golgi outpost synthesis impaired by toxic … reports, volume 20 supplemental information golgi...
TRANSCRIPT
Cell Reports, Volume 20
Supplemental Information
Golgi Outpost Synthesis Impaired by Toxic
Polyglutamine Proteins Contributes to Dendritic
Pathology in Neurons
Chang Geon Chung, Min Jee Kwon, Keun Hye Jeon, Do Young Hyeon, Myeong HoonHan, Jeong Hyang Park, In Jun Cha, Jae Ho Cho, Kunhyung Kim, Sangchul Rho, GyuRee Kim, Hyobin Jeong, Jae Won Lee, TaeSoo Kim, Keetae Kim, Kwang PyoKim, Michael D. Ehlers, Daehee Hwang, and Sung Bae Lee
w1118
A
78Q OE
B
Rac1 OE
C
78Q OE + Rac1 OE
DCrebA OE
E
78Q OE + CrebA OE
F
78Q OE +Rac1 OE+ CrebA OE
G
Fig. S1
0 1 2 3 4 5 6 7 8 90
50
100
150
200
250
300
350
400
450Rac1 OE
log2 (length)
#of
term
inal
de
ndrit
es
J<10um >10um
0 1 2 3 4 5 6 7 8 90
50
100
150
200
250
300
350
400
45078Q OE + Rac1 OE
log2 (length)
# of
term
inal
de
ndrit
es
I
<10um >10um
0 1 2 3 4 5 6 7 8 90
50
100
150
200
250
300
350
400
450
Kw1118
log2 (length)
# of
term
inal
de
ndrit
es
<10um >10um
0 1 2 3 4 5 6 7 8 90
50
100
150
200
250
300
350
400
45078Q OE
log2 (length)
# of
term
inal
de
ndrit
es
H
<10um >10um
L w1118 vs. Rac1 OE
log2 (length)
Prob
abili
ty D
ensi
ty
0
0.1
0.2
0.3
0.4
0.5
0.6
0 1 2 3 4 5 6 7 8 90
<10um >10um P<0.001
Figure S1. Rac1 Overexpression Alters the Number and the Length of Terminal Dendrites in the PolyQ-
Expressing C4 da Neurons (Related to Figure 1).
(A-G) Original confocal images of the C4 da neurons used for dendrite analyses in Figures 1, 5A, 5B, 5D, 5E, S1H-
L, and S6. Fluorescence intensity has been non-linearly adjusted for better visualization of the dendrites. For all the
denoted genotypes, n = 6. (H-K) Histograms of the log2(length of terminal dendrites) in 78Q OE (H), 78Q OE +
Rac1 OE (I), Rac1 OE (J), and w1118 (K). Short (<10 μm) and longer (>10 μm) terminal dendrites are partitioned by
a red line. n = 6. (L) Comparison of the probability density functions (PDFs) of log2(length of terminal dendrites) in
C4 da neurons expressing the denoted transgenes: w1118 (black) vs. Rac1 OE (red). Normalized histograms are also
provided. Short (<10 μm) and longer (>10 μm) terminal dendrites are partitioned by a green line.
A
Fig. S2
02468
101214
****
****
Primary + secondary dendrites
# of
GO
Ps
w1118 Ataxin1-82Q OEManII-eGFP
Merged
MJDFL-78Q OE
B
CD4-tdTom
HA-MJD-75QControl
75QControlSpin
e #
per µ
m
0
0.2
0.4
0.6
***
C
D
E
F w1118 78Q OEMJDFL-78Q OEAtaxin1-82Q OE High
Low
CD
4-td
GFP
inte
nsity
HA
Merged
pGolt
Control HA-MJD-75Q
Figure S2. PolyQ Toxicity Reduces the Number of GOPs both in the Fly and the Mammalian Neuronal
System (Related to Figure 2).
(A) Representative images of GOPs labeled by ManII-eGFP in w1118 and C4 da neurons expressing the denoted
transgenes. GOPs are indicated by arrowheads. The scale bar represents 20 μm. (B) Quantification of the number of
GOPs in both primary and secondary dendrites of w1118 and C4 da neurons expressing the denoted transgenes. ****
p < 0.10×10-3 by one-way ANOVA with Tukey’s post-hoc correction; error bars, SEM; n ≥ 15. (C) Images that show
dendrites and dendritic spines of mouse primary neurons with or without MJD-75Q transfection at 6 DIV (turned on
at 8 DIV). The images were obtained at 12~13 DIV. The scale bar represents 50 µm. (D) Quantification of the spine
number per µm of dendrites in mouse primary neurons with or without MJD-75Q overexpression. ***p < 0.001 by
Student’s t-test; error bars, S.D.; n = 7. (E) Images of pGolt-mCherry-positive dendritic Golgi (GOPs and/or GSs) in
rat primary neurons with or without MJD-75Q overexpression using an inducible system. MJD-75Q was transfected
at 6 DIV and turned on at 8 DIV. Neurons were imaged at 14~15 DIV. Magnified images are to the right. Arrows
indicate pGolt-mCherry-positive dendritic Golgi. The scale bar represents 50 μm and the scale bar in the inset
represents 20 μm. (F) Representative CD4-tdGFP-labeled dendrite images (pseudocolored Thal) of w1118 and C4 da
neurons expressing the denoted transgenes. The scale bar represents 50 μm.
