Proteomics-based monitoring of pathway activity reveals that blocking diacylglycerol biosynthesis rescues from alpha-synuclein toxicity

Cell Systems, Volume9

Supplemental Information

Proteomics-Based Monitoring of Pathway Activity

Reveals that Blocking Diacylglycerol Biosynthesis

Rescues from Alpha-Synuclein Toxicity

Martin Soste, Konstantina Charmpi, Fabienne Lampert, Juan Atilio Gerez, Marc van

Oostrum,

Liliana

Malinovska,

Paul

Jonathan

Boersema,

Natalia

Cecilia

A

B

C

E

F

G

log2 fold change (α-Syn/vector)

-log10(q-value)

-log10(q-value)

500 most abundant proteins

vector3 vector1 vector2 α-Syn2 α-Syn1 α-Syn3

D

500 most abundant phosphopeptides

vector2 vector3 vector1 α-Syn1 α-Syn2 α-Syn3

log2 fold change (α-Syn/vector)

H

I

J

Figure S1. Phenotypic characterization, proteomics and genetic modulation of α-Syn-expressing yeast cells, related to Figure 1, Table S5, STAR methods. The previously described yeast model for α-Syn toxicity (Outeiro and Lindquist, 2003) was assessed (A) for the formation of inclusions by fluorescence microscopy of YFP-tagged α-Syn (scale bars represent 10 µm), (B) for mean growth inhibition (n=3), and (C) for mean cell viability (n=2) relative to vector control. (D-E) Hierarchically clustered heatmaps of label-free proteomic quantification displaying the 500 most intense (D) proteins and (E) phosphopeptides (n=3) (F-G) Volcano plots showing significantly regulated (F) proteins and (G) phosphopeptides in the α-Syn strain relative to vector control. C.V., coefficient of variation. (H) Rescue due to overexpression of each modulator calculated relative to growth of the α-Syn-expressing strain without modulation. According to previous studies (see also Table S5), enhancers are shown with solid bars and suppressors with checkered bars. All genes were previously classified as overexpression modulators, except VPS35 and COX5a, which were found to be deletion modulators. Log2 Rescue displayed as mean for modulators analyzed in batches 1-3 (n=3) and as single values for batch 4 (n=1) (I) Data from Figure 1B is displayed with clustering by modulator. Modulators are annotated according to their cellular process. (J) Mean pairwise corre-lations (n=6) of sentinel abundances in unmodulated α-Syn-expressing samples from four different proteomic analysis batches. All error bars repre-sent standard deviation.

Pah1 S744 & S748

D

A

B

C

Car1 Cwp1 Dot5 Ero1

Gpp1 Met10 Ncp1

She3 S361

Osh1 T1053 & S1056 Num1 S2494

Svl3 S666 Pho84 T316

Bcy1 T129

E

Figure S2. Regression analyses and relationship between Pah1 phosphorylation and modulator distance to the closest Pah1 regulator, related to Figure 1, Figure 2, Table S12 and STAR methods. Additional examples of correlation plots show the extent of rescue plotted as a function of peptide abundance relative to the

vector strain. (A) Bcy1 T129 is a sentinel for Mck1 kinase activity, PKA and TOR inhibition, and Bcy1 activation (rank 86/103 phosphopeptides). (B) Svl3 S666 is an

α-Syn deregulated phosphopeptide (rank 95/103 phosphopeptides). (C) Pho84 T316 is an α-Syn deregulated phosphopeptide (rank 11/103 phosphopeptides). (D)

Partial regression is shown for each (phospho)protein sentinel with a maximal selection probability ≥ 0.6. Each panel shows the association of the variables, rescue and abundance, after removing the linear effects of the other sentinels that also passed the stability selection thresholds for both variables. Phosphosites and proteins were treated separately. The best fit line has been added to all plots. r, Pearson correlation coefficient. p, p-value from application of the correlation test (n=37). Note that the residuals are not independent and therefore the degrees of freedom need to be adapted for computing the p-values. (E) The log2 fold-change (FC) values of Pah1 S744

and S748 plotted as a function of the shortest path lengths found for the corresponding genetic modulator. If no shortest path was found, the path length was labeled as infinite (Inf). A one-tailed (two-sample) t-test was applied testing if the average FC was greater in observations at distance 1 or 2 (n1=14) compared to observations at distance 3, 4, or Inf (n2=19), *, p-value < 0.05.

