www.pnas.org/cgi/doi/10.1073/pnas.
1807449
115
Supplemental Figures
:
Figure S1
–
The design of HcRed7. Two core libraries targeted unique structural regions of the protein core. The first
region (
green
) surrounds the phenolate side chain of the chromophore. The second (
yellow
), is a very highly mut
ated
region in RFP monomer evolution. This region holds an internal water channel, key catalytic residues, and abuts the AC
interface.
Figure S
2
–
The diversity of
spectroscopic properties of 41 colored variants from
the HcRed core libraries cLibA an
d
cLibB. Brightness is plotted against maximum intensity emission wav
elength. RFPs with emission peaks at longer
wavelengths are
noticeably dimmer than some of the
bright,
hypsochromically shifted RFPs.
Figure S
3
–
Consensus design compares favorably
to error
-
prone mutagenesis in the improvement of fluorescence in
HcRed7 ∆6. The two libraries were compared by screening in 96
-
well plates with ~4,000 screened variants from the
consensus design library and ~8,000 variants screened from the error
-
prone li
brary. Both individual variants and the
population from the consensus library outperform the error
-
prone library, although both contributed valuable variants
to the engineering process.
Figure S
4
–
A map displaying
the intermolecular contacts
made by
each residue
of HcRed
in the
AC
interface
. The map
is
color coded by
the
frequency
with which the residues are mutated during past instances of fluorescent protein
monomerization
. Residues 146, 159,
167, 168,
170,
174,
191, 193
, 197, 201, and 214
were
mut
ated during mGinger
engineering
. Residues 222
-
223 were deleted as part of the C
-
terminal tail.
4
3
2
1
0
6
5
7
Color-coding
for mutational
frequency per
site
Figure S
5
–
Spectral scans of the monomeric fluorescent proteins mGinger1, mGinger2, mKelly1, and mKelly2, along with
the dimeric proteins they were d
erived from.
Figure S
6
–
Mutations during mGinger and mKelly engineering categorized by solvent
-
accessible surface area (SASA). (A)
The number of new mutations introduced in each step of protein engineering. (B) The cumulative number of mutations
that
separate each variant from the wild
-
type progenitor (HcCP in the case of HcRed variants and eqFP578 in the case of
mCardinal variants).
Figure S
7
–
Mutations during mGinger
1
and mKelly
1
engineering categorized by
structural clustering. The tail and AC
i
nterface are shown in
r
ed in spheres, as they both participate in the AC interface, but are broken out separately in the
pi
-
charts. SAP refers to mutations that occur near positions with high surface aggregation propensity (SAP). Many
mutations cluster int
o more than one structural region, and in th
is
case they are categorized hierarchically from top to
bottom according to the legend of the pi
-
chart. So Tail Deletion residues are not included in the AC Interface count, and
AC interface residues are not incl
uded in the SAP residue count, and so on.
Figure S
8
–
Maturation kinetics of the monomeric fluorescent proteins mGinger1, mGinger2, mKelly1, and mKelly2,
along with the dimeric proteins they were derived from.
Figure S
9
–
Photobleaching kinetics of
the monomeric fluorescent proteins mGinger1, mGinger2, mKelly1, and mKelly2,
along with the dimeric proteins they were derived from.