C
0 2 4 6 8 x 1070
2
4
6
8x 10
7
w1118, AUC
78Q
OE,
AU
C
PC
y=0.81x+4863.1R2 = 0.99
0 2 4 6 8 x 1050
2
4
6
8x 10
5
w1118, AUC
78Q
OE,
AU
C
SM
y=0.71x+3774.8R2 = 0.93
0
0.2
0.4
0.6
0.8
1
w1118 78Q OE
Rel
ativ
e le
vels
of P
CR
elat
ive
leve
ls o
f SM
0
0.2
0.4
0.6
0.8
1
w1118 78Q OE
*
*
D
E F 0
0.2
0.4
0.6
0.8
1
w1118 78Q OE
PCSM
00.20.40.60.8
1
w1118 78Q OE
*** **
G
Fig. S3
lipid biosynthetic process (LP)30
20
10
0
No.
of g
enes
sphingolipid steroid fatty acidglycolipid
brn
SMSr
alph
a4G
T1la
ceG
lcA
T-S
ifc eas
CG
3311
6C
G93
76In
osC
G38
12Pi
sbb
cC
G64
01PI
G-C
Cct
1C
G77
18C
G48
25C
G64
09C
G83
11C
G22
01PI
G-U
CG
7149
min
ofu
12C
G59
91C
G45
85C
G30
381
norp
AC
dsA
ecd
Npc
2bC
G77
24N
pc2a
CG
1026
8da
reC
yp6t
3C
yp4g
15sa
dbe
ta4G
alN
AcT
AC
1Gal
TAC
yt-b
5-r
HLH
106
Des
at1
CG
1998
CG
9804
fa2h
CG
2781
CG
1217
0A
CC
CG
1093
2C
G31
523
Bal
dspo
tLa
s
78Q OEw1118
lipid biosynthetic process (LP)
phospholipid
log2-fold-changes
210-1-2
A
B
Figure S3. PolyQ Toxicity Impairs the Lipid Biosynthetic Process and Decreases the Amount of SM and PC
(Related to Figure 3).
(A) The number of down-regulated genes involved in daughter GOBPs of the lipid biosynthetic process (LP). (B)
Heat map representation of down-regulated genes involved in daughter GOBPs of LP. Color bar in the heat map,
gradient of log2-fold-changes of mRNA expression levels in 78Q OE and w1118 with respect to the median mRNA
expression levels. (C-F) Lipidomics analysis of Sphingomyelin (SM) and Phosphatidylcholine (PC). Comparison of
SM (C) and PC (E) abundances (area under the elution curves; AUCs) measured by multiple reaction monitoring
methods between 78Q-expressing and control fly brains. Relative quantified levels of total SM (D) and PC (F) in
78Q OE to those in w1118 are denoted by the slope of the regression lines: 0.71 for SM and 0.81 for PC. (G)
Relative quantified amounts of SM and PC in 78Q-expressing fly brains using Lipid assay kits (Cell Biolabs, San
Diego, CA, USA), compared to those in w1118 controls. *p < 0.05, **p < 0.01, ***p < 0.001 by Student’s t-test;
error bars, SEM; n ≥ 3, controls set to 1.
020406080
100120
w1118 Sec31 RNAi
***
Ave
rage
fluo
resc
ence
in
tens
ity (a
.u.)
A w1118 Sec31 RNAi
Man
II-eG
FPM
anII-
eGFP
mem
bran
e m
arke
r
# of
GO
Ps in
prim
ary
+ se
cond
ary
dend
rites
B
02468
101214
w1118 Sec31RNAi
Sec23RNAi
*******
C
w1118
C4da neuronal dendrites
Low High
Sec31 RNAim
CD
8-G
FP in
tens
ity
D
Fig. S4
Figure S4. Disruption of the COPII Pathway Decreases the Number of GOPs and the PM Protein Supply
(Related to Figure 4).
(A) Representative images of GOPs labeled by ManII-eGFP in w1118 and C4 da neurons expressing Sec31 RNAi.
PM marker is labeled in red (ppk-CD4-tdTom). GOPs are indicated by arrowheads. The scale bar represents 10 μm.
(B) Quantification of the number of GOPs in both primary and secondary dendrites of w1118 and C4 da neurons
expressing Sec31 RNAi or Sec23 RNAi. ***p < 0.001, ****p < 0.10×10-3 by one-way ANOVA with Tukey’s post-
hoc correction; error bars, SEM; n ≥ 8. (C) Representative mCD8-GFP-labeled dendrite images (pseudocolored
Thal) of w1118 and C4 da neurons expressing Sec31 RNAi. The scale bar represents 25 μm. (D) Quantification of
mCD8-GFP pixel intensity in w1118 and C4 da neurons expressing Sec31 RNAi. ***p < 0.001 by Student’s t-test;
error bars, SEM; n = 8.
B
Fig. S5
Man
II-eG
FP
78Q OE + CrebA OE78Q OE
78Q OE+ HLH106 OE 78Q OE + Dref OE78Q OE + gt OE
AM
anII-
eGFP
0.8
0.85
0.9
0.95
1
CREB3L1 Sec13 Sec23A
control MJD-77Q
Rel
ativ
e m
RN
A le
vel
*****
**
Figure S5. PolyQ Proteins Transcriptionally Inhibit the CrebA/CREB3L1-COPII Pathway in HEK293T Cells
(Related to Figure 4).
(A) The images of GOPs in da neuronal cluster expressing the denoted transgenes. ManII-eGFP was used as GOP
marker and 10(2)80-gal4 was used to drive expression in da neurons. Magnified images of the blue box at the
bottom of each picture. Red arrows indicate somatic Golgi. The scale bar represents 50 μm and scale bar in the inset
represents 20 μm. (B) mRNA levels of CREB3L1, Sec13, and Sec23A in HEK293T cells expressing MJD-77Q or
control (transfection reagents only). **p < 0.01, ***p < 0.001 by Student’s t-test; error bars, SEM; n ≥ 3.
0 1 2 3 4 5 6 7 8 90
50
100
150
200
250
300
Fig. S6
78Q OE + CrebA OE78Q OEA
78Q OE + Rac1 OE + CrebA OE
log2 (length)
# of
term
inal
den
drite
s
B
<10um >10um
C<10 µm >10 µm
Frac
tions
of
term
inal
den
drite
s
0
0.2
0.4
0.6
0.8
78Q OE +Rac1 OE
78Q OE +Rac1 OE +CrebA OE
** **
0
0.2
0.4
0.6
78Q OE +Rac1 OE
78Q OE +Rac1 OE +CrebA OE
Figure S6. Co-overexpression of Rac1 and CrebA Synergistically Restores the Terminal Dendrite Elongation
in the PolyQ-Expressing Neurons (Related to Figure 5).
