Figure Legends
Figure 1: Yeast/mammalian cell-cell selection platform layouts.
(a) Schematic of yeast surface display of an scFv (left) and mammalian
cell expression of a target MP (right). (b) In the biofloating platform,
yeast and mammalian cells are co-incubated in suspension. (c) In the
biopanning platform, yeast are incubated on top of mammalian cell
monolayers.
Figure 2: Effects of scFv affinity on kinetics of
yeast/mammalian cell interactions for biofloating versus biopanning
platforms. Yeast expressing the anti-PD-L1 atezolizumab (High), D12
(Medium), or A1 (Low) scFvs or the anti-PD-1 (control) scFv were
co-incubated with PD-L1+ CHO-K1 cells for varying
lengths of time. (a) Flow cytometry histogram overlays depicting yeast
binding (as measured by cmyc detection) within the mammalian cell
population via biofloating for various incubation timepoints. (b) Phase
contrast microscopy images of yeast binding to mammalian cells via
biopanning for various incubation timepoints. (c) Quantification of
biofloating results shown in (a), displaying the percentage of mammalian
cells that are bound to yeast cells at various timepoints. (d)
Quantification of biopanning results shown in (b), presenting the number
of bound yeast cells per image at various timepoints.
Figure 3: Effects of scFv affinity on equilibrium
yeast/mammalian cell interactions for biofloating versus biopanning
platforms. Yeast expressing the anti-PD-L1 atezolizumab (High), D12
(Medium), or A1 (Low) scFvs or the anti-PD-1 (control) scFv were
co-incubated with PD-L1+ CHO-K1 cells over a range of
yeast:mammalian cell ratios. (a) Flow cytometry histogram overlays
showing yeast binding (as measured by cmyc detection) within the
mammalian cell population via biofloating for various yeast:mammalian
cell ratios. (b) Phase contrast microscopy images of yeast binding to
mammalian cells via biopanning for various yeast:mammalian cell ratios.
(c) Quantification of the data shown in (a), displaying the percentage
of mammalian cells that are bound to yeast cells at various
yeast:mammalian cell ratios. (d) Quantification of the results shown
(b), displaying the number of bound yeast cells per image at various
yeast:mammalian cell ratios.
Figure 4: Effects of avidity on kinetics of yeast/mammalian
cell interactions for biofloating versus biopanning platforms.
Atezolizumab (PD-L1-specific) or nivolumab (control) scFv-expressing
yeast were co-incubated with PD-L1+ CHO-K1 (Dense),
H2444 (Medium), or MDA-MB-231 (Sparse) cells for varying lengths of
time. (a) Flow cytometry histogram overlays depicting yeast binding (as
measured by cmyc detection) within the mammalian cell population via
biofloating for various incubation timepoints. (b) Phase contrast
microscopy images of yeast binding to mammalian cells via biopanning for
various incubation timepoints. (c) Quantification of biofloating results
shown in (a), displaying the percentage of mammalian cells that are
bound to yeast cells at various timepoints. (d) Quantification of
biopanning results shown in (b), presenting the number of bound yeast
cells per image at various timepoints.
Figure 5: Effects of avidity on equilibrium yeast/mammalian
cell interactions for biofloating versus biopanning platforms.
Atezolizumab (PD-L1-specific) or nivolumab (control) scFv-expressing
yeast were co-incubated with PD-L1+ CHO-K1 (Dense),
H2444 (Medium), or MDA-MB-231 (Sparse) cells over a range of
yeast:mammalian cell ratios. (a) Flow cytometry histogram overlays
showing yeast binding (as measured by cmyc detection) within the
mammalian cell population via biofloating for various yeast:mammalian
cell ratios. (b) Phase contrast microscopy images of yeast binding to
mammalian cells via biopanning for various yeast:mammalian cell ratios.
(c) Quantification of the data shown in (a), displaying the percentage
of mammalian cells that are bound to yeast cells at various
yeast:mammalian cell ratios. (d) Quantification of the results shown
(b), displaying the number of bound yeast cells per image at various
yeast:mammalian cell ratios.
Figure 6: Enrichment of a specific clone spiked into a naïve
yeast-displayed scFv library using suspension cell-based versus adherent
cell-based selections. (a) Suspension cell-based selection scheme using
MACS wherein biotinylated target antigen-expressing mammalian cells are
incubated with a yeast-displayed scFv library and immobilized within a
magnetic column via streptavidin-coated magnetic beads to allow for
enrichment of target antigen-specific yeast. (b) Analysis of the yeast
population following each round of evolution performed using the
suspension cell-based selection scheme shown in (a). A naïve
yeast-displayed scFv library doped with 1:1000 PD-L1-specific
atezolizumab scFv-displaying yeast was evolved against
PD-L1+ CHO-K1 cells. Enrichment of atezolizumab
scFv-expressing yeast (red) and depletion of other scFvs (blue) was
monitored by flow cytometry. (c) Adherent cell-based selection scheme
wherein a yeast library is incubated over mammalian cell monolayers and
washed to allow for enrichment of target antigen-specific yeast. (d)
Analysis of the yeast population following each round of evolution
performed using the adherent cell-based selection paradigm shown in (c).
A naïve yeast-displayed scFv library doped with 1:1000 PD-L1-specific
atezolizumab scFv-displaying yeast was evolved against
PD-L1+ CHO-K1 cells. Enrichment of atezolizumab
scFv-expressing yeast (red) and depletion of other scFvs (blue) were
monitored by flow cytometry.
Figure S1: Expression of PD-L1 on various mammalian cell lines.
Flow cytometry histograms depicting PD-L1 levels on the
PD-L1+ CHO-K1 (left), H2444 (middle), and MDA-MB-231
(right) cell lines. Staining with fluorescently labeled isotype control
antibody (red) and anti-PD-L1 antibody (blue) are presented. Note that
antigen expression is uniform on each cell line.
Figure S2: Selective binding of anti-PD-L1 scFv-expressing
yeast clones to PD-L1+ but not
PD-L1- CHO-K1 cells on biofloating and biopanning
platforms. Yeast expressing the anti-PD-L1 atezolizumab (High), D12
(Medium), or A1 (Low) scFvs were co-incubated with
PD-L1- CHO-K1 or PD-L1+ CHO-K1 cells
for 180 min. (a) Flow cytometry histogram overlays depicting yeast
binding (as measured by cmyc detection) within the mammalian cell
population via biofloating. (b) Phase contrast microscopy images of
yeast binding to mammalian cells via biopanning.