C

F

D

E

G

YFP Nile Red Merge

α-Syn α-Syn Δnem1 α-Syn pah1-7A

α-Syn α-Syn Δpah1 α-Syn pah1-D398E

α-Syn ******* ** α-Syn Δnem1 α-Syn pah1-7A

H

*

**

A

I

J

****

B

Figure S3. Regulation of α-Syn phenotypes by Pah1 activity, related to Figure 3, Figure 4 & STAR methods. (A) Yeast strains

spotted onto galactose-containing medium to induce expression of GFP or other aggregation-prone proteins (n=3) (B) α-Syn

abun-dance, assessed by western blot (n=4 for α-Syn and Δnem1, n=1 for α-Syn pah1-WT and α-Syn pah1-7A) (C) Representative images

of α-Syn-YFP cells stained with Nile red. (D) Co-localization of α-Syn and cellular lipids was assessed by correlation of YFP and Nile

red. Mean Pearson R values were compared (one-way ANOVA adjusted for multiple comparisons, n≥2, alpha=0.05). (E)

Representa-tive fields of view used to quantify the percentage of indicated cells with α-Syn inclusions. (F) Representative fields of view of

α-Syn-YFP in Pah1 mutants. (G) Cells with and without α-Syn-YFP inclusions were counted in three fields of view. Mean counts

expressed as percentages with standard deviation between biological replicates for solid data points and between fields of view for hollow data points were compared (n>3 marked with solid data points and n=1 marked with hollow data points; one-way ANOVA

adjusted for multiple comparisons, alpha=0.05). (H) Mean Integrated Nile Red intensity was used to quantify lipid abundance (one-way

ANOVA adjusted for multiple comparisons, n≥2, alpha=0.05). (I) Mean normalized biomasses of pah1-7D/E mutant or pah1-WT yeast

strains grown in 1 nM estradiol to induce α-Syn expression were monitored over 24 h (one-way ANOVA adjusted for multiple

compari-sons, n=4, alpha=0.05) (J) Observed and expected biomass of strains expressing α-Syn and a Pah1 mutant. Biomass data from (I)

were used and means were compared (two-way ANOVA adjusted for multiple comparisons, alpha=0.05). ****, p-value < 0.0001; ***, p-value < 0.001; **, p-value < 0.01; *, p-value < 0.05. All error bars represent standard deviation. All scale bars represent 10 µm.

LPIN1 LPIN2 LPIN3

HSC70 HSC70 HSC70

LPIN1

siRNA: Control LPIN2 Control LPIN3 Control

IES1 IES1 MED6 MED6 SSK1SSK1 SSK2 SSK2 MED1 MED1 RFC1RFC1 STH1 STH1 RSC4 RSC4 RTG3RTG3 MSC3MSC3 BDF1BDF1 MKS1 MKS1 SEG1SEG1 OSH2 OSH2 PHO4