(A) Comparison of dendrite traces of representative C4 da neurons expressing the denoted transgenes: 78Q OE
(orange) vs. 78Q OE + CrebA OE (light blue). The scale bar represents 100 μm. (B) Histograms of the log2(length of
terminal dendrites) in 78Q OE + Rac1 OE + CrebA OE. Short (<10 μm) and longer (>10 μm) terminal dendrites are
partitioned by a red line. (C) Comparison of the fractions of short (left) and longer (right) terminal dendrites neurons
expressing the denoted transgenes: Rac1 OE + CrebA OE (purple) and 78Q OE + Rac1 OE + CrebA OE (red). **p <
0.01 by Student’s t-test; error bars, SEM; n = 6.
78Q OE
w1118
ANF-EMDCD4-tdTom overlay
0
5
10
15
20
w1118 Sp1RNAi
# of
GO
Ps in
prim
ary
+
seco
ndar
y de
ndrit
es N.S.C
D
A
Fig. S7
B
ddaE
ddaD
27Q OE 78Q OE
ddaFddaC
ddaA
ddaB
ddaE
ddaDddaF
ddaC
ddaA
ddaB
α-HRPα-cut
78Q OE 78Q OE + Cut OE78Q OE + Scr OE
ManII-eGFP
# of
cle
aved
cas
pase
-3
(+) c
ells
in a
dult
fly b
rain
N.S.
78Q OE
w1118
nc82 Caspase-3 overlay
E
F
0
30
60
90
120
150
w1118 78Q OE
G
0
0.2
0.4
0.6
0.8
1
1.2
w1118 78Q OE
Rel
ativ
e rR
NA
leve
l N.S.
I
H
18h
APF
27Q OEw1118
120h
AEL
78Q OE
0
20
40
60
80
100
w1118 27Q OE 78Q OE
Perc
ent o
f neu
rons
with
de
ndrit
e br
anch
at 1
8h A
PF (%
)
Figure S7. Overexpression of Cut or Scr Cannot Restore the Loss of GOPs, and Neuronal Cell Death is Not
Obvious in One-Day Old Fly Brains Expressing PolyQ Proteins (Related to Figure 6 and Discussion).
(A) Black and white images of GOPs in da neuronal cluster expressing the denoted transgenes. ManII-eGFP was
used as GOP marker (white) and 10(2)80-gal4 was used to drive expression in da neurons. The scale bar represents
50 μm. (B) Immunostaining of Cut using anti-Cut antibody in da neuronal clusters expressing 27Q (left) and 78Q
(right). Membrane was stained with anti-HRP antibody. (C) Quantification of the number of GOPs in both primary
and secondary dendrites of w1118 and Sp1 RNAi. N.S., not significant by Student’s t-test; error bars, SEM; n = 8. (D)
Images of neuropeptide vesicles (ANF-EMD) in da neuronal clusters with or without 78Q expression. CD4-tdTom
was used as a membrane marker. The scale bar represents 50 μm. (E) Images of one-day-old adult fly brains with or
without 78Q expression. The brain is marked in green by nc82, a synaptic marker. Cleaved caspase-3 were
immunostained using anti-cleaved caspase-3 antibody to examine cell death. (F) Quantification of the number of
cells positive for caspase-3 puncta in adult fly brain with or without 78Q expression. N.S., not significant by
Student’s t-test; error bars, SEM; n = 3. (G) Comparison of rRNA band intensity between w1118 fly heads and
polyQ-expressed heads. N.S., not significant, by Student’s t-test; error bars, SEM; n = 4. (H) Dendrite images of C4
da neurons expressing denoted transgenes at 120h after egg laying (AEL) and 18h after puparium formation (APF).
ppk-gal4-driven expression of CD4-tdGFP or mCD8-RFP was used to label dendrites. Red arrows indicate cell
bodies. The red scale bar represents 20 μm and blue scale bar represents 40 μm. (I) Percentage of C4 da neurons that
have dendrite branches attached to the cell body at 18h APF in each genotype.
Supplemental Experimental Materials and Procedures
Fly stocks
The third chromosome weak allele of UAS-MJDtr-78Q (78Q) was used for genetic interactions unless otherwise
specified, while the strong, second chromosome allele (UAS-MJDtr-78Qs; 78Qs) was used for RNA and protein
work. For all RNA and protein work, transgenes were driven by pan-neuronal elav-gal4 (III) driver except for the
cases where the transgenic expression induced significant early lethality (i.e. Cut and CBP overexpression). To
overexpress CBP and/or Cut pan-neuronally while bypassing early lethality, the transgenes were induced at 1 day
after eclosion using elavGS-gal4 (III) (Figures 5E and 5H). The flies were fed 100 µM Mifepristone (RU486)
(SIGMA) for 7 days for induction before being collected for subsequent RT-PCR analysis. UAS-HLH106-FLAG-
HA was obtained from the National Centre for Biological Sciences Drosophila Facility. UAS-gt-3xHA and UAS-
Dref-3xHA were obtained from FlyORF (Switzerland) (Bischof et al., 2013). UAS-Ataxin1-82Q (Fernandez-Funez
et al., 2000) was kindly provided by J. Botas (Baylor College of Medicine, Houston, TX). UAS-MJDFL-78Q
(Warrick et al., 2005) was gifted by N.M. Bonini (University of Pennsylvania, PA). The following lines were
kindly provided by Y.N. Jan (University of California, San Francisco, CA): ppk-gal4, ppk1a-gal4, 109(2)80-gal4,
ppk-CD4-tdtomato, UAS-mCD8-GFP, UAS-mCD8-RFP, UAS-Mannosidase II-eGFP (ManII-eGFP), UAS-CD4-
tdGFP, and UAS-cut.
