% IMPORTANT: The following is UTF-8 encoded. This means that in the presence
% of non-ASCII characters, it will not work with BibTeX 0.99 or older.
% Instead, you should use an up-to-date BibTeX implementation like “bibtex8” or
% “biber”.
@PHDTHESIS{Mller:956251,
author = {Müller, Carolin},
othercontributors = {Oldiges, Marco and Blank, Lars M.},
title = {{I}nvestigation of protein secretion in microscale
cultivation systems with novel tools},
school = {RWTH Aachen University},
type = {Dissertation},
address = {Aachen},
publisher = {RWTH Aachen University},
reportid = {RWTH-2023-04029},
pages = {Online-Ressource : Illustrationen, Diagramme},
year = {2023},
note = {Veröffentlicht auf dem Publikationsserver der RWTH Aachen
University; Dissertation, RWTH Aachen University, 2023},
abstract = {Until today, it is impossible to predict a suitable signal
peptide for Sec-type secretion of heterologous proteins in
Gram-positive bacteria. Instead, signal peptides have to be
tested for each host and target protein under process
conditions. In addition, the ribosome binding site and in
particular the spacer between the Shine-Dalgarno sequence
and the start codon of the signal peptide can also affect
protein secretion. To accelerate the identification of
suitable combinations of signal peptides and target
proteins, automated workflows for targeted strain
construction and high-throughput screening for heterologous
protein secretion in Corynebacterium glutamicum were
established, which can be easily adapted to different target
proteins. A plasmid library with different Bacillus subtilis
signal peptides was constructed in the newly designed
pPBEx2-based plasmid pCMEx12, which allows the exchange of
the ribosome binding site including the spacer sequence as
well as the signal peptide sequence by cassette mutagenesis.
In this method, the inserts are provided as hybridized
oligonucleotides that are not fully complementary, but have
overhangs that can be ligated to the restriction digested
backbone. For target protein secretion with pCMEx-based
plasmids, a reporter gene coding for a blue chromoprotein
under the control of a constitutive promoter can be
exchanged with the gene of interest by Golden Gate assembly,
combining restriction and ligation in a one-pot setup. Since
the chromoprotein leads to blue colonies after
transformation, successful cloning can be detected by a
change in colony color from blue to white, in addition to a
restriction enzyme digest in which the number and size of
DNA fragments depend on the insert. The gene of interest is
then expressed in frame with an amino-terminal signal
peptide and carboxyl-terminally linked to the 11th β-sheet
of the green fluorescent protein (GFP, GFP11-tag) under the
control of the inducible tac promoter. The molecular cloning
steps of the Golden Gate assembly, the Escherichia coli
heatshock transformation, the plasmid purification and the
restriction digest were automated using the Opentrons OT-2
liquid handling robot with integrated Temperature or
Magnetic Module. This reduced the process time for molecular
cloning to about $58\%$ of that for the manual process. For
testing cultivation workflows with online product
monitoring, the C. glutamicum pPBEx2-PhoDCg-GFP enabling
tightly controlled induction of GFP secretion was
successfully prepared. An automated high-throughput
screening workflow was developed on a Tecan Freedom EVO®
robotic platform with an integrated centrifuge, microplate
reader, and BioLector® for microscale cultivation with
online measurement of the backscatter signal that correlates
to cell dry weight. Automated preculture handling and
backscatter-triggered inoculation of main cultures and
induction ensure high comparability of bacterial growth.
Using this workflow, suitable combinations of ribosome
binding sites and B. subtilis signal peptides were
identified for Fusarium solani f. sp. pisi cutinase-GFP11
secretion by C. glutamicum. Cutinase-GFP11 in the
cultivation supernatant was detected 4 h after induction via
activity measurement and activity-independently by assembly
of the GFP11-tag with GFP1-10 in added detector solution by
holo-GFP fluorescence in split GFP assay. The process time
from cultivation of up to 12 different strains to detection
of the target protein in the supernatant is about 1.5 days,
with manual operations only required at the start of
cultivation and prior to the assays. For high-throughput
screening approaches, sufficient quantities of detector
solution is needed. Therefore, a fed-batch cultivation
process for the GFP1-10 production in laboratory-scale
bioreactors was established and detector solution for up to
385 screenings in 96-well plates could be obtained. Aspects
of GFP1-10 detector protein stability, storage and assay
incubation conditions have been investigated. After a
proof-of-concept using the secretion model protein cutinase,
a B. subtilis signal peptide screening for secretion of
polyethylene terephthalate degrading enzymes leaf-branch
compost cutinase (LCC) and PE-H was conducted. For this, the
cultivation workflow was optimized to allow the comparison
of up to 24 different strains in one run. Using a process
model combining Bayesian inference and Thompson sampling,
the best of 24 signal peptides was identified with a
probability of $80\%$ for the PE-H and $75\%$ for the LCC
after only three batch cultivations.},
cin = {162610 / 160000},
ddc = {570},
cid = {$I:(DE-82)162610_20140620$ / $I:(DE-82)160000_20140620$},
typ = {PUB:(DE-HGF)11},
doi = {10.18154/RWTH-2023-04029},
url = {https://publications.rwth-aachen.de/record/956251},
}
Gast ::