Parkinson’s VPS35[D620N] mutation induces LRRK2-mediated lysosomal association of RILPL1 and TMEM55B

We demonstrate that the Parkinson’s VPS35[D620N] mutation alters the expression of ~220 lysosomal proteins and stimulates recruitment and phosphorylation of Rab proteins at the lysosome. This recruits the phospho-Rab effector protein RILPL1 to the lysosome where it binds to the lysosomal integral membrane protein TMEM55B. We identify highly conserved regions of RILPL1 and TMEM55B that interact and design mutations that block binding. In mouse fibroblasts, brain, and lung, we demonstrate that the VPS35[D620N] mutation reduces RILPL1 levels, in a manner reversed by LRRK2 inhibition and proteasome inhibitors. Knockout of RILPL1 enhances phosphorylation of Rab substrates, and knockout of TMEM55B increases RILPL1 levels. The lysosomotropic agent LLOMe also induced LRRK2 kinase–mediated association of RILPL1 to the lysosome, but to a lower extent than the D620N mutation. Our study uncovers a pathway through which dysfunctional lysosomes resulting from the VPS35[D620N] mutation recruit and activate LRRK2 on the lysosomal surface, driving assembly of the RILPL1-TMEM55B complex.

).The blue dots represent lysosomal proteins annotated in the GO terms database (GO:0005764), while the red dots represent significantly enriched proteins with fold change > 1.     S4).The blue dots represent lysosomal proteins annotated in the GO terms database (GO:0005764), while the red dots represent significantly enriched proteins with fold change > 1.         Legends for Tables S4 to S6 that are Excel sheets submitted separately as auxiliary supplementary files Table S4.Search result of DIA MS data: VPS35 WT vs D620N WCL and LysoTag-IP Table S5.Search result of DIA MS data: VPS35 D620N ± MLi-2 WCL and LysoTag-IP Table S6.Search result of TMT MS data: RILPL1 WT vs R293A GFP-IP

Fig. S1 .
Fig. S1.Further Mass Spectrometry analysis of wild type and D620N MEFs.(A) Littermate-matched wild type (WT) and VPS35[D620N] homozygous knock-in (KI) MEFs were treated ± 100 nM MLi-2 for 2 h prior to lysis.Lysate were subjected to quantitative immunoblot analysis using the LI-COR Odyssey CLx Western Blot imaging and the indicated antibodies.Technical replicates represent cell extract obtained from different dishes of cells.Quantitation of immunoblotting data (performed using ImageStudioLite software version 5.2.5, RRID:SCR_013715) is shown as mean ± SEM.Data were analyzed using two-tailed unpaired student's t-test ( *** p< 0.001, **** p< 0.001).( B) The volcano plot show the result of MEFs VPS35[D620N] not 5 and p-value < 0.05.The black dotted line box indicates proteins confidently detected in the LysoTag-IP samples but not in the Mock-IP samples.(C) The log2 MS intensities of these proteins are shown.(D) The volcano plot (Curtain link: https://curtain.proteo.info/#/18c40d8a-f678-4d7e-976a-f70c85683ccc) shows the proteome changes of all detected proteins in MEFs VPS35-WT versus VPS35[D620N] LysoTag-IP experiments.The red dots represent mock-enriched proteins with fold change > 1.5 and p-value < 0.05, as determined in the experiments shown in Supplementary Figure 1A.The gray dots represent proteins bound to magnetic beads, which do not show significant enrichment.(D) The raw MS intensities of two non-specific proteins that associate with the magnetic beads (TINAGL1 and H2BC21) were shown for reference.

Fig. S3 .
Fig.S3.Gene Ontology analysis of proteins whose levels are most affected by the D620N mutations.The GO term pathway analysis of significant down-regulated (green bar) and up-regulated proteins (red bar) observed in (A) WCL (Fig.1C) and (B) LysoTag-IP (Fig.1D) experiments with fold change > 1.5 and p-value <0.05 using Metascape software (RRID:SCR_016620, version 5.3) with an enrichment of p-value < 0.01.