Microscope image acquisition
Imaging of live animals was carried out by using Leica SP5, Zeiss LSM700, or LSM780 confocal microscopes.
Unless otherwise specified, 3rd instar larvae were taken directly from the vial to the slide glass for mounting. As
the mounting solution to paralyze the larvae, 1:5 ratio of diethyl ether to halocarbon oil was used. Acquired images
were processed by using Adobe Photoshop program, Image J, or Neurolucida 360 (MBF Bioscience). Images of
the C4 da neurons were obtained from the abdominal segments A2-A4.
Image processing and analyses
Adobe Photoshop was used to edit and invert images of dendrites and to merge multiple images. Image J was used
to convert the original colors to pseudocolors (Thal) for better comparison of CD4-tdGFP or mCD8-GFP intensity.
Neurolucida (MBF Bioscience) was used to quantify dendritic branch points and to create pseudocolored images
of dendrite traces. Microsoft Excel was used to create basic graphs and perform two-tailed Student’s t-test.
GraphPad Prism (Version 6.01) was used to perform one-way and two-way ANOVA with Tukey’s post-hoc
correction.
Analysis of terminal dendritic lengths
For each imaged C4 da neuron, a dendritic trace was first drawn by using Neurolucida 360 (MBF Bioscience).
Then, the automatic dendrite ordering function (Strahler’s method) of Neurolucida 360 (MBF Bioscience) was
used to identify terminal dendrites (dendrite order 0) and to measure the lengths of all terminal dendrites. The
measured terminal dendritic lengths were collected and then converted to log2-scaled lengths. The Gaussian kernel
density estimation method (Bowman and Azzalini, 1997) (MATLAB 2011a) was applied to the log2-scaled lengths.
A histogram was drawn with the bin size of 0.2 within the range between 0 and 9. Also, a normalized histogram
was generated by matching the trapezoidal integrals of the histogram and the estimated kernel density function.
To evaluate the statistical significance of the difference between the distributions (probability density
functions) of terminal dendritic lengths for two different types of C4 da neurons (n = 6 per type) expressing the
indicated transgenes, we performed the following procedure. First, as a statistic value, we computed the maximum
difference between cumulative density functions (CDFs) for the two real distributions of terminal dendritic lengths.
Second, we computed the statistic value for the two distributions of terminal dendritic lengths obtained from
randomly permuted samples. This step was repeated 1000 times, resulting in 1000 statistic values from the random
permutation experiments. Third, an empirical null distribution was estimated for the 1000 statistic value using the
Gaussian kernel density estimation method (Bowman and Azzalini, 1997). Finally, the p value for the observed
static value for the two real distributions was computed by the right-sided test using the empirical distribution.
To estimate the number of samples that ensures the statistical power in the comparison, we performed
power analysis for effect size=2.0 for the mean difference of terminal dendrite lengths in the two comparisons.
The power analysis revealed that the total number of neurons being compared should be larger than 12 (i.e., n =
6 in each of the two types of C4 da neurons being compared) to achieve statistical power 0.95. The effect size of
2.0 was estimated from the comparison of 78Q OE + Rac1 OE versus 78Q OE + Rac1 OE + CrebA OE (effect
size = 2.03), which showed a smaller mean difference than the comparison of 78Q OE versus 78Q OE + Rac1 OE
(effect size = 4.70).
Quantification of GOPs
The numbers of GOPs marked by ManII-eGFP were counted in primary and secondary dendrites of C4 da neurons
expressing specific transgenes of interest. Primary dendrites are defined as those originating from soma until the
first branch point (Ye et al., 2007). Secondary dendrites are defined as those dendrites from the first branch point
to the second branch point.
Larval locomotion
Third instar larvae were individually picked off the vial wall and placed into an agarose (2%) 90mm petri dish
with 1X PBS and left there for roughly 5 min for acclimation. Next, the larva was placed carefully onto the center
of another 2% agarose 90mm petri dish and began timing. For each genotype, we recorded the localizations of the
larvae (n = 20 per genotype) in 1 min into a 20×1 binary vector where ones and zeros indicate whether the larvae
reached the edge of the petri dish or not, respectively. We then computed the fraction of larvae that reached the
edge of the petri dish within 1 min (i.e, the number of ones divided by 20). To evaluate the statistical significance
of the difference between the fractions of two different genotypes (G1 and G2) of larvae, we performed the
following procedure. First, we combined the localization data for G1 (20×1) and G2 (20×1) into a 40×1 vector,
randomly permuted the elements in the combined vector, split the permuted vector into G1_rand (20×1) and
G2_rand (20×1), and computed the difference between the fractions from G1_rand and G2_rand. This step was
repeated 100,000 times, resulting in 100,000 fraction differences. Second, an empirical null distribution for the
fraction difference was estimated by applying the Gaussian kernel density estimation method (Bowman and
Azzalini, 1997) to the 100,000 fraction differences from the random permutation experiments. Third, the p value
for the observed fraction difference (Figure 5F) was computed by the right-sided test using the empirical
distribution. These steps were applied to all pairs of genotypes. Finally, we performed a multiple testing correction
for the p values resulted from the above procedure for all pairs of genotypes using Bonferroni post-hoc test.
Immunostaining
Third instar larval fillet was fixed with 4% formaldehyde (Junsei, Japan) for 20 min at room temperature (RT).
After rinsing with washing buffer (0.3% Triton X-100 in Phosphate Buffered Saline), samples were incubated in
the blocking buffer (Normal Donkey Serum at a concentration of 1:20 in washing buffer) for an hour at RT.
