b2 5¢-UTR that may inhibit both transcription and scanning translation. It also suggests that the initiation of sGC b2 translation is through an alternative mechanism that is less efficient than scanning translation. The 5¢-UTR of sCG b2 mRNA harbors an IRES The sGC b2 5¢-UTR was then inserted downstream to the hairpin in the pSL3 bicistronic vector to determine whether this sequence contained an IRES (Fig. 5A). This plasmid incorporates the renilla and firefly luciferases as first and second cistrons, respectively. The

Fig. 2. Predicted secondary structure of the sGC b2 5¢-UTR. Predicted RNA secondary structures of the sGC b2 5¢-UTR calculated by the MFOLD algorithm. Both secondary structures have extendable and stable stem loops with a semiconserved Y-structure characteristic of the cellular IRES. The dots denote the position of the uAUG in the secondary structures. The folding energy (DG) appears below each structure.

3600



A


A pCMV

FLuc

pA

pFLuc sGC β2 5′ UTR pFLuc-β2

B pFLuc pFLuc-β2

B

β2 -FLuc mRNA FLuc mRNA

Relative luciferase activity/mRNA

C 100 90 80

*

60 40 20

pFLuc

C

Fig. 3. The sGC b2 5¢-UTR inhibits the in vitro translation of the downstream reporter gene. (A) Genetic constructs for in vitro translation. Both 5¢-UTRs were cloned downstream of the T7 promoter of the pGEM-11ZP(+) vector. (B) The in vitro translation products were resolved via SDS ⁄ PAGE, and [35S]Met-labeled polypeptides were detected by autoradiography. The pGEM-11ZP(+) vector was used as control. The relative position of the luciferase is indicated. (C) Luciferase activity from in vitro transcription ⁄ translation reactions in the presence or absence of 7-methyl-G(5¢)ppp(5¢) guanosine cap (Cap). The error bars represent the mean ± standard deviation of five independent experiments. *P < 0.01 with respect to the Cap-FLuc group as calculated by one-way ANOVA followed by the Duncan test (n = 5).

β2

Fig. 4. The sGC b2 5¢-UTR reduces in vivo transcription and translation of a downstream reporter gene. (A) Monocistronic reporter vectors. The sGC b2 5¢-UTR was inserted between pCMV and the firefly luciferase gene. (B) Northern blot analysis of monocistronic mRNAs. Total mRNAs from transfected cells were blotted to nylon membranes and probed with a radiolabeled DNA specific for luciferase. (C) Relative luciferase activity of transfected HeLa cells. The luciferase activity was normalized to the transfection control activity of pRL-TK and to the mRNA concentration determined by TaqMan RT-PCR. The error bars represent the mean ± standard deviation. *P < 0.01 as calculated by a t-test of unequal variances (n = 4).

hairpin between cistrons inhibits scanning translation and eliminates the possibility of ribosomal readthrough. The relative IRES activity (ratio of firefly to renilla luciferase) was determined in transiently transfected HeLa cells. The insertion of the sense-oriented sGC b2 5¢-UTR in the intercistronic position increased the translation of the downstream cistron 6.3-fold over the control (pSL3) and two-fold over the vector containing the EMCV IRES (Fig. 5B). The IRES activity of the bicistronic vector with the antisense leader was two times less than that of the one with the sense sGC


