Initiation of protein synthesis is regulated by sequence and structure of a mRNA
[reviewed in (
1
)]. In prokaryotes, a ribosome binding site (RBS) (AGGAGG) suitably positioned
at the 5'-end of the mRNA directs efficient translation. Activity of a RBS
may be influenced by the length and nucleotide composition of the spacer which
separates the RBS and the initiator (AUG) (
2
). In vertebrates, a Kozak sequence GCCG/ACCAUGC juxtaposed to the initiator
promotes efficient and accurate translation on a mRNA with a short 5' untranslated region (5' UTR) (
1
). A mRNA which lacks the Kozak consensus may be translated efficiently if it
possesses a moderately long 5' UTR that lacks stable secondary structures (
3
).
Frequently, it is desirable for investigators to be able to express a single
protein both in prokaryotes and eukaryotes. However, efficient translation of
proteins in these two systems dictates the use of two different 5' UTRs, as discussed above, to fulfill the requirements of the prokaryotic
and eukaryotic translational machineries. To circumvent this problem, we tested
the hypothesis that a single upstream element which combines some features of
both eukaryote and prokaryote 5' UTRs may function with both prokaryotic and mammalian ribosomes. A RBS
and spacer (AGGAGGGTTTTTA) of the
Ner
gene from bacteriophage Mu has been used for high level expression of proteins
in
Escherichia coli
and
Streptomyces
(
4
,
5
). Herein, we demonstrate that a variant of this
Ner
RBS-spacer which was altered to incorporate two nucleotides (CC) next to the
initiator ATG, termed
Ner
RBS-hybrid-spacer
ACC
, can sponsor very efficient translation of proteins in a vertebrate (rabbit
reticulocyte) system. A control RBS-spacer from the
LacZ
gene of
E.coli
was ineffective in this cross-species translational stimulation function exhibited by the
Ner
RBS-spacer.
By high fidelity PCR-mediated mutagenesis (
6
), the
Ner
RBS-hybrid-spacer
ACC
(AGGAGGGTTTTTACC) was linked to the 5' region of two test coding sequences, alkaline phosphatase of
E.coli
(
7
), and a glycosyl phosphatidylinositol phospholipase C (GPI-PLC) from the protozoan parasite
Trypanosoma brucei
(
8
). Primers and templates used were as follows. Alkaline phosphatase [Delta]2-22: Template, pAD135 (
7
); 5' primer, (F1 AP[Delta]2-22,
TAAGGATCCTTAACTTAGGAGGGTTTTTACCATGACACCAGAAATGCCTGTTCTGG); reverse primer (R446-BG2-AP, GATCGGATCCTTAAGATCTGCCCCAGAGCGGCTTTCATGG). For GPI-PLC mutagenesis, the template was pDH4 (
8
), the 5' primer was KCR10 (TAAGGATCCTTAACACAGGAGG
G
TTTTTACCATGTTTGGTGGT) and the reverse primer was KCR5
(TATGTGGATCCTTATGACCTTGCGGTTTGGTT). All PCR primers contained a
Bam
HI site for subcloning purposes. PCR products were digested with
Bam
HI and cloned into pBluescriptII (SK) (+) (Stratagene) under transcriptional
regulation of the phage T7 promoter therein. A GPI-PLC plasmid (pKMW2) (
9
) containing the RBS (AGGAGG) and spacer (CAGCTA) from the
LacZ
gene of
E.coli
was used as control for the
in vitro
translation experiments. All other sequences in the primers were identical to
those used in the construction of the
Ner
RBS-spacer clone. Identical amounts of plasmids purified by anion exchange
chromatography (Qiagen tip-20, Qiagen) were transcribed with purified T7 RNA polymerase and
translated in the same reaction mixture using rabbit reticulocyte lysate (TNT
T7 coupled transcription/translation reticulocyte lysate system, Promega)
containing [
35
S]methionine as directed by the manufacturer. Translation products were resolved
by SDS-PAGE and detected by fluorography. To assess the efficiency of
translation of the GPI-PLC containing the
Ner
RBS-spacer
CC
, the quantity of product was compared with that produced from a GPI-PLC cDNA clone with the authentic eukaryotic 5' UTR [pDH4 (
8
)]. Because the cDNA was cloned downstream of the T7 promoter in pBluescript
(Stratagene), identical reagents used for analysis of the constructs with the
E.coli
5' UTRs could be used for the experiments involving expression of the cDNA.
The nine bases preceding the initiator ATG in this cDNA are ATCATTGTA.
Transcripts from GPI-PLC constructs which contained the RBS-hybrid-spacer
ACC
from
Ner
were robustly translated by the rabbit reticulocyte lysates (Fig.
1
B, lanes 1 and 2, representing two clones of this construct), producing a full-length 39 kDa product. Indeed, translation from the
Ner
RBS-hybrid-spacer
ACC
construct was just as prolific as that obtained with the authentic 5' UTR of the trypanosome gene (Fig.
1
B, compare lane 3 with lanes 1 and 2). More importantly, translation initiated
from the correct AUG codon, since a 39 kDa GPI-PLC was produced; a protein of identical size is generated when a cDNA of
GPI-PLC is transcribed with the identical RNA polymerase and translated in the
reticulocyte lysate (Fig.
1
B, compare lanes 1 and 2 with lane 3). To test the possibility that downstream
(coding sequences) might be influencing the translation of GPI-PLC mRNA from the
Ner
RBS-hybrid-spacer
ACC
plasmids, the coding sequence of
E.coli
alkaline phosphatase [Delta]2-22 which has no protein sequence similarity with GPI-PLC was placed downstream of
Ner
RBS-hybrid-spacer
ACC
. This alkaline phosphatase construct was translated very efficiently to produce the expected 47 kDa
protein (Fig.
1
B, lanes 5 and 6; two representative clones), indicating that the
Ner
RBS-hybrid-spacer
ACC
was solely responsible for the efficient translation of coding sequences
positioned 3' to it.
REFERENCES
Return