Samples were then incubated overnight at 4°C with primary antibodies of rat anti-HA (3F10, Roche Applied
Sciences; 1:100 dilution) for detection of 78Q, mouse anti-V5 (46-0705, Invitrogen; 1:100 dilution) for detection
of CBP, and rabbit anti-Cut for detection of Cut in blocking buffer. After rinsing samples several times with
washing buffer for 30 min, they were incubated further with fluorescent dye-conjugated secondary antibodies
(Jackson Immunoresearch Laboratories; 1:200 for anti-rat Cy2 & 1:50 for anti-mouse cy5) for 2~4 hours at
RT. Cy3-conjugated anti-HRP (Jackson Immunoresearch Laboratories; 1:200 dilution) staining was used together
with the aforementioned secondary antibodies to label membrane of dendrites of whole da neurons in larvae
fillet. After washing, samples were mounted with 70% glycerol in phosphate buffered saline (PBG) for imaging.
Lipid Extraction
The heads of Drosophila melanogaster were collected and used to measure sphingomyelin (SM) and
phosphatidylcholine (PC) using the lipid assay kits (STA-601 and STA-600, respectively) according to the
manuals provided by Cell Biolabs, Inc. After homogenization in chloroform:methanol solution (2:1 ratio) and
centrifugation at 15000xg for 1 min, the supernatants were incubated at RT for 1 hour on the shaker. Phase
separation with deionized water was performed, and the samples were incubated at RT for 10 min. The samples
were centrifuged at 1000xg for 10 min, and their lower organic phases were collected. The sample solutions were
completely dried out by using SpeedVac Concentrator Savant SPD1010 (Thermo Scientific). The pellet obtained
from each sample was dissolved in chloroform:methanol:water solution (60:30:45 ratio), and before performing
each lipid assay kit, the sample was diluted from 1:50 to 1:400 accordingly.
Analysis of the PM lipids (Lipidomics)
Sample Preparation- For the lipid extraction from fly heads, a two-staged extraction procedure, including neutral
and acidic extractions, was used. First, for the neutral extraction, the heads were added to 1 mL of
chloroform/methanol (1:2, v/v). The sample was vortexed for 30 s every 3 min. After the centrifugation (13,800xg,
2 min at 4 °C), the 950 µL of supernatant was transferred to a new Eppendorf tube. Second, for the acidic extraction,
the remaining pellet was resuspended in 750 µL chloroform/methanol/37% (1N) HCl (40:80:1, v/v/v) and
incubated for 15 min at RT while vortexing the sample for 30 s every 5 min. After transferring the tube to ice, 250
µL cold chloroform and 450 µL cold 0.1 M HCl were added. The sample tube was vortexed briefly and then
centrifuged (6,500 ×g, 2 min at 4 °C). The organic phase was collected and dried with the SpeedVacTM and re-
suspend with 500 µL of 0.1% formic acid in H2O. The injection volume was 2 µL for LC-MS/MS.
LC-MS/MS analysis- The quantitative lipid profiling was performed by 6490 Accurate-Mass Triple Quadrupole
(QqQ) LC-MS coupled to a 1200 series HPLC system (Agilent Technologies, Wilmington, DE, USA) with a
Hypersil GOLD column (2.1 × 100 mm ID; 1.9 μm, Thermo science). This provides high sensitivity by iFunnel
technology that consists of three components: Agilent Jet Stream technology, a hexabore capillary, and a dual ion
funnel. The used column was a Hypersil GOLD column (2.1 × 100 mm ID; 1.9 μm, Thermo science). The
temperatures of column oven and sample tray were set to 40 and 4 °C, respectively. Solvent A consisted of a
acetonitrile–methanol–water mixture (19:19:2) with 20 mmol/L ammonium formate and 0.1% (v/v) formic acid,
and solvent B consisted of 2-propanol with 20 mmol/L ammonium formate and 0.1% (v/v) formic acid. The flow
rate was set to 250 μL/min, and the injection volume was to 2 μL. The following elution gradient program was
used: holding solvent steady (A/B: 95/5) for 5 min, a first linear gradient to solvent (A/B: 70/30) for 10 min, a
second linear gradient to solvent (A/B: 10/90) for 7 min, an isocratic elution at solvent (A/B: 10/90) for 3 min,
and a third linear gradient to solvent (A/B: 95/5) for 1 min. The column was equilibrated at 5% solvent B for 4
min before reuse. Total run time was 30 min for each analysis. The following parameters were used for data
acquisition: 3500 V positive mode of capillary voltage, 3000 V negative mode of capillary voltage, a sheath gas
flow of 11 L/min (UHP nitrogen) at 200ºC, a drying gas flow of 15 L/min at 150ºC, and nebulizer gas flow at 25
psi. Multiple reaction monitoring (MRM) conditions, including transition and MS/MS collision energy, to analyze
each target lipid were optimized using several lipid standards.
Cell Culture and Transfection
Immortalized cell line- HEK293T cells were generally grown in DMEM/High Glucose media (Thermo Scientific
HyClone) with 10% FBS (Thermo Scientific HyClone). Cells were counted and seeded one day before the
transfection. When the seeded cells became 70-80% confluent, the media was changed to FBS-free DMEM media.
Based on the protocol provided by Invitrogen, Lipofectamine 2000 (Invitrogen, USA) was used to transfect 5μg
of DNAs in Opti-MEM media. After the addition of DNA-lipid complex to cells and 4-hour incubation thereafter,
their media was changed back to the original 10% FBS-containing DMEM media and then incubated for 1-2 days.
Primary neuron culture and inducible expression of polyQ proteins- For dendrite imaging analysis, embryonic
hippocampal neurons were isolated from mouse embryos (E16~E17) as previously described (Fu et al., 2007) and
plated at the initial density of 1.5 × 105 cells/well. A modified bidirectional inducible pBI vector system
(Clontech), a generous gift from E. H. Jho (The University of Seoul, Korea), was used to optimize the expression
of toxic polyQ proteins in cultured neurons. 1ug of DNA per well were transfected by using Lipofectamine 2000
(Invitrogen) at 6 DIV based on the protocol provided by Invitrogen. For the induction of expression, 15ng/mL of
Doxycycline was treated at 8 DIV. Cultured cells were immunostained and imaged at 12~13 DIV.