3601


A

pCMV

RLuc


h

of monocistronic mRNA by splicing [31], total RNAs from transfected HeLa cells with the bicistronic vectors pSL3 (control) and pSL3-b2 were analyzed by northern blot. The full-length mRNA and a faint upper band were detected (Fig. 5C). The band above the bicistronic mRNA was likely produced by an incomplete transcription termination of those genetic constructs. The absence of firefly luciferase monocistronic mRNAs ruled out the possibility of splicing sites within the sGC b2 5¢-UTR. This was further confirmed by RT-PCR. However, three times more firefly luciferase mRNA was observed in the control cells than in those transfected with pSL3b2. The L-myc and Apaf-1 IRESs also decreased the abundance of luciferase mRNA in HeLa cells [20,25]. To exclude the possibility that the sGC b2 5¢-UTR contains cryptic promoters, it was inserted upstream of the FLuc gene in the promoterless plasmid pGL3-Basic. The pFLuc with the CMV promoter was used as positive control (Fig. 6A). The absence of promoter abolished the luciferase activity in transfected HeLa cells regardless of the presence of the sGC b2 5¢-UTR (Fig. 6B). This experiment demonstrated that no cryptic promoter resides in this 5¢-UTR.

pA

FLuc

pSL3 1.6

1.0 3.2 Kb

EMCV pSL3-EMCV sGC β2 5′UTR

pSL3-β2 1.0

Relative IRES activity folds control (pSL3)

B

8

0.3 3.5 Kb

1.6

*

6

4

2

0 pSL3

EMCV

β2

C Kb 8 6 5 4 3

β2 -FLuc mRNA

The alternative sGC b2 5¢-UTR promotes the IRES-mediated translation

FLuc mRNA

2

Fig. 5. The sGC b2 5¢-UTR has IRES activity. (A) Genetic constructs utilised to demonstrate the presence of IRES in the sGC b2 5¢-UTR. The EMCV IRES was used as a positive control. (B) Relative IRES activity of HeLa cells transfected with bicistronic genetic constructs. Relative IRES activity was calculated by the ratio of firefly and renilla luciferase activity. The error bars represent the mean ± standard deviation of five independent experiments. *P < 0.01 as calculated by one-way ANOVA analysis followed by the Duncan test. (C) Northern blot analysis that shows the integrity of the bicistronic mRNA containing the sGC b2 5¢-UTR.

b2 5¢-UTR. However, the mRNA harboring the antisense leader was not totally bicistronic as detected by RT-PCR (data not shown). These data suggested that the 5¢-UTR of sGC b2 harbors an IRES only in the sense orientation. IRES activity in the 5¢-UTR of sGC b2 mRNA is neither due to the presence of a cryptic promoter nor to splicing sites in the bicistronic mRNA To demonstrate that the IRES activity observed in the sGC b2 5¢-UTR was not due to the production 3602

The ability of alternative sGC b2 5¢-UTRs to promote internal ribosome entry on a bicistronic mRNA was also assessed (Fig. 7A). This alternative leader lacks exon III along with five uAUGs that may have inhibitory effects on translation (Fig. 1C). The alternative leader increased the IRES activity 53-fold over pSL3 (control) and six-fold over the full-length 5¢-UTR (Fig. 7B). The mRNA containing this 5¢-UTR b2 was completely bicistronic, as shown by RT-PCR with two independent sets of primers (Fig. 7C). No cryptic promoters were found in this leader either (Fig. 6). Exon III alone slightly, but not significantly, increased IRES activity over the control (one-way anova analysis, n = 6, Fig. 7B). The elevated IRES activity in the alternative 5¢-UTR may be explained by the appearance of more efficient IRES structures and ⁄ or the absence of inhibitory uORFs. The alternative leader also possess an inframe AUG at +53 that may allow a more efficient translation of the downstream cistron (Fig. 1C). Taken together, these data suggest that both sGC b2 5¢-UTR isoforms are able to promote IRES-mediated translation.