Fig. S4 .
Fig. S4.Further Mass Spectrometry analysis of D60N MEFs treated ± MLi-2.(A) The volcano plots show the result of MEFs VPS35[D620N] (no LysoTag) (Mock-IP) 5 and p-value < 0.05.The black dotted line box indicates proteins confidently detected in the LysoTag-IP samples but not in the Mock-IP samples.The log2 MS intensities of these proteins are shown in Supplementary Figure (B).(C) The volcano plot (Curtain link: https://curtain.proteo.info/#/e10d5d9a-3214-4a5d-88cb-7ca1b1460e1c) shows the proteome changes of all detected proteins in MEFs VPS35[D620N] MLi-2 versus DMSO LysoTag-IP experiments.The red dots represent mock-enriched proteins with fold change > 1.5 and p-value < 0.05.The gray dots represent proteins bound to magnetic beads, which do not show significant enrichment.(D) The raw MS intensities of two non-specific proteins that associate with the magnetic beads (MALSU1 and PENK) were shown for reference.

Fig. S5 .
Fig. S5.Relative levels of selected proteins in lysate and lysosomes of wild type and D620N MEFs and impact of MLi-2.(A to D) Violin plots of the indicated proteins from MS data presented in Fig. 1, A and C; and Fig. 2, B and D).Data were analyzed using two-tailed unpaired t-test (** p< 0.01, *** p< 0.001, **** p< 0.001).

Fig. S6
Fig. S6 Enrichment of proteins related to the retromer complex.Volcano-plots show the enrichment of VPS26A, VPS26B, and VPS29, proteins which are the part of the retromer complex apart from VPS35.Comparison between the (A) Whole cell lysate (WCL) and (B) LysoTag-IP from the VPS35 WT and D620N MEFs.Comparison between the (C) Whole cell lysate (WCL) and (D) LysoTag-IP from the VPS35 D620N MEFs treated with ±MLi-2.

Fig. S7
Fig. S7 Proteasome mediated degradation of RILPL1.VPS35 D620N MEFs were treated with either 50 µg/mL of cycloheximide alone, or cycloheximide+10 µM MG-132, or cycloheximide + protease inhibitor cocktails (PIC -5 µM of E64D + 10 µM of Leupeptin + 10 µM of pepstatin A) for 8 h and 12 h prior to lysis.Lysates were subjected to quantitative immunoblot analysis using the LI-COR Odyssey CLx Western Blot imaging and the indicated antibodies.Technical replicates represent cell extracts obtained from three different dishes of cells.Quantitation of immunoblotting data (performed using ImageStudioLite software version 5.2.5, RRID:SCR_013715) is shown as mean ± SEM.Data were analyzed using two-tailed unpaired student's t-test ( *p< 0.05, ***p< 0.001).

Fig. S9 .
Fig. S9.Further confirmation that RILPL1 interacts with TMEM55B but not RILP or RILPL2.(A) Domain structure of full length and truncated RILPL1 mutants.(B) HEK293 cells were transiently transfected with the indicated proteins and lysed 24h

Fig. S10 .
Fig. S10.Confirmation that the TMEM55B Conserved domain binds RILPL1 and the AlphaFold2 model of this interaction.(A) Domain structure of full length and truncated mutants of TMEM55B used in this study.(B) HEK293 cells were transiently transfected with the indicated proteins and lysed 24h post transfection.Halo-

Fig. S12 .
Fig. S12.Relative levels of proteins in D620N Lysates and lysosomes that have been reported to be recruited to the lysosome in LLOMe treated cells.(A to G) Violin plots of the indicated proteins from MS data presented in Figure 1 (data on left panels) and Figure 2 (data on right panels).Data were analyzed using two-tailed unpaired t-test (*** p< 0.001, **** p< 0.0001).