Biomedical Engineering Reference
In-Depth Information
7.5 EXPRESSION AND PURIFICATION OF
MONOMERIC Fc-FUSION PROTEINS
been expressed in mammalian cells, namely CHO and
HEK293, which are frequently used cell lines for protein
expression [12,31,32,61].
Two proteins (effector protein-Fc fusion and Fc alone) are
produced using the monomeric Fc-fusion protein expression
system described above. After protein folding and the
formation of disulfide bonds between the Fc moieties, three
forms of products are secreted in the cell culture media, Fc
fusion dimer (effector protein-Fc homodimer), Fc fusion
monomer (one effector protein-Fc and one Fc), and Fc dimer
(Fc homodimer) (Figure 7.6) [12,61]. Because the Fc fusion
monomer is the only desired product, it is important that the
expression system is tailored to minimize the expression of
the other two products (Fc fusion homodimer and Fc homo-
dimer). It can be challenging to separate Fc fusion monomer
from Fc fusion homodimer during purification because of
the relatively similar biochemical properties of these two
molecules, while the Fc dimer can generally be more readily
separated. Over-expression of Fc has been shown to increase
the ratio of Fc fusion monomer to Fc fusion homodimer, and
thereby facilitates purification. Depending on the activity of
Fc fusion homodimer, it can contribute significantly to the
activity assay of the Fc monomer fusion cell culture harvest
when the expression level is analyzed by activity assays,
which may pose additional challenges. Analytical size exclu-
sion chromatography HPLC or nonreducing SDS-PAGE can
be helpful tools to analyze and quantify the monomer Fc
fusion expression levels in the cell culture, and can be coupled
to protein A chromatography to facilitate detection.
A flexible polypeptide linker can also be designed to
connect the two Fc fragments so that the monomer Fc-fusion
protein is expressed as a single protein. This approach
eliminates the need to separate the Fc fusion monomer
from the Fc fusion homodimer, and can lead to more
efficient expression. If desired, intracellular cleavage sites
can be added flanking the linker peptide so that the mono-
meric Fc protein is translated as a single polypeptide and
before secretion, processed intracellularly at the cleavable
linker to produce Fc-effector protein and Fc heterodimer
(the monomer Fc-fusion protein). This may require co-
transfection of additional processing enzymes to fully
remove the linker from the final proteins.
Similar to traditional dimeric Fc-fusion proteins, the
presence of Fc provides an advantage of affinity chroma-
tography purification for Fc monomeric fusion proteins.
Monomeric Fc-fusion proteins can usually be purified
using Protein A or Protein G affinity chromatography.
Both Protein A and Protein G are immunoglobulin binding
proteins found in Streptococcal bacteria. While Protein G
binds to all four subclasses of human IgG, Protein A binds
well to IgG1, IgG2, and IgG4. The binding of IgG to
Protein G is often stronger, making elution, and complete
recovery of the immunoglobulin more difficult. Addition-
ally, because of the lower cost of Protein A compared to
The Fc fragment used as fusion partner usually contains the
hinge region and CH2 and CH3 domains of immunoglobulin
heavy chain. For therapeutic purpose, most Fc-fusion pro-
teins employ the Fc fragment of human IgG1. All seven
currently marketed Fc-fusion proteins contain IgG1 Fc
(Table 7.2), while Fc fragments of human IgG2 and IgG4
have also been explored as fusion partners in development
[80-82]. The Fc domain of human IgG3 is not a preferred
fusion partner, presumably in part because of its high
complement activation property. Depending on the applica-
tion and structure-function of the Fc-fusion protein, the Fc
fragment is commonly coupled at the C-terminus and some-
times the N-terminus of the effector proteins. In some cases,
a small spacer (e.g., a glycine- and serine-rich peptide)
between Fc and the effector protein is helpful in preserving
the function of the effector protein. In contrast to the
expression of traditional dimeric Fc-fusion proteins, which
require one expression cassette for the effector protein-Fc
fusion, expression of monomeric Fc-fusion proteins usually
employs two expression cassettes, one controls the expres-
sion of the effector protein coupled with Fc and the other
controls expression of Fc alone. The two expression cas-
settes can be built within one bi-cistronic vector with two
expression cassettes, or in two separate vectors.
Several expression systems have been used for Fc-fusion
protein expression. For therapeutic purposes, most tradi-
tional dimeric Fc-fusion proteins have been produced in
mammalian expression systems. Six out of the seven Fc
fusion molecules on the market are expressed in CHO cells,
while only romiplostim is expressed in Escherichia coli
[83-89] (Table 7.2).
Expression of Fc-fusion proteins in insect cells (such as
Sf9) and yeast systems (such as Pichia pastoris) has also
been reported [90,91]. The E. coli expression system pro-
vides high productivity but it cannot carry out post-
translational modifications such as glycosylation and may
not form disulfide bonds and proper conformation. Refold-
ing of Fc-fusion proteins produced from E. coli is often
necessary and it can result in scrambled disulfide forms and
other misfolded conformations [92,93]. Proteins expressed
in yeast cells can form disulfide bonds and are glycosylated,
however, glycans made by yeast cells contain high mannose
structures, have different oligosaccharide structure from
human and are often immunogenic to humans [94]. Over
the last decade, however, significant progress has been
achieved through genetic engineering of glycosylation path-
ways in yeast cells for human like glycosylation [94,95].
Expression of Fc-fusion proteins in mammalian systems
provides advantages such as correct protein folding,
disulfide bond formation, and proper posttranslational mod-
ification. All monomeric Fc-fusion proteins reported have
Search WWH ::
Custom Search