A

pCMV

RLuc

FLuc

β2 5′UTR pSL3-β2

I

II III PCR #1 729 bp

PCR #2 525 bp

IV

A-β2 5′UTR II IV I

pSL3-A-β2 525 bp

576 bp

pSL3-EIII-β2

III

pSL-3

B

60 50 40 30 1 0

422 bp

65

Relative IRES activity folds control (pSL3)

Relative luciferase activity Folds of control (pGL3-Basic)

525 bp

*

55 45 35 25 15 5

pGL3-basic

β2

Aβ2

pSL3

pFLuc

β2

A-β2

EIII-β2

C Ladder Plus (bp)

1000 850 650 500

pSL-3

pSL3-β2

pSL3-Aβ2 +RT

pSL3-Aβ2 –RT

PCR #1

Fig. 6. The sGC b2 5¢-UTRs have no cryptic promoters. (A) Promoterless genetic constructs. The pFLuc was used as a positive control. (B) Genetic constructs were transfected into HeLa cells, and FLuc activity was measured and normalized to transfection efficiency and protein content. The data are expressed as fold increase over pGL3-Basic. The error bars represent the mean ± standard deviation of four independent experiments.

400 300

To investigate how widely the sGC b2 IRES is utilized in mammalian cells, a panel of cell lines from different tissues was used. These cell lines were transfected using either pSL3 (control) or pSL3-b2. The sGC b2 5¢-UTR was most active in human cervix epithelial adenocarcinoma (HeLa) cells and moderately active in transformed monkey kidney (Cos7) cells, in human lung carcinoma (CCL-185) cells, and in human hepatocellular carcinoma (Hep G2) cells (Fig. 8).

Discussion The mechanism by which sGC b2 mRNA is translated has remained elusive thus far. Herein, we present evidences that the human sGC b2 mRNA is translated via a nonconventional mechanism and that its 5¢-UTR has a strong IRES element comparable with that of picornaviruses [32,33]. Our demonstration of IRES activity in the 5¢-UTR of sGC b2 accounts for the translation of this gene despite the presence of abortive uORFs and stable secondary structures in its leader

1000 850 650

PCR #2

The sGC b2 IRES is functional in a wide range of cell lines

500

Fig. 7. The alternative sGC b2 5¢-UTR promotes IRES-mediated translation (A) Bicistronic genetic constructs utilized in this experiment. Exons within the 5¢-UTR are indicated (see Fig. 1B and C). (B) Relative IRES activity recovered from HeLa cells transfected with above genetic constructs. The error bars represent the mean ± standard deviation of six independent experiments. *P < 0.01 as calculated by one-way ANOVA analysis followed by the Duncan test. (C) RT-PCR analysis to demonstrate bicistronic RNA integrity using two sets of primers, PCR#1 and PCR#2 [as indicated in (A)]. The first lane is the 1 kb ladder Plus DNA molecular marker. Lane 5 shows the reverse transcriptase-negative control for each set of primers.

sequence that may inhibit scanning-dependent translation [18]. We initially characterized the sGC b2 5¢-UTR in vitro and in vivo. The use of monocistronic genetic construct demonstrated that the insertion of the sGC b2 5¢-UTR upstream of the FLuc gene inhibited its


3603



Fig. 8. The sGC b2 5¢-UTR is active in a range of cell lines. Cell lines were transfected with the pSL3-b2 bicistronic vector using Lipofectamine 2000. Luciferase activity was measured and normalized on the basis of transfection control and protein contents. The error bars represent the mean ± standard deviation of five independent experiments.

translation. This is consistent with the observations that only low levels of the sGC b2 protein are detected in tissues and with the low levels of b2-based sGC obtained in transfected cells [34, 9]. Transfection of cells with a bicistronic construct containing the sGC b2 5¢-UTR between cistrons demonstrated the presence of an IRES. The IRES activity was found, although differing in degree, in several human cell lines. The differences in the sGC b2 IRESmediated translation among cell lines may be due to the levels of translation factors require for a proper IRES function. Recently, the idea of IRES-mediated translation in eukaryotes has been challenged on the basis of the methods typically used for IRES identification [31]. IRES activity in the cells transiently transfected with bicistronic vectors is often confused with aberrant RNA cleavage, splicing, and ⁄ or the presence of the cryptic promoter within the construct itself that lead to the formation of a low amount of monocistronic messengers translated via the conventional scanning mechanism. No evidence of either RNA cleavage or cryptic promoter was found in cells transfected with the bicistronic constructs containing the sGC b2 5¢-UTR. These data confirmed that the sGC b2 gene could be produced by an IRES-mediated translational mechanism. We also investigated whether the alternative sGC b2 5¢-UTR had IRES activity. The IRES activity in this sequence was six-fold stronger than that in the fulllength and most abundant 5¢-UTR. This suggests that the ability of the sGC b2 5¢-UTR to initiate IRES translation resides in exons I, II and IV. It also suggests that the production of sGC b2 could be regulated 3604

