The introduction of
unnatural amino acids into proteins can be achieved by the use of an amber
suppressor tRNA, that has been chemically acylated with the desired amino acid
and that is not a substrate for the natural aminoacyl tRNA synthetases. The
addition of the charged amber suppressor tRNA to the protein
biosynthesis reaction results in site specific incorporation of the amino acid
into the protein. However very often the usually applied methodology does not
lead to sufficient amounts of mutant protein. The goal of this work was to
expand our knowledge of in vitro amber suppression expecting that
the strategy mentioned above can be applied more general.
Within
the presented work the site specific incorporation of an unnatural amino acid
could be demonstrated by the introduction of ?-dansyl lysine into FABP (fatty acid binding
protein). As expected the incorporation of the unnatural amino acid was very
inefficient. A further limitation of the methodology proved to be the
preparation of large amounts of shortened tRNA molecules, which are necessary
to produce the chemically acylated tRNA. The high level of 3'-end heterogeneity
of the in vitro transcription products strongly impedes the purification
of homogenous tRNA. The homogeneity and the amount of the tRNAs during
T7-transcription could clearly be improved by increasing the reaction
temperature from 37°C to an optimum temperature of 44°C. Additionally, various
preparations of T7-RNA-Polymerase showed different n+1-activities.
To
investigate the activities of amber suppressor tRNAs expressed in
vivo, total tRNA from 10 different Escherichia coli amber suppressor
strains (Kleina et al. 1990) was prepared and employed in cell-free
translation. The relations between the suppression activities of the amber
suppressors for leucine, histidine, tyrosine and serine in vitro
corresponded to the in vivo results of other authors. In contrast to
those four the suppression activities of other amber suppressors were
decreased in vitro indicating that the activities of the corresponding
amino acyl tRNA synthetases may be reduced.
Processing,
repair and aminoacylation of in vitro transcribed tRNAs and the relation
of these processes to each other were investigated in-depth. Incubation of
3'-shortened or 3'-prolonged heterogenous in vitro transcription products
in the S100 enzyme fraction of the total translation system resulted in
homogenous tRNA populations with correct 3'-terminal CCA-ends. Processing of
prolonged tRNAs and repair of shortened tRNAs in the whole translation system
were shown not to be limiting for protein biosynthesis in vitro.
The
suppression activities of the transcripts of seven different amber suppressor
tRNA species (tRNASerCUA {su+1},
tRNATyrCUA {su+3}, tRNALeuCUA {su+6}, tRNALeu5CUA, tRNAPheCUA,
tRNAHisCUA and tRNAAla1CUA) were
shown not to be limited by aminoacylation. Therefore the suppression activities
of these tRNAs reflects structural properties of their aminoacylated
counterparts. Suppression efficiency was defined as the frequency of ribosomal
tRNA selection divided by the frequency of RF1 selection. The anticodon loops
of all tRNAs contained the sequence 5'C34U35A36A37A383'
which is ideal for efficient suppression as known from other studies. Still, in
the presence of this sequence suppression efficiencies varied over a large
range. The rate of tRNA selection was 20 times higher for the strongest
suppressor, tRNASerCUA,
compared to the efficiency of the weakest one, tRNAAla1CUA,
which represents the suppression efficiency of the actual tRNAs used for
chemical aminoacylation. In general the more the sequence of the amber suppressor
tRNA reflected the sequence of the original wild type tRNA, from which it was
deviated, the better the suppression efficiencies became. With increasing
number of nucleotide exchanges, that were necessary to get the sequence 5'C34U35A36A37A383'
into the anticodon loop of the amber suppressor, suppression
efficiencies decreased, indicating that tRNA sequences have been evolved to
support optimal interaction between codon and anticodon, as it is postulated in
the "Extended Anticodon" (Yarus 1982). The two best amber
suppressors by far, tRNASerCUA and tRNATyrCUA,
both contain the base C32 inside their anticodon loop. There is also
some evidence, that certain structural features important for suppression are
localized outside the anticodon arm and that these structural features are
found mainly in typeII-tRNAs.
The most important
conclusion resulting from this work within the field of biotechnology is, that
the actual amber suppressor tRNAs used for chemical aminoacylation are
comparably weak suppressors. A logical step from this work is the construction
of new amber suppressor tRNAs with higly improved suppression
efficiences for chemical aminoacylation. Therefore the present work should
allow a much improved incorporation of unnatural amino acids into proteins in
the future.
