Transcripts expressed using a bicistronic vector pIREShyg2 are sensitized to nonsense-mediated mRNA decay

Abstract

Background: pIREShyg2 has been widely used as a bicistronic expression vector. However, it is not known if the vector would affect the expression of cloned genes via nonsense-mediated mRNA decay (NMD), an mRNA surveillance system that degrades mRNA with a premature termination codon (PTC). In mammalian cells, the induction of NMD requires either a long 3'UTR or the presence of an exon-junction complex downstream of a PTC. The efficiency of NMD is greater when a PTC generates longer 3'UTR. pIREShyg2 provides the first cistron gene with a long 3'UTR consisting of a downstream intervening sequence (IVS), an internal ribosomal entry site (IRES) and the second cistron. Therefore, we hypothesized that the first cistron genes in pIREShyg2 are sensitized to NMD, which affects their expression levels. To examine this hypothesis, cDNAs encoding human granulocyte-macrophage colony-stimulating factor receptor beta chain (betac) and its splice variant (betac79), in which the retention of a 79-base intron caused a frameshift generating 18 PTCs, were cloned into pIREShyg2 and stably expressed in a murine cell line, Ba/F3.

Results: Compared with wild-type betac, the mRNA levels of betac79 were less than one tenth and decayed faster. Both translation inhibition and Upf1 knockdown led to significantly greater up-regulation of betac79 than wild-type betac. However, the use of a monocistronic pMT21 vector abolished the up-regulatory effects of translation inhibition and Upf1 knockdown on both wild-type betac and betac79, suggesting that the NMD is attributable to a structural determinant in pIREShyg2. The elimination of the intron and the proximal 3' 17 PTCs did not alter the greater effects of translation inhibition on betac79, suggesting that the first PTC, which determines 3'UTR length, was sufficient to enhance NMD efficiency. Thus, transcripts of PTC-harboring genes with longer 3'UTR are more efficiently degraded by the vector-dependent NMD than those of wild-type genes with relatively shorter 3'UTR, resulting in minimized expression of truncated mutants.

Conclusions: We conclude that pIREShyg2, which sensitizes its bicistronic transcripts to NMD, may be useful for studying NMD but should be avoided when maximum expressions of PTC-harboring genes are required.

Figure 1

Schematic presentation of wild-type βc- and βc79-constructs and positions of PCR primers. βc cDNA from exon II to exon XIV, which contains the full length of the protein coding region, was inserted into a pIREShyg2 vector that had an IVS, an internal ribosome entry site (IRES), and hygromycin resistance gene between its cloning site and poly(A) region. The distance from the normal stop codon in wild-type βc to the IVS was 254 nucleotides. The retention of 79-base intron V in βc79 caused a frameshift that generated 18 PTCs. The first termination codon in βc79 was "TGA". Hyg^r: hygromycin resistance gene, P_CMV: human cytomegalovirus major immediate early promoter/enhancer.

Figure 2

DNA and RNA expression of introduced bc. (a) DNA and RNA were isolated from a wild-type βc-transfected clone 5C and four βc79-transfected clones mE, m821, m951, and 63m. The isolated DNA and reverse-transcribed RNA were subjected to the amplification of βc and GAPDH by PCR. Primers 1S/1A, 2S/2A, and 3S/2A were used for detection of βc. cDNA obtained from 5C was diluted by 10 times, while cDNA from βc79-expressing clones were directly applied to PCR without dilution. M: Marker. (b) The βc-specific transcripts were quantified by real-time PCR using primer 1S/1A. The amounts of βc-specific transcripts relative to those of endogenous GAPDH transcripts were compared among three wild-type and four βc79 clones. The average values from 3 runs of real-time RT-PCR are presented. Error bars express S.E.M. (c) The mE and 5C cells were cultured in the presence of 25 μg/mL DRB and the ratios of βc-specific transcripts to GAPDH transcripts were plotted at various time points. *: significantly lower than the ratio from 5C cells at the same time points (p < 0.05).

