Cell Reports, Volume 14

Supplemental Information

Alternative Splicing of G9a

Regulates Neuronal Differentiation

Ana Fiszbein, Luciana E. Giono, Ana Quaglino, Bruno G. Berardino, LorenaSigaut, Catalina von Bilderling, Ignacio E. Schor, Juliana H. Enriqué Steinberg, MarioRossi, Lía I. Pietrasanta, Julio J. Caramelo, Anabella Srebrow, and Alberto R. Kornblihtt

Figure S1

N2a Cells qPCR B

siLUC siE10 siG9a A

D

Figure S2

C

B

A

RA 2 days

RA 2 days +

control 2 days RA 2 days +

BIX 2 days

Figure S3

A

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

1

1,1

A B C D E F G H I J K

Dif

Undif

0

0,2

0,4

0,6

0,8

1

1 2

nC

hIP

: IP

H3

K9

me

2/H

3to

tal

Promoter Ex 4 Ex10 Ex 7 Ex 14 Ex 22

Ex 27

A B C D E F G

H I J K

Intergenic

nC

hIP

: IP

H3

K9

me

2/H

3to

tal

B C

CpG Island

- DMSO

D

RA

-

Figure S4

- DMSO

N2a cells

E9 E10 E11

Globin

Promoter

I9 I10 I globin

E globin

B

A

Figure S5

Figure S6 A B

Figure S7

- DMSO

ED1+

ED1-

Exon 25

Intergenic region

Promoter

Exon 1 Exon 20 Exon 33 (ED1)

Exon 47

0,

4,

8,

12,

16,

Promoter Exon 10 Exon 20 Exon 28 Exon 35 Exon 45

(H3K

9m

e2

-IgG

)/In

put

rel to

HP

RT

E2

FN Differ

Undif -

DMSO

1

Supplemental Information

Supplemental Figure Legends

Figure S1. Selectively down regulation of G9a E10+ isoform does not affect G9a levels. Related to

Figure 3

N2a cells were differentiated by treatment with 1 µM RA (RA) or maintained undifferentiated (-) and

transfected with siLUC as a control, siE10 or siG9a previously to differentiation induction. (A) mRNA

levels of total G9a were assessed by RT-qPCR and relativized to HSPCB. (B) Pictures are shown and

levels of total G9a fluorescence intensity were measure after immunofluorescence.

Figure S2. Neuronal differentiation triggers a nuclear accumulation of G9a protein. Related to Figure 5

N2a cells were differentiated with 2% (v/v) DMSO or 1 µM RA for 6 or 3 days, respectively. Cells

were then fixed and G9a protein localization was assessed by indirect immunofluorescence. (A and B)

Plots show total fluorescence intensity of G9a in the nuclear and cytoplasmic compartments for N2a

cells differentiated with DMSO (A) or RA (B). (C) HaCaT, HeLa, MCF7 and T47D cells were fixed

and G9a protein localization was assessed by indirect immunofluorescence. Alternative splicing of the

G9a gene was assessed by radioactive RT-PCR.

Figure S3. BIX does not inhibit differentiation when cells were treated after the differentiation program

was already established. Related to Figure 7

N2a cells were differentiated by treatment with 1 µM RA for 2 days and then 3 µM BIX or vehicle was

added. The percentage of cells displaying neurite outgrowth (differentiated cells) was determined and

representative pictures are shown. Scale bars represent 10 µm.

Figure S4. Chromatin structure regulates G9a alternative splicing. Related to Figure 7

(A) Schematic diagram of G9a showing the exon-intron organization and the distribution of qPCR

amplicons (top). N2a cells were differentiated with RA for 3 days and H3K9me2 levels at the indicated

regions were determined by nChIP assay followed by qPCR (bottom). Values of two independent

immunoprecipitations, relativized to the mean value for total H3, are shown for each region. An

amplicon located at an intergenic region (left) was used as a control. (B-C) N2a cells were

2

differentiated with 2% DMSO for 6 days and treated with 5 µM 5-azacytidine (B) or 2 ng/µl

trichostatin A (C) for 48 and 16 h, respectively. Cells were then harvested, RNA was extracted and

alternative splicing of the G9a gene was assessed by radioactive RT-PCR. Statistical analysis

correspond to two-way ANOVA, Bonferroni post hoc test (*) p<0.05, (***) p<0.0001. (D) N2a cells

were differentiated with DMSO for 6 days and DNA methylation of a CpG island close to the

alternative exon 10 (red box) was determined by sodium bisulfite modification followed by DNA

sequencing. Black circles represent methylated CpG dinucleotides.

Figure S5. Alternative splicing elicited by a G9a minigene is modulated during neuron differentiation.

Related to Figure 7

(A) Schematic diagram of the G9a minigene. (B) N2a cells were transfected with the minigene vector

and differentiated with 2% DMSO for 6 days. Cells were harvested, RNA was extracted and alternative

splicing was assessed by radioactive RT-PCR as described.

Figure S6. Position of G9a (EHMT2) and SUV39H1 in a functional interactions splicing network

(Papasaikas et al., 2015). Related to Discussion

Connections between factors indicate consistent effects of their knockdowns on an extensive set of

alternative splicing events, suggesting mechanistic associations in splicing regulation. Positive or

negative functional interactions are represented by green or red edges, respectively. Edge thickness

indicates the strength of the functional interaction, while node size is proportional to the overall impact

of a given knockdown in the regulation of alternative splicing. Node coloring corresponds to the

natural separation of the network in distinct, coherent modules. Black dotted lines represent

known physical interactions as reported in the STRING database (Franceschini et al., 2013). Factors

implicated in chromatin structure modification are shown with red labeling.

Figure S7. Chromatin structure and alternative splicing of Fibronectic gene during neuron

differentiation. Related to Discussion

Schematic diagram of Fibronectin showing the exon-intron organization (top). N2a cells were

differentiated with DMSO for 6 days. Cells were harvested, RNA was extracted, and alternative

splicing of the Fibronectin gene was assessed by radioactive RT-PCR. H3K9me2 levels at the

3

indicated regions of Fibronectin gene were determined by nChIP assay followed by qPCR. Values of

two independent immunoprecipitations are shown for each region.

Supplemental Experimental Procedures

Bisulfite sequencing

Unmethylated cytosine was converted to uracil by sodium bisulfite using the EpiTect Bisulfite Kit

(Quiagen). Two µl of the converted samples were used as PCR template for amplification with primers

designed with the online tool Bisearch. PCR products were sequenced using a BI 3130xl Genetic

Analyzer (Applied Biosystems).

nChIP

Native chromatin immunoprecipitation was performed as previously described (Schor et al., 2009). For

each immunoprecipitation we used 10 µg of H3K9me2 (07-441) antibody from Millipore or 5 µg of

H3total (07-690) antibody from Millipore. Protein A agarose beads, pre-blocked with salmon sperm

(Millipore), were used to recover the immuno-complexes. DNA was purified using QIAquick

purification kit (QIAGEN) and quantitative PCR analysis was performed using SYBR green.

Immunoprecipitated chromatin was normalized to input chromatin, and H3 total amount. All primer

sequences and real time PCR conditions are available upon request.