Similarly, for Golgi Outpost (GOP) imaging analysis, embryonic hippocampal neurons were dissected from rat
embryos (E16~E17). The initial density of cells was 7 × 104 ~ 1 × 105/well. The same modified bidirectional
inducible pBI vector system (Clontech) used in dendrite imaging analysis was also used to optimize the expression
of toxic polyQ proteins in cultured neurons. For labeling of GOPs, pGolt-mCherry was used. pGolt-mCherry /
pGolt3-mCherry was a gift from Michael Kreutz (Addgene plasmid # 73297) (Mikhaylova et al., 2016). 1.2ug of
each DNA per well were transfected by using CalPhosTM mammalian transfection kit (Clontech) at 6 DIV based
on the user manual provided by Clontech Laboratories, Inc. For the induction of expression, 15ng/mL of
Doxycycline was treated at 8 DIV. Cultured cells were immunostained and imaged at 14~15 DIV.
Chromatin immunoprecipitation (ChIP)
The ChIP assay was performed as previously described with slight modifications (Kim et al., 2016). After fly
heads were collected in ice cold PBS, the PBS was totally eliminated, and the fly heads were lysed within 1.5%
formaldehyde [dissolved in 0.143M NaCl, 1.43mM EDTA, 71.45mM HEPES-KOH (pH 7.5), and 1X protease
inhibitor (#87786, Thermo Scientific)] to crosslink CBP with unknown chromatin regions. After 20 min
incubation at RT, glycine and Tris were added for the final concentration to be 370mM and 2.5mM, respectively.
After crosslinking termination for 5 min at RT, lysates from wild-type or CBP OE were centrifugated at 4,000 XG
for 5 min. Next, the isolated pellets were washed with Tris-buffered saline (20mM Tris-HCL (pH 7.5) and 150mM
NaCl), FA lysis buffer [50mM HEPES-KOH (pH7.5), 150mM NaCl, 1mM EDTA, 1% Triton X-100, 0.1% SDS
and 0.1% Na-deoxycholate], FA lysis buffer/0.5% SDS [50mM HEPES-KOH (pH7.5), 150mM NaCl, 1mM
EDTA, 1% Triton X-100, 0.5% SDS and 0.1% Na-deoxycholate], and FA lysis buffer sequentially. Then the final
pellets were re-suspended within FA lysis buffer.
The suspended sample was sonicated under ‘Peak power’ 70.0, ‘Duty factor’ 25.0, and ‘Cycles/Burst’
200 (Average power 17.5) condition using Covaris M220 for 30 min, and lysates containing the sheared
nucleosomes were isolated. Input control was prepared from the lysates, and the remaining lysates containing
nucleosomes were incubated with 1μl of anti-V5 antibody (ab9116, Abcam) and protein A/G agarose (SC-2003,
Santa Cruz) at 4℃ overnight. Next, incubated agarose beads were precipitated and washed with FA lysis buffer
and washing buffer [10mM Tris-HCl (pH 8.0), 0.25M LiCl, 1mM EDTA, 0.5% NP-40, 0.5% Na-deoxycholate]
in turn. After washing, nucleosomes bound to the agarose beads were collected in elution buffer [50mM Tris-HCl
(pH7.5), 10mM EDTA, and 1% SDS] at 65℃ for 20 min. Then, the eluted nucleosome lysates and previous input
control were incubated with RNase (1mg/ml) at 37℃ for 1 hour followed by incubation of proteinase K (20mg/ml)
at 42℃ for 1 hour. Next, the crosslinks were reversed by heating at 65℃ for 5 hours. The DNA was extracted
with phenol-chloroform-isoamyl alcohol method and final DNA products were dissolved in ultrapure water.
Quantitative PCR (qPCR)
Amount of DNA region of interest was analyzed by qPCR method using QuantiSpeed SYBR Green kit (QS105-
10, PhilKorea) and CFX96 TouchTM Real-Time PCR Detection System (Bio-Rad), and all qPCR experiments
were performed in triplicates to ensure fidelity.
Analysis for qPCR
To calculate the relative enrichment of CBP on the promoter region of CrebA compared to the CrebA ORF region,
following equations were used: ΔΔCq = ΔCq (Binding candidateIP - ORFIP) - ΔCq (Binding candidateInput - ORFInput)
and Relative fold change = 2(−𝛥𝛥𝐶𝑞).
Primers used in qPCR
Four putative CRE sites were identified using MotifLocator in the promoter region of CrebA gene. Thus, we
designed four sets of primers starting at 1.6kb upstream of the transcription start site (TSS) and ending at 5’UTR.
Primers at the ORF region of CrebA was also designed as negative control.
Gene Sequences
CrebA P1 5’ - GCCAATTTACTATGTGGTGTAAGC
3’ - ACAGATACAGCGCAGACAAAC
CrebA P2 5’ - GACAGCAGCAGCAGCATAAG
3’ - TGGTAACACTCGCTCCGTTG
CrebA P3 5’ - CCATAGCGTCGGAAGAAACTTG
3’ - GCCACCAATCTACCCTTGATACTG
CrebA P4 5’ - CAGAGGACTGACGTTTCGATTC
3’ - TGTGTGTGAATGGGCAGTAAAC
CrebA ORF 5’ - CCGGAATCCTTGAAAATCAGCC
3’ - GTATGACGGTGGGATCTTTGAC
RT-PCR
Total RNAs were extracted from the heads of Drosophila melanogaster or HEK293T cells with Easy Blue system
(iNtRON Biotechnology, Korea). Transgenic expression in fly heads were driven by pan-neuronal elav-gal4 driver
and the samples collected at 1 day after eclosion. The extracted RNAs were converted into cDNAs by using
GoScriptTM Reverse Transcription System (Promega). For RT-PCR, cycle numbers of 28 (for Sec13, Sec23,
Sec31, Sec63, and Rac1), 27 (for CrebA), 26 (for CREB3L1), 22 (for human Sec13), and 24 (for Sec23A) were
performed with C1000 TouchTM Thermal Cycler or T100TM Thermal Cycler system (Bio-Rad).
