84
Combinatorial Phage Antibody Libraries
has been used successfully to display
scFvs on their surface. Potential advan-
tages of yeast display are that flow cy-
tometry can be used to sort the positive
and negative binding clones, and equilib-
rium binding experiments have shown a
good correlation between binding af±nity
and selection allowing ±ne discrimina-
tion between clones of similar af±nity.
Ribosome display has also been used to
select antibodies in a totally cell-free sys-
tem. The potential advantages of ribosome
display are the large and diverse reper-
toires that can be achieved (up to
10
14
)
and flexibility in selection conditions. The
principles of selection for virtually all
combinatorial antibody libraries are sim-
ilar, so because phage display is the
most common mode of display, we will
next consider the selection from antibody
phage libraries.
3
Selection from Antibody Phage Libraries
The strength of phage display lies in
the ability to select positive clones occur-
ring in a library at frequencies of one
in a million or less. This selection takes
place over consecutive rounds of panning
in which positive phages are progres-
sively enriched against a background of
unwanted clones. In effect, panning is
performed by af±nity puri±cation of an
antibody phage against an antigen immo-
bilized on a solid surface such as a column
matrix or a well in an enzyme-linked
immunosorbent assay (ELISA) setup. Pro-
teins on the surface of whole cells may
also be panned in this manner in tissue-
culture plates or using flow cytometry.
Phages may also be bound to biotiny-
lated antigen in solution and then captured
on streptavidin-coated magnetic beads. In
each case, nonantigen-bound phages are
removed by thorough washing and bound
phages are recovered by elution from the
antigen, generally with acid. Eluted phages
are infected into
E. coli
,amp
l
ed
,and
reselected in further rounds of panning
(see Fig. 4). When monovalent antibody
display is used, typically three to ±ve
rounds of panning will select a popula-
tion of clones with the highest af±nity
for antigen.
The panning process is unlike the an-
tibody response to antigen
in vivo
in
which mutation and recombination occur
simultaneously with selection. Thus, an
antibody library can be screened by alter-
nating antigens in successive rounds to
select for cross-reactive antibodies without
worrying about ‘‘eliciting’’
de novo
anti-
bodies to each new antigen. Moreover, it
is possible to pan with mutant proteins,
unfolded proteins and peptide fragments
that are either unsuitable for immuniza-
tion or are simply not immunogenic. Since
panning is not limited by a requirement
for puri±ed antigen, phage libraries may
be selected against an enormous range
of molecules. Panning experiments per-
formed using lysates of virus-infected cells
and intact cells indicate that the technique
is sensitive to antigen present in small
amounts. Different panning strategies are
detailed in Section 4.
Antibody clones isolated at the phage
level from the ±nal round of panning need
to be examined as soluble Fab to further
investigate their binding properties. In
the pComb3H vector, the transition from
phage display to soluble expression is
made after removal of the g3 sequence by
restriction digestion and religation of the
vector. Soluble Fab, which accumulates
in
the
periplasmic
space,
is
released
following lysis of the bacteria. If necessary,
recombinant
Fab
may
be
puri±ed
to
previous page 1404 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online next page 1406 Encyclopedia of Molecular Cell Biology and Molecular Medicine read online Home Toggle text on/off