at the post-transcriptional level through differential splicing, and at the translational level through an IRES. We can predict that under conditions where high levels of protein are required, more alternative sGC b2 mRNA is produced along with the factors needed for an efficient IRES-mediated translation. It is important to note that cellular IRES are often found in mRNA coding for regulatory proteins which expression is timely coordinated within the cell [18]. Furthermore, the IRES allows translation of transcripts in situations where cap-dependent translation is attenuated, such as stress, apoptosis, mitotic-phase transition, and development. There are several lines of evidence indicating a possible role of sGC b2 as negative modulator of NO signaling. For example, Gupta et al. demonstrated that cotransfection of sGC b2 along with a1 and b1 into Cos7 cells blunted sGC activity in response to NO [13]. In addition, they found that the expression of the b2-subunit was increased in the kidney of hypertensive rats where sGC activity was diminished [13]. Collectively, these findings support the hypothesis that sGC b2 could be a modulator of NO signaling under stress or pathological conditions where IRES translation is favored. Future studies will be needed to determine the specific role of IRES-mediated translation of sGC b2 in the control of NO signaling.

Experimental procedures Cloning of 5¢-UTRs of the human sGC b2 gene The sGC b2 5¢-UTRs were isolated from human kidney total RNA (Stratagene, La jolla, CA, USA) by RT-PCR with the following primers: 5¢-GCTTGGTGCTGCATCT CAATCCC-3¢ (forward) and 5¢-CTTCAGAATTGAAAG TATTCTCC-3¢ (reverse). Exon III was isolated in a similar manner but using 5¢-AAAGGTACCAACTTCTGC AGAAGTAC-3¢ (forward) and 5¢-CCATGGCTCGAGC CAGAATGTTGCAGG-3¢ (reverse) as primers. The PCR conditions per cycle were set as 30 s at 95 C, 30 s at 50 C, and 30 s at 72 C. PCR products were resolved on a 2% agarose gel, and the intensity of each band was measured using NIH imagej software (NIH, Bethesda, MA, USA). PCR products were further cloned into a pMosBlue-TA vector (Amersham Biosciences, Piscataway, NJ, USA), and their sequences were determined at Sequetech (Mountain View, CA, USA).

Genetic constructs Genetic constructs are depicted at the tops of the relevant figures. pSL3 was generated from pSL-EMCV by deleting



the IRES [35]. pSL3-b2, pSL3-Ab2 and pSL3-EIII-b2 were generated by inserting the corresponding sGC b2 5¢-UTR into the pSL-3. PGEM-FLuc and pGEM-b2 were generated by either cloning the firefly luciferase (FLuc) alone or fused to the sGC b2 5¢-UTR downstream of the T7 promoter in the pGEM-11ZP(+) (Promega, Madison, WI , USA). The pSL-FLuc was constructed by NcoI removal of the renilla luciferase gene (RLuc) from the bicistronic vector pSL 3. The sGC b2 5¢-UTR was KpnI-XhoI inserted in pSL-FLuc to generate pSL-b2. pGL3-b2 was generated by inserting the sGC b2 5¢-UTR upstream of the FLuc gene in the promoterless plasmid pGL-3 Basic (Promega).

