We focused our project on coupling and optimizing the characteristics
of a restriction endonuclease with short oligonucleotides to develop a
programmable and highly specific enzyme-oligo-complex. As a restriction
endonuclease we chose the cleavage domain of the well studied
endonuclease FokI from Flavobacterium okeanokoites. Normally FokI acts
as a homodimer, each dimer divided in cleavage and restriction domain.
Chandrasegaran and Miller have already made experiments to uncouple the
cleavage and restriction domains of FokI and created a novel
site-specific endonuclease by linking the cleavage domain to zinc
finger proteins.
For our project we generated two Fok heterodimers (Miller, Nature
biotech, 2007). For the catalytic active Fok partner, named Fok_a, the
first 1158 nucleotides, i.e. the recognition domain, were deleted and
glutamate 490 was switched to lysine (GAA->AAA) as well as
isoleucine 538 to lysine (ATC->AAA) for the heterodimer
formation.
For the catalytic inactive Fok partner, named Fok_i, the heterodimeric
amino acids glutamine 486 was switched to glutamate (CAA->GAA)
and
isoleucine 499 to leucine (ATC->CTG) and the catalytic amino
acids
aspartate 450 was switched to alanine (GAC->GCG) and aspartate
467
to alanine (GAT->GCG).
|
Association of linker FluA and Dig with
DNA and Fok_a
and Fok_i monomers |
The two heterodimeric
partners were fused to different anticalins
binding different adapter molecules. Thus Fok_i is fused to anticalin
on Fluorescein and Fok_a to anticalin on Digoxigenin. These adapter
molecules are linked to oligonucleotides mediating the binding of the
DNA site of interest. Now the heterodimerization comes into play. If
the different Fok_i and Fok_a constructs bind their target oligos and
come together, the inactive domain will serve simply as an activator of
the active domain, cutting only one strand of the DNA. In our 3D models
we showed that Fok domains are positioned in such a way that Fok_a will
cut the DNA and Fok_i the modified oligonucleotide. Thus the
inactivation of Fok_i allows the reuse of our oligonucleotides.
Different linkers were designed and fused between cleavage domain and
binding protein to test the optimal distance to preserve the most
possible flexibility and most possible precision of the heterodimeric
Foks.
|
complete universal restriction enzyme.
Blue: DNA strand; Red: two 16bp
long Oligos, tags as indicated in the
picture. Ochre: Fluorescein A
binding lipocalin; Orange: digoxigenin binding
lipocalin; Yellow:
tagged Base with C6 linkers and attached tags; light Blue: inactive
FokI cleavage domain; Red: FokI active
cleavage domain; Green: three
catalytically active aminoacids; White: two Calcium ions |
The place where all the
cutting events will take place in our scenery
is the thermocycler. Therefore all the ingredients, i.e. chromosomal or
plasmid DNA of interest, modified oligonucleotides and the different
heterodimers Fok_a and Fok_i are mixed together. At high temperatures
the DNA will denaturate allowing the different partners to find each
other, cut the DNA, and fall apart in course of one thermocycle. The
whole procedure can be repeated undefinately. To reach this step, we
need to improve the thermostability of our enzyme via phage display.
Creating a universal restriction enzyme provides not only the
possibility to improve routine cloning but also to enhance therapeutic
gene repair via triplex technology. Many genetic diseases and
especially ones arising from single nucleotide polymorphisms (SNPs) or
monogenetic disease can be alleviated by the replacement of mutated
genes using this method. To cut double stranded DNA the
oligonucleotides have to be replaced by triple helix forming oligos
(TFO). They can bind double-stranded DNA in homopurin- or
homopyrimidine-rich areas. But developments are also made to widen the
possible interaction domains of the DNA and hence make the TFOs as
programmable as our conventional oligonucleotides. In case of the human
genome of 3×10^9 bp size, a highly specific artifical
endonuclease
would be necessary to address the mutated gene explicitly. The used
TFOs therefore have to possess a minimum length of 16 bp to cut just
once in the human genome (4^16 bp = 4.3*10^9 bp).
Additionally, we provide
an alternative way of binding by the coupling
of the cleavage domain to the transcription factor Fos. This concept
also includes the activity regulation by photo-switching. We are also
modifying an argonaute protein into a DNA endonuclease using phage
display technics.