Molecular Cell, Volume
84
Supplemental information
Denaturing puri
fi
cations demonstrate that PRC2
and other widely reported chromatin proteins
do not appear to bind directly to RNA
in vivo
Jimmy K. Guo, Mario R. Blanco, Ward G. Walkup IV, Grant Bonesteele, Carl R.
Urbinati, Abhik K. Banerjee, Amy Chow, Olivia Ettlin, Mackenzie Strehle, Parham
Peyda, Enrique Amaya, Vickie Trinh, and Mitchell Guttman
A
B
Tagged
Endogenous
*
SUZ12
EZH2
Tx
UTx
Tx
UTx
*
*
PTBP1
Tx
UTx
*
250
EED
Tx
UTx
*
75
50
100
150
37
0
2
4
6
8
Fold enrichmen
t
(
T
agged/Endogenous
)
Protein quantification
EED
EZH2
SUZ12
PTBP1
D
(kDa)
Imaged: Halo-ligand (AlexaFluor-660)
Halo-EZH2-V5
Halo-SUZ12-V5
Halo-EED-V5
250
150
100
75
IgG
Endogenous
EZH2
Endogenous
SUZ12
EED
EZH2
SUZ12
EED
SUZ12
EED
EZH2
IP
Expressed
(Halo-V5)
C
250
150
100
75
50
1
1
2
αV5 WB
Halo-EZH2-V5
1
2
αV5 WB
Halo-SUZ12-V5
2
αV5 WB
Halo-EED-V5
αV5 WB
1
2
L
1
2
Halo-PTBP1-V5
POI
POI
Halo
V5
αV5 WB
Untransfected
Figure S1. Controls for tagged PRC2 components (Related to Figure 1).
(A)
Western blots performed against endogenous proteins within HEK293T cell lysates that were transfected with Halo-V5-tagged proteins
(PTBP1, EED, EZH2, SUZ12). (*) denotes the molecular weight of the tagged version of the protein of interest (POI) (endogenous protein + ~33
kDa HaloTag + ~1.5 kDa V5 tag), black arrow depicts the molecular weight of the endogenous protein. “Tx” and “UTx” refer to transfected and
untransfected cells, respectively.
(B)
Quantification of (A), depicting fold enrichment of tagged protein expression relative to endogenous protein.
(C)
Western blots performed against V5 epitope after CLIP captures (via heat elution) of tagged PRC2 components (EED, EZH2, SUZ12, PTBP1)
shown in replicate. Black arrows depict the molecular weight of the tagged proteins. An untransfected control was subject to V5 IP and run in
replicate on the same blot as PTBP1 (and SUZ12, which is cropped) to demonstrate the specificity of the V5 antibody.
(D)
Endogenous PRC2 components (EZH2 or SUZ12) were immunoprecipitated from cell lysates expressing Halo-EED-V5, Halo-EZH2-V5, or
Halo-SUZ12-V5 protein. The amount of tagged protein that was associated with the endogenous protein was visualized using a fluorescently
labeled Halo-ligand (AlexaFluor-660) on a gel.
K
D
= N/A
Time (Seconds)
Response (RU)
Halo-
GFP
-V5
K
D
= 3.9E-08 M
Time (Seconds)
Response (RU)
Halo-
GFP-3xLamdaN
-V5
K
D
= 6.0E-08 M
Time (Seconds)
Response (RU)
Halo-
PTBP1
-V5
K
D
= 2.57 E-07 M
Time (Seconds)
Response (RU)
Halo-
EED
-V5
K
D
= 5.6E-05 M
Time (Seconds)
Response (RU)
Halo-
EZH2
-V5
K
D
= 1.8E-06 M
Time (Seconds)
Response (RU)
Halo-
SUZ12
-V5
Figure S2. Tagged PRC2 components bind to RNA
in vitro
(Related to Figure 1).
HaloTag fusion proteins were immobilized on a Biacore chip functionalized with HaloTag ligand.
In vitro
transcribed RNA was injected and the
affinity of RNA for HaloTag fusion proteins was measured. RNA substrates were the Maltose Binding Protein (MBP) fused with five tandem
repeats of BoxB RNA hairpin or the Xist A-repeat. Sensorgrams for negative control (HaloTag-GFP-V5) and Halo-tagged proteins of interest
(GFP-3xLambdaN, PTBP1, EED, EZH2, SUZ12) are shown. Panels show one data set (colored lines) fit globally to a 1:1 Langmuir binding
interaction model with bulk refractive index (RI).
A
B
32
P RNA
Protein
250
150
100
75
Halo-EED-V5 CLI
P
-UV
+UV
RNase
RNase
250
150
100
75
32
P RNA
Protein
Halo-EZH2-V5 CLI
P
RNase
RNase
-UV
+UV
250
150
100
75
50
32
P RNA
Protein
Halo-SUZ12-V5 CLI
P
RNase
RNase
-UV
+UV
Untransfected
(
32
P
)
RNase
32
P RNA
250
150
100
75
50
+UV
C
D
250
150
100
75
50
+UV
-UV
RNase
RNase
32
P RNA
Protein
Halo-PTBP1-V5 CLIP
0
0.001
0.002
0.003
0.004
Local maximum
50
75
100
150
250
Fraction of
32
P intensity
Halo-PTBP1-V5 CLIP RNase Titration
0.005
Size (kDa)
High
Mid
Low
Halo-PTBP1-V5 CLIP
-100 nts
+100 nts
HYUUUYU
1
0
2
3
Crosslink-induced
truncation frequency
Figure S3. Visualization of RNA-protein complexes purified by CLIP at different RNase concentrations (Related to Figure 1).
(A, B)
CLIP was performed from cells expressing proteins tagged with N-terminal Halo and C-terminal V5 tags with (+UV, right) and without
UV-crosslinking (-UV, left). RNase I titration was carried out to resolve RNA-protein complexes at RNase I dilutions of 1:50, 1:3,000, and 1:50,000.
(A) Tagged EED, EZH2, SUZ12, and (B) untransfected lysates were captured and visualized using the same CLIP conditions. Captured RNA was
radiolabeled (
32
P), run on an SDS-PAGE gel, and transferred to nitrocellulose membrane. Red arrow indicates the expected size of the immuno
-
precipitated protein. Protein inputs are shown for each lane by Halo-ligand labeling (below).
(C)
(Left) Same as in (A, B), but with tagged PTBP1. (Right) Quantification of
32
P Halo-PTBP1-V5 CLIP intensity for each RNase condition (High,
Mid, Low, corresponding to 1:50, 1:3,000, and 1:50,000) measured by ImageJ (y-axis), integrated across the length of the gel which is scaled to
expected molecular weights (x-axis).
(D)
Crosslink-induced truncation frequency (percentage) of Halo-PTBP1-V5 CLIP relative to known PTBP1 motif (HYUUUYU, shown in blue).