Cell culture and transient transfection Transformed monkey kidney (Cos7), human lung carcinoma (CCL-185), human hepatocellular carcinoma (HepG2) and human cervix epithelial adenocarcinoma (HeLa) cells were obtained from American Tissue Culture Collection (Manassas, VA, USA). Cells were cultured under conventional conditions in DMEM supplemented with penicillin ⁄ streptomycin and 10% fetal bovine serum. Transfections were performed with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Cells were cultured with Opti-MEM I Reduced Serum Medium for 1 h before transfection. The DNA–Lipofectamine suspension (100 lL) containing 1.6 lg of plasmid and 4 lL of Lipofectamine 2000 was added drop by drop to the cells. After 4 h of incubation, 1 mL of growth medium was added and cells were cultured for 48 additional hours.

Luciferase activity The firefly luciferase activity of cells transfected with monocistronic reporter constructs was measured with the Luciferase Assay System (Promega). In these experiments, luciferase activity was normalized on the basis of the renilla luciferase activity of the transfection control vector pRLTK (Promega) and of the mRNA levels determined by TaqMan Real Time PCR. The luciferase expression of the bicistronic vector was determined using the dual luciferase assay system (Promega) in a Tuner Biosystems Lumminometer Model TD 20 ⁄ 20 (Mountain View, CA, USA). All assays were performed in triplicate on three to six different occasions. Final values were expressed as fold of the experimental control, indicated on each figure.

Northern blotting, RT-PCR, and TaqMan real time RT-PCR Total RNA was prepared with the RNeasy Midi kit (Qiagen, Valencia, CA, USA). Four micrograms of RNA were denatured in formaldehyde and separated on 1% formalde-


hyde ⁄ agarose gels. RNA was capillary blotted onto a nylon membrane. Filters were UV cross-linked, and hybridizations were performed in Ultrahigh hybridization solution (Ambion, Austin, TX, USA). The hybridization probe specific for luciferase were prepared with the T7 Ribroprobe system (Promega) in the presence of [32P]CTP[aP]. Total RNAs were treated with DNase before retrotranscription. RT-PCRs were carried out with no retrotranscripted RNAs (RT-Minus control) to rule out any possibility of DNA contamination. The integrity of bicistronic mRNA was also assessed by RT-PCR with two set of primers. PCR#1 (F-5¢-AACTTTCGAAGTCATGGTGG-3¢ and R-5¢-GAC TTTCCAAAATGTCGTAATAACC-3¢) amplified the intercistronic region. PCR#2 (F-5¢-TTCCATCTTCCAGCGGA TAG-3¢ and R-5¢-CTACGTGCAAGTGATGATTTAC-3¢) amplified a 525 bp internal fragment of the first cistron. The PCR conditions per cycle were set as 30 s at 95 C, 30 s at 52 C, and 1 min at 72 C. Firefly and renilla luciferase mRNAs were quantified using custom TaqMan Gene Expression Assays, respectively, according to the manufacturer’s instructions (Applied Biosystems, Foster City, CA, USA). Probes were FAM ⁄ MGB and primers were not limited. Real-time PCR was performed on an ABI Prism 7500 Fast Real-Time PCR System (96-well plate).

In vitro transcription and translation In vitro coupled transcription and translation were performed with the TnT Quick Coupled Transcription ⁄ Translation System (Promega). Reactions were performed with 10 fm of circular plasmids in the presence of [35S]Met at 37 C for 90 min. The translated proteins were resolved via 12.5% SDS ⁄ PAGE and labeled polypeptides were detected by autoradiography. The band intensity was measured using NIH imagej software. The luciferase activity of in vitro translated proteins was measured as described above.

Secondary structure modeling Minimal free energy calculations and secondary structure predictions were generated using the web-implemented version of the mfold algorithm incorporating version 3.0 of the Turner rules.

Acknowledgements We thank Mireya Hernandez for her technical assistance and Deborah Georges for her assistance with the preparation of the manuscript. This work was supported by NIH grant R01 HL63426 granted to Dr Si M. Pham.


