Introduction
In order to create a universal
restriction
enzyme we followed several ideas and models. The first and most
promising idea
we had was to create a monomeric restriction enzyme as it represents the optimal and
easiest
model for a universal restriction enzyme
If we are able to employ a
monomeric enzyme,
this protein would have a couple of advantages. Most importantly we
would no longer
need two separate oligonucleotides to achieve specific binding and
cutting of
the target DNA. Also only one protein has to be purified –
thus saving time and
money. A scientific advantage is the option to optimize the monomer by
phage
display. Using this technology, we would have the chance to create a
thermostable, specific, universal restriction enzyme whose DNA binding
activity
is only created by a single oligonucleotide.
Our goal was firstly to create
a Fok-monomer which
is able to cut DNA without a primary dimerization step and secondly to
show
that our heterodimeric interface design works properly. To reach this,
we had
to clone all the required parts one after another, resulting in a huge
fusion
protein.
Results
3D modelling and design goals
|
Fig1.:
Scheme
I.
lipocalin II.
Foka/i (heterodimers) III.
Linker IV. His-Tag |
The necessary parts in the
model of an
monomeric restriction encyme are: Tag-binding protein - Fok
heterodimer1 -
linker - Fok herterodimer2 (Fig1).
Tag |
binding protein |
linker |
Fok heterodimer |
HisTag (BBa_K157011)
StrepTag(BBa_K157012)
|
FluA(BBa_K157004)
DigA(BBa_K243003)
Fos(BBa_K243027)
|
GSAT-Linker(BBa_K243029)
SEGLinker(BBa_K243030)
|
FokA (K243000)
Foki (K243001)
|
Tab.:1 list of
parts usable for
Fok Monomer
In order to purify this fusion
protein we had
to employ a N-terminal tag. The binding protein (a lipocalin) mediates
between the
FokI protein and the tagged oligo. Both parts of the Fok heterodimers are
associated with a long linker allowing
them to establish the cutting site conformation. This artificial
restriction
nuclease domain is the functional part of the monomer and catalyzes the
DNA
cleavage.
Before
we started in the wet lab we planned all
the individual steps and to figure out the best way to create the Fok
monomer,
so the first thing we did was to design a Fok monomer in
silico.
|
Fig2.:
3D model of Fok Monomer |
The model (Fig2.) shows that we
have to use a
linker which is something about 70 Angström in length to
provide the required
distance between the two Fok parts. Therefore we decided to order two
complement oligonucleotides encoding one 36 amino acid glycine-serine
linker.
5'-CTAGATGGCCGGCGGTTCTGGTGGTGGTTCTGGCGGTGGTTCTGGAGGTAGTTCTGGCGGTGGATCTGGAGGCGGTTCTGGGTCAGGATCTGGTGGAGGTTCTGGCTCTGGGAATCAGA-3'
3'-TACCGGCCGCCAAGACCACCACCAAGACCGCCACCAAGACCTCCATCAAGACCGCCACCTAGACCTCCGCCAAGACCCAGTCCTAGACCACTACCAAGACCGAGACCCTTAGTCTGGCC-5
Cloning
However at this stage the first
problem
appeared. This are the oligonucleotides, each
120aa long
that we ordered from Mr.Gene. Unfortunately we made a mistake when
ordering the
complement oligonucleotide. As you can see we have two mismatches
(compare
coloured bases in the sequence). But somehow we managed to dimerize the
two
oligos. After dimerization in the thermo cycler
and
cloning into a pMA (BBa_K243031) vector (XbaI/AgeI) some mutations
appeared in
the 36GSLinker gene caused by the mismatches. But since there was no
frame
shift we decided to carry on.
Firstly we picked the parts we were going to
use. We decided to use one we already completed in our earlier
experiments. His
– FluA – Split -
Foki (BBa_K243010) has all the functions we needed. These are (I) a tag
to
purify, (II) a Fok domain and (III) the anticalin which binds the
modified
oligonucleotide Fig.1. Cloning the linker (no part) behind those was
the next
step. We followed the assembly standard 25 and cut the vector with AgeI
and
PstI and the inserted with NgoMIV and PstI.
|
Fig3. gele
picture of test digest
|
However, we ran into some more problems. After cloning Foka
(BBa_K243000) behind this construct, again assembly standard 25, we
performed
some test digestions to cut out the insert again. These showed that the
plasmid
became even smaller as it would be without an insertion (pEX: ~ 4,5kb
and after
test digestion ~ 2,2kb). The Insert should be about 1,9kb but as you
can see in
Fig.3 it´s about 900k. Also
the sequencing
went wrong.
Most likely the active Fok did
cut the plasmid
and promoted deletion of the Fok gene, what gave mutated cells a
significant
growth advantage. After that we decided to order the 36GS-linker (GSAT
Linker:
BBa_K243029) as gene synthesis.
5´-GAGCTCGAATTCGCGGCCGCTTCTAGATGGCCGGCGGTGGTTCTGCCGGTGGCTCCGGTTCTGGCTCCAGCGGTGGCAGCTCTGGTGCGTCCGGCACGGG
TACTGCGGGTGGCACTGGCAGCGGTTCCGGTACTGGCTCTGGCACCGGTAATACTAGTAGCGGCCGCTGCAGGGTACC-3´
This
time the order was well planned but it
took over 4 weeks to get them. Unfortunately the genes arrived too
late. Once
again we have to make all the cloning steps. As of mid October 2009 the
project
is still in progress.