Primer Information
Primers for the fly genome
Gene Sequences
Sec13 5’– ACTTCTACGGCCTACTGTTG
3’–TCGTGGTTGCTGTATTCGTA
5’– TACGAATACAGCAACCACGA
3’– TAAATTATCGCAACCACCGC
Sec23
5’– TCCGTTTCCCCCTCAATATG
3’– GGTTATCAGACCCACGAGAG
5’– AGAGCTATGTGTTTCGTGGT
3’– CACGAAGGTCATAATGCGAC
Sec31 5’– GAACCTCGAAAACGGTCAAG
3’– TGCCATCCGTTTCCTTGATA
5’– ATCCCTTCCAGAACAACCTG
3’– GACTCTTACGCAGATCCCAA
Sec63 5’– TCGACGATGAGAACACCAAT
3’– GTTCTTTTGGTTCTGGGCTG
5’– CATGAAAATGTCGCCGATGA
3’– ACCGAATAGCGTCTTCATGT
CrebA 5’– CAACTACCTCAGCACCTATACGAC
3’– GTTACCTTCGGAATCATCGCTGG
Rac1 5’– CAGCTACACGACCAATGCCTTTCC
3’– CCGAGCACTCCAGATACTTGACC
Primers for the human genome
Gene Sequences
CREB3L1
5’– GATCTGAACGAGTCGGACTTCC
3’ – CATGCTCCATGCTCTGCATC
Sec13
5’ – GTTGTACCTGGAAGCCTCATAG
3’ – CACCTTATTGTCTCCACCAGAG
Sec23A
5’ – CTCCTGGAGATGAAATGCTGTC
3’ – GCTAAGGTTGTAGTGGGACTAAG
Co-Immunoprecipitation (coIP)
Equal number of fly heads from each line was collected inside the lysis buffer (150mM NaCl, 1% Triton X100,
50mM Tris-HCl pH7.5) with 1:100 ratio of protease inhibitor cocktail (stock concentration of 100x) (Thermo
Scientific, USA). Samples were completely homogenized and centrifuged at 4ºC with the maximum speed for 10
min. Protein amount was measured via Bradford protein assay (Bio-Rad). After matching the total protein amount
of lysate, α-V5 antibody-ChIP grade (ab9116, Abcam) was added in each sample with 1:1000 ratio and incubated
on the shaker at 4ºC overnight. Pre-cleaned protein A/G plus-agarose beads (Santa Cruz Biotechnology, Inc.) were
added into the samples and incubated on the shaker at 4ºC for 3 hours. The samples were centrifuged at 1000xg
for 3 seconds at 4ºC, and the supernatant was discarded. After washing the beads with 500ul lysis buffer 3 times,
4x laemmli buffer (Bio-Rad; diluted to 2x) was added with 2-Mercaptoethanol (BIOSESANG, Inc.) for 9:1 ratio
and boiled at 95ºC for 5 min. The sample was then centrifuged at 1000xg for 3 seconds, and the supernatant was
taken for immunoblotting. The samples were loaded onto the gel and detected using α-Cut (DSHB: 2B10) antibody.
mRNA-sequencing and data analysis
We expressed 78Q using a pan-neuronal elav-gal4 driver and isolated the adult fly heads for NGS analysis. First,
total RNAs were prepared from the fly heads with or without expressing 78Q. Poly(A) mRNA isolation from total
RNA (2 µg) and fragmentation were performed using the Illumina Truseq Stranded mRNA LT Sample Prep Kit
with poly-T oligo-attached magnetic beads, according to the manufacturer’s instructions. Reverse transcription of
RNA fragments was performed using Superscript II reverse transcriptase (Life Technologies). The adaptor ligated
libraries were sequenced using an Illumina HiSeq 2500 (DNA Link, Korea). The mRNA-sequencing analysis was
performed for two independent replicates in each condition. From the resulting read sequences for each sample,
adapter sequences (TruSeq universal and indexed adapters) were removed using the cutadapt software (Martin,
2011). Remaining reads were then aligned to the Drosophila melanogaster reference genome (Flybase r5.57)
using TopHat aligner (Trapnell et al., 2009). Considering the variations in individual genomes and presence of
multiple gene copies, we used two mismatches in a read and allowed the reads to be aligned in up to 20 different
locations, which are default options in the TopHat aligner. After the alignment, we counted the numbers of reads
mapped to gene features (GTF file of BDGP5.73) using HTSeq (Anders et al., 2014) and also estimated fragments
per kilobase of transcript per million fragments mapped (FPKM) using Cufflinks (Trapnell et al., 2010).
Identification of differentially expressed genes (DEGs)
On average, 21.8 million reads were obtained in individual samples and aligned to the fly genome, resulting in
2.2 Giga bps of mapped sequences, which correspond to 74.1-fold coverage of the annotated fly transcriptome.
We first identified ‘expressed’ genes as the genes with fragments per kilobase of transcript per million fragments
mapped (FPKM) larger than 1 under at least one of the four samples (two independent samples for w1118 or
MJDtr-78Q). For these expressed genes, the numbers of reads counted by HTseq were converted to log2-read-
counts after adding one to the read counts. The log2-read-counts for the samples in each condition were then
normalized using the quantile normalization method (Bolstad et al., 2003). For each gene, we calculated a T-
statistic value using Student’s t-test and also a log2-fold-change in the comparison of 78Q versus w1118. We then
estimated empirical null distributions for T-statistic value and log2-fold-change by random permutation of the four
samples (300 permutations). Using the estimated empirical distributions, we computed adjusted p values for the
two tests for each gene and then combined these p values with Stouffer’s method (Hwang et al., 2005). Finally,
we identified DEGs as the ones that have the combined p values < 0.05 and absolute log2-fold-changes > 1 (2-
fold). To identify cellular processes represented by the DEGs, we performed the enrichment analysis of gene
ontology biological processes (GOBPs) for the down-regulated genes using DAVID software (Huang et al., 2009)
and selected the GOBPs with p value < 0.05 as the processes enriched by the down-regulated genes.