3605



References 1 Andreopoulos S & Papapetropoulos A (2000) Molecular aspects of soluble guanylyl cyclase regulation. Gen Pharmacol 34, 147–157. 2 Lucas KA, Pitari GM, Kazerounian S, Ruiz-Stewart I, Park J, Schulz S, Chepenik KP & Waldman SA (2000) Guanylyl cyclases and signaling by cyclic GMP. Pharmacol Rev 52, 375–414. 3 Koesling D & Friebe A (1999) Soluble guanylyl cyclase: structure and regulation. Rev Physiol Biochem Pharmacol 135, 41–65. 4 Gupta G, Kim J, Yang L, Sturley SL & Danziger RS (1997) Expression and purification of soluble, active heterodimeric guanylyl cyclase from baculovirus. Protein Expr Purif 10, 325–330. 5 Lee YC, Martin E & Murad F (2000) Human recombinant soluble guanylyl cyclase: expression, purification, and regulation. Proc Natl Acad Sci USA 97, 10763– 10768. 6 Nedvetsky PI, Kleinschnitz C & Schmidt HH (2002) Regional distribution of protein and activity of the nitric oxide receptor, soluble guanylyl cyclase, in rat brain suggests multiple mechanisms of regulation. Brain Res 950, 148–154. 7 Harteneck C, Wedel B, Koesling D, Malkewitz J, Bohme E & Schultz G (1991) Molecular cloning and expression of a new alpha-subunit of soluble guanylyl cyclase. Interchangeability of the alpha-subunits of the enzyme. FEBS Lett 292, 217–222. 8 Russwurm M, Behrends S, Harteneck C & Koesling D (1998) Functional properties of a naturally occurring isoform of soluble guanylyl cyclase. Biochem J 335, 125–130. 9 Yuen PS, Potter LR & Garbers DL (1990) A new form of guanylyl cyclase is preferentially expressed in rat kidney. Biochemistry 29, 10872–10878. 10 Behrends S, Budaeus L, Kempfert J, Scholz H, Starbatty J & Vehse K (2001) The beta 2 subunit of nitric oxide-sensitive guanylyl cyclase is developmentally regulated in rat kidney. Naunyn Schmiedebergs Arch Pharmacol 364, 573–576. 11 Gibb BJ, Wykes V & Garthwaite J (2003) Properties of NO-activated guanylyl cyclases expressed in cells. Br J Pharmacol 139, 1032–1040. 12 Nighorn A, Byrnes KA & Morton DB (1999) Identification and characterization of a novel beta subunit of soluble guanylyl cyclase that is active in the absence of a second subunit and is relatively insensitive to nitric oxide. J Biol Chem 274, 2525–2531. 13 Gupta G, Azam M, Yang L & Danziger RS (1997) The beta2 subunit inhibits stimulation of the alpha1 ⁄ beta1 form of soluble guanylyl cyclase by nitric oxide. Potential relevance to regulation of blood pressure. J Clin Invest 100, 1488–1492.