Figure 3

Dose-response relationship between βc79 transcripts and translation inhibitors. The βc79-expressing clone mE was incubated with various concentrations of puromycin, emetin, and CHX for 4.5 hours. The ratios of βc-specific transcripts to GAPDH transcripts were measured by real-time RT-PCR with 1S/1A primers. The ratios in NMD inhibitor-treated cells were divided by those in untreated cells, which are presented as fold increases. The mean values ± S.E.M. (error bars) were obtained from 3 to 4 independent experiments.

Figure 4

Time course of βc transcript levels in response to puromycin. (a) Two βc79 clones, mE and 63m, and a wild-type βc clone 5C were incubated with 100 μg/mL puromycin for the indicated periods, and the ratios of βc-specific transcripts to endogenous GAPDH transcripts were measured by real-time RT-PCR with 1S/1A primers. The values at 0 hours are plotted as 1.0. The average values ± S.E.M. (error bars) were obtained from 3 to 8 experiments. * and **: significantly higher than the value in 5C incubated for the same time with p < 0.05 and p < 0.01, respectively. (b) After 8-hour treatment, the βc-specific transcripts were quantified and normalized by those of GAPDH. The values from mE and 63m cells were compared with those from 5C cells, which are plotted as 1.0. (c) The siRNA targeting Upf1 was introduced into 5C and mE cells, and the amounts of Upf1 and βc transcripts were quantified after 48 hours. The presented data are the average values and S. E. M. (error bars) f3rom four different experiments. *: significantly different (p < 0.05).

Figure 5

Effects of puromycin and Upf1 knockdown on the constructs cloned in a pMT21 vector. (a) The structure of the pMT21 vector. SV40ori: simian virus 40 (SV40) origin of replication and enhancer element; AdMLP: adenovirus major late promoter; TPL: the tripartite leader; IVS: intervening sequence; DHFR: murine dihydrofolate reductase coding region; poly(A), polyadenylation signal. (b) The cDNAs of wild-type βc and βc79 were stably expressed in Ba/F3 cells using a pMT21 vector. These transfectants and 5C cells that expressed the wild-type βc in pIREShyg2 were incubated with or without 100 μg/mL puromycin for 4.5 hours. The values for βc/GAPDH in puromycin-treated cells were divided by the values in untreated cells, which are presented as fold increases of βc/GAPDH. Data shown are the average values ± S.E.M. (error bars) (pIREShyg2: n = 8, pMT21: n = 3). **: significantly smaller than the value in the wild-type βc cloned into pIREShyg2 (p < 0.01). (c) Upf1 was knocked down by siRNA in the cells expressing wild-type βc in pIREShyg2 (5C) and pMT21. The amounts of Upf1 and βc transcripts relative to GAPDH from three independent experiments are presented. *: significantly different (p < 0.05).

Figure 6

Replacement of the intron V with a single nucleotide. (a) βc-G was constructed by replacing the intron V-derived 79 nucleotides with a single nucleotide G in βc79. The frameshift generating eighteen PTCs were preserved in βc-G. βc258 was a βc with a single PTC at the 259 th codon. Both βc79 and βc-G had 2238 nucleotides downstream of their termination codon, and βc258 had 2183 nucleotides. (b) βc-G and βc258 were cloned into a pIREShyg2 vector and stably expressed in Ba/F3 cells. The cells expressing βc-G, βc258, βc79 (mE) and wild-type βc (5C) were cultured with or without 100 μg/mL puromycin for 4.5 hours. The amounts of βc-specific transcripts relative to GAPDH transcripts in puromycin-treated cells were divided by the amounts in untreated cells, and are presented as-fold increases of βc/GAPDH. Average values and S.E.M. (error bars) were obtained from three independent experiments. * significantly higher than 5C (p < 0.05). (c) EpoR215 was an EpoR with a termination codon (TAG) at the 215 th codon generating 1099 nucleotides downstram of the termination codon. In wild-type EpoR, there were 190 nucleotides downstream of its normal termination codon. (d) Wild-type EpoR and EpoR215 were stably expressed in Ba/F3 cells using a pIREShyg2 vector, and cultured with or without 100 μg/ml puromycin for 4.5 hours. The ratios of EpoR-specific transcripts to GAPDH transcripts in puromycin-treated cells were divided by those in untreated cells. Average values and S. E. M. (error bars) were obtained from three different experiments. * significantly higher than the value in wild-type EpoR (P < 0.05).