Identification of key transcription factors (TFs)
We first selected the DEGs involved in membrane organization (MO), vesicle-mediated transport (VT), and lipid
biosynthetic process (LP) associated with GOP synthesis. Using the TF-target relationship data for 150 TFs from
DroID (v2014_01) (Yu et al., 2008) and previous literatures (Abrams and Andrew, 2005; Kunte et al., 2006), for
each TF, we counted the number of target genes in the selected DEGs and then calculated its fraction of the total
number of target genes. We also calculated the significance of the number of the target genes based on the
following random sampling experiments: 1) the same number of the genes with the selected DEGs were randomly
selected from the whole genes of Drosophila melanogaster (NCBI RefSeq release 68); 2) the number of target
genes for the TF was counted; 3) Steps 1-2 were repeated 100,000 times; 4) the empirical distribution for the
number of target genes was estimated using the Gaussian kernel density estimation method (Bowman and
Azzalini, 1997); and 5) p value for the observed number of the target genes in the selected DEGs was estimated
using the right-sided test using the empirical null distribution. Finally, we selected the 29 differentially expressed
TFs with p < 0.05, fraction > 0.01 for at least one of the processes (MO, VT, and LP), and the number of target
genes larger than 3.
Supplementary References
Abrams, E. W. & Andrew, D. J. (2005) CrebA regulates secretory activity in the Drosophila salivary gland and
epidermis. Development, 132(12), 2743-58.
Anders, S., Pyl, P. T. & Huber, W. (2014) HTSeq - A Python framework to work with high-throughput
sequencing data.
Bischof, J., Bjorklund, M., Furger, E., Schertel, C., Taipale, J. & Basler, K. (2013) A versatile platform for
creating a comprehensive UAS-ORFeome library in Drosophila. Development, 140(11), 2434-42.
Bolstad, B. M., Irizarry, R. A., Å strand, M. & Speed, T. P. (2003) A comparison of normalization methods for
high density oligonucleotide array data based on variance and bias. Bioinformatics, 19(2), 185-193.
Bowman, A. W. & Azzalini, A. (1997) Applied smoothing techniques for data analysis : the kernel approach
with S-Plus illustrations. OxfordNew York: Clarendon Press ;Oxford University Press.
Fernandez-Funez, P., Nino-Rosales, M. L., de Gouyon, B., She, W. C., Luchak, J. M., Martinez, P., Turiegano,
E., Benito, J., Capovilla, M., Skinner, P. J., McCall, A., Canal, I., Orr, H. T., Zoghbi, H. Y. & Botas, J. (2000)
Identification of genes that modify ataxin-1-induced neurodegeneration. Nature, 408(6808), 101-6.
Fu, W. Y., Chen, Y., Sahin, M., Zhao, X. S., Shi, L., Bikoff, J. B., Lai, K. O., Yung, W. H., Fu, A. K. Y.,
Greenberg, M. E. & Ip, N. Y. (2007) Cdk5 regulates EphA4-mediated dendritic spine retraction through an
ephexin1-dependent mechanism. Nature Neuroscience, 10(1), 67-76.
Huang, D. W., Sherman, B. T. & Lempicki, R. A. (2009) Systematic and integrative analysis of large gene lists
using DAVID bioinformatics resources. Nature Protocols, 4(1), 44-57.
Hwang, D., Rust, A. G., Ramsey, S., Smith, J. J., Leslie, D. M., Weston, A. D., de Atauri, P., Aitchison, J. D.,
Hood, L., Siegel, A. F. & Bolouri, H. (2005) A data integration methodology for systems biology. Proc Natl
Acad Sci U S A, 102(48), 17296-301.
Kim, J. H., Lee, B. B., Oh, Y. M., Zhu, C., Steinmetz, L. M., Lee, Y., Kim, W. K., Lee, S. B., Buratowski, S. &
Kim, T. (2016) Modulation of mRNA and lncRNA expression dynamics by the Set2-Rpd3S pathway. Nat
Commun, 7, 13534.
Kunte, A. S., Matthews, K. A. & Rawson, R. B. (2006) Fatty acid auxotrophy in Drosophila larvae lacking
SREBP. Cell Metab, 3(6), 439-48.
Martin, M. (2011) Cutadapt removes adapter sequences from high-throughput sequencing reads. 2011, 17(1).
Mikhaylova, M., Bera, S., Kobler, O., Frischknecht, R. & Kreutz, M. R. (2016) A Dendritic Golgi Satellite
between ERGIC and Retromer. Cell Rep, 14(2), 189-99.
Trapnell, C., Pachter, L. & Salzberg, S. L. (2009) TopHat: discovering splice junctions with RNA-Seq.
Bioinformatics, 25(9), 1105-11.
Trapnell, C., Williams, B. A., Pertea, G., Mortazavi, A., Kwan, G., van Baren, M. J., Salzberg, S. L., Wold, B. J.
& Pachter, L. (2010) Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and
isoform switching during cell differentiation. Nat Biotechnol, 28(5), 511-5.
Warrick, J. M., Morabito, L. M., Bilen, J., Gordesky-Gold, B., Faust, L. Z., Paulson, H. L. & Bonini, N. M.
(2005) Ataxin-3 suppresses polyglutamine neurodegeneration in Drosophila by a ubiquitin-associated
mechanism. Mol Cell, 18(1), 37-48.
Ye, B., Zhang, Y., Song, W., Younger, S. H., Jan, L. Y. & Jan, Y. N. (2007) Growing dendrites and axons differ in
their reliance on the secretory pathway. Cell, 130(4), 717-729.
Yu, J., Pacifico, S., Liu, G. & Finley, R. L., Jr. (2008) DroID: the Drosophila Interactions Database, a
comprehensive resource for annotated gene and protein interactions. BMC Genomics, 9, 461.