3606

14 Koglin M, Vehse K, Budaeus L, Scholz H & Behrends S (2001) Nitric oxide activates the beta 2 subunit of soluble guanylyl cyclase in the absence of a second subunit. J Biol Chem 276, 30737–30743. 15 Meijer HA & Thomas AA (2002) Control of eukaryotic protein synthesis by upstream open reading frames in the 5¢-untranslated region of an mRNA. Biochem J 367, 1–11. 16 Kozak M (1978) How do eucaryotic ribosomes select initiation regions in messenger RNA? Cell 15, 1109– 1123. 17 Rogozin IB, Kochetov AV, Kondrashov FA, Koonin EV & Milanesi L (2001) Presence of ATG triplets in 5¢ untranslated regions of eukaryotic cDNAs correlates with a ‘weak’ context of the start codon. Bioinformatics 17, 890–900. 18 Hellen CU & Sarnow P (2001) Internal ribosome entry sites in eukaryotic mRNA molecules. Genes Dev 15, 1593–1612. 19 Vagner S, Galy B & Pyronnet S (2001) Irresistible IRES. Attracting the translation machinery to internal ribosome entry sites. EMBO Rep 2, 893–898. 20 Coldwell MJ, Mitchell SA, Stoneley M, MacFarlane M & Willis AE (2000) Initiation of Apaf-1 translation by internal ribosome entry. Oncogene 19, 899–905. 21 Creancier L, Morello D, Mercier P, Prats AC (2000) Fibroblast growth factor 2 internal ribosome entry site (IRES) activity ex vivo and in transgenic mice reveals a stringent tissue-specific regulation. J Cell Biol 150, 275– 281. 22 Coldwell MJ, deSchoolmeester ML, Fraser GA, Pickering BM, Packham G & Willis AE (2001) The p36 isoform of BAG-1 is translated by internal ribosome entry following heat shock. Oncogene 20, 4095–4100. 23 Lang KJ, Kappel A & Goodall GJ (2002) Hypoxiainducible factor-1alpha mRNA contains an internal ribosome entry site that allows efficient translation during normoxia and hypoxia. Mol Biol Cell 13, 1792– 1801. 24 Rubtsova MP, Sizova DV, Dmitriev SE, Ivanov DS, Prassolov VS & Shatsky IN (2003) Distinctive properties of the 5¢-untranslated region of human hsp70 mRNA. J Biol Chem 278, 22350–22356. 25 Jopling CL, Spriggs KA, Mitchell SA, Stoneley M & Willis AE (2004) L-Myc protein synthesis is initiated by internal ribosome entry. RNA 10, 287–298. 26 Peri S & Pandey A (2001) A reassessment of the translation initiation codon in vertebrates. Trends Genet 17, 685–687. 27 Le SY & Maizel JV Jr (1997) A common RNA structural motif involved in the internal initiation of translation of cellular mRNAs. Nucleic Acids Res 25, 362–369. 28 Le QuesneJP, Stoneley M, Fraser GA & Willis AE (2001) Derivation of a structural model for the c-myc IRES. J Mol Biol 310, 111–126.



29 Pedersen SK, Christiansen J, Hansen TO, Larsen MR & Nielsen FC (2002) Human insulin-like growth factor II leader 2 mediates internal initiation of translation. Biochem J 363, 37–44. 30 Mitchell SA, Spriggs KA, Coldwell MJ, Jackson RJ & Willis AE (2003) The Apaf-1 internal ribosome entry segment attains the correct structural conformation for function via interactions with PTB and unr. Mol Cell 11, 757–771. 31 Kozak M (2003) Alternative ways to think about mRNA sequences and proteins that appear to promote internal initiation of translation. Gene 318, 1–23. 32 Jang SK, Krausslich HG, Nicklin MJ, Duke GM, Palmenberg AC & Wimmer E (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. 33 Pelletier J, Kaplan G, Racaniello VR & Sonenberg N (1988) Cap-independent translation of poliovirus mRNA is conferred by sequence elements within the 5¢ noncoding region. Mol Cell Biol 8, 1103–1112. 34 Mundel P, Gambaryan S, Bachmann S, Koesling D & Kriz W (1995) Immunolocalization of soluble guanylyl


cyclase subunits in rat kidney. Histochem Cell Biol 103, 75–79. 35 Schiavi A, Hudder A & Werner R (1999) Connexin43 mRNA contains a functional internal ribosome entry site. FEBS Lett 464, 118–122.

Supplementary material The following supplementary material is available online: Fig. S1. Abundance of sGC b2 5¢-UTRs in the kidney (K) and in several tumors (T) and their normal adjacent tissues (NTA). This material is available as part of the online article from http://www.blackwell-synergy.com Please note: Blackwell Publishing are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.


3607