Analysis of cis-acting expression determinants of the tobacco psbA 5’UTR in vivo

Chloroplast gene expression is predominantly regulated at the
posttranscriptional levels of mRNA stability and translation
efficiency. The expression of psbA, an important
photosynthesis-related chloroplast gene, has been revealed to be
regulated via its 5’- untranslated region (UTR). Some cis-acting
elements within this 5’UTR and the correlated trans-acting factors
have been defined in Chlamydomonas. However, no in vivo evidence
with respect to the cis-acting elements of the psbA 5’UTR has been
so far achieved in higher plants such as tobacco. To attempt this,
we generated a series of mutants of the tobacco psbA 5’UTR by base
alterations and sequence deletions, with special regard to the
stem-loop structure and the putative target sites for ribosome
association and binding of nuclear regulatory factors. In addition,
a versatile plastid transformation vector pKCZ with an insertion
site in the inverted repeat region of the plastid genome was
constructed. In all constructs, the psbA 5’UTR (Wt or modified) was
used as the 5’ leader of the reporter gene uidA under control of
the same promoter, Prrn, the promoter of the rRNA operon. Through
biolistic DNA delivery to tobacco chloroplasts, transplastomic
plants were obtained. DNA and RNA analyses of these transplastomic
plants demonstrated that the transgenes aadA and uidA had been
correctly integrated into the plastome at the insertion site, and
transcribed in discrete sizes. Quantitative assays were also done
to determine the proportion of intact transplastome, the uidA mRNA
level, Gus activity, and uidA translation efficiency. The main
results are the following: 1) The insertion site at the unique MunI
between two tRNA genes (trnR-ACG and trnNGUU) is functional. Vector
pKCZ has a large flexibility for further DNA manipulations and
hence is useful for future applications. 2) The stem-loop of the
psbA 5’UTR is required for mRNA stabilisation and translation. All
mutants related to this region showed a 2~3 fold decrease in mRNA
stability and a 1.5~6 fold reduction in translation efficiency. The
function of this stem-loop depends on its correct sequence and
secondary conformation. 3) the AU-box of the psbA 5’UTR is a
crucial translation determinant. Mutations of this element almost
abolished translation efficacy (up to 175-fold decrease), but did
not significantly affect mRNA accumulation. The regulatory role of
the AU-Box is sequencedependent and might be affected by its inner
secondary structure. 4) The internal AUG codon of the psbA 5’UTR is
unable to initiate translation. An introduction of mRNA
translatability from this codon failed to direct the translation of
reporter uidA gene, overriding the mutation of the AU-Box. 5) The
5’end poly(A) sequence does not confer a distinct regulatory
signal. The deletion of this element only insignificantly affected
mRNA abundance and translation. However, this mutation might
slightly disturb the conformation of the stem-loop, resulting in a
moderate decrease in translation efficiency (~1.5 fold). 6) The
SD(Shine-Dalgarno)-like RBS (ribosome binding site) of the psbA
5’UTR appears to be an indispensable element for translation
initiation. Mutation of this element led to a dramatically low
expression of the uidA gene as seen by Gus staining. 7) The 5’end
structural sequence of the rbcL 5’UTR does not convey a high mRNA
stabilising effect to the psbA 5’UTR in a cycling condition of the
light and the dark. Their distinct roles appear to be involved in
darkness adaptation. Furthermore, with respect to the overall
regulatory function of the psbA 5’UTR, two models are proposed,
i.e. dual RBS-mediated translation initiation, and cpRBPs-mediated
mRNA stability and translation. The mechanisms for mRNA
stabilisation entailed by the rbcL 5’UTR are also discussed. Direct
repeat-mediated transgene loss after chloroplast transformation and
other aspects related to the choice of insertion site and plastid
promoter are also analysed.