PHO4 KAR3KAR3

INP53 INP53 PHO84

PHO84 _TCO89TCO89

GCN2 GCN2

ISW2 ISW2 ORM2

ORM2 GGA1GGA1 SCS2SCS2 NTC20NTC20 RGA2RGA2 DLS1DLS1

GCS1 GCS1 CWC2 CWC2 INP52INP52 MSS4MSS4 KOG1 KOG1 GUS1 GUS1 ESC1 ESC1 BUL1

BUL1 ORC2ORC2

TAF2 TAF2 TAF12 TAF12 _MCM3MCM3 NUP157 NUP157 EPO1 EPO1 ENA1 ENA1 _SLM1SLM1 NHA1 NHA1 MOT2

MOT2 _NUP84NUP84

SEC7

SEC7 _MES1MES1 AVO2AVO2

SIR3 SIR3 SIR4 SIR4 LDB19 LDB19 NUP145NUP145 CSL4 CSL4 CDC39 CDC39 NOP6NOP6 TSC11 TSC11 NUP1 NUP1 NOP19 NOP19 RRP6 RRP6 _NUP100NUP100 SPO12 SPO12 NET1 NET1 _CDC14CDC14 ECM21 ECM21 XRN1 XRN1 REG1REG1 TIF3

TIF3 PMA1PMA1

SLI15 SLI15 NUR1 NUR1 RPS7A RPS7A PAT1 PAT1 CDC55CDC55 EDC3

EDC3 RPS16ARPS16A _RTS1RTS1

PKH3 PKH3 TIF4632

TIF4632 _PDI1PDI1

KAR2 KAR2 SFB3 SFB3 FPK1 FPK1 FUN12

FUN12 _{TIF4631TIF4631} YPK2YPK2

MLC1 MLC1 RPL8B

RPL8B

PKH2

PKH2 CYR1CYR1 RIM15RIM15 CCH1CCH1

SEC16 SEC16 BUD3 BUD3 SNF3 SNF3 NOP12 NOP12 AXL2 AXL2 STM1 STM1 _SPA2SPA2 ATG1

ATG1 BUD6BUD6

UME6 UME6 BCK1 BCK1 NOP2 NOP2 HOS4 HOS4 ATG13

ATG13 TOD6TOD6

MIG1

MIG1 _ATG19ATG19 LAS1

LAS1 BNI1BNI1

RPA190 RPA190

CNM67

CNM67 SNT1SNT1 RCO1RCO1 _STB3STB3

PAH1

PAH1 _{PUP2PUP2 STE20STE20}

CYC8 CYC8 GIC2 GIC2 STB1 STB1 SPC110 SPC110 RPC82 RPC82 NPR3 NPR3 BLM10 BLM10 SYP1 SYP1 URA2 URA2 _CDS1CDS1 GPT2 GPT2 NPR2 NPR2 _WHI5WHI5 SKP1 SKP1 _MTC5MTC5 SPC29 SPC29 HXK1 HXK1 TGL4TGL4 RTC1 RTC1 EDE1EDE1 BBC1BBC1 NTH1 NTH1 YLR352W

YLR352W _{YAP1802YAP1802} STF2 STF2 UBR1 UBR1 VRP1VRP1 BBP1 BBP1 TSL1 TSL1 _TGL1TGL1 SLA2 SLA2 HSP26 HSP26 LAS17LAS17 MSC1 MSC1 _PAN1PAN1 MPS2 MPS2 SAC6 SAC6 SLA1 SLA1 RTN2 RTN2 _ARK1ARK1 HBT1 HBT1 ABP1 ABP1 CRN1 CRN1 PRK1 PRK1

A

B

Figure S4. Effects of propranolol on the α-Syn yeast phosphoproteome and siRNA-mediated inhibition of lipins in HEK293 cells perturbed by α-Syn fibrils, related to Figure 4, STAR methods. (A) STRING network of regulated phosphoproteins upon

propranolol treatment of estradiol-induced α-Syn yeast. A confidence threshold of 0.9 interaction score was applied. Disconnected nodes were omitted. K-means clustering was set with K=10 and is displayed with different colors. Intra-cluster edges are represented

by solid lines and inter-cluster edges are represented by dashed lines. (B) siRNA efficiency was assessed by RT-PCR. Cells were

transfected with the indicated siRNAs and lipin expression (LPIN1-3) was quantified by non-saturating RT-PCR. An unrelated mRNA (HSC70) was quantified as a loading control.