Fig 2: ULK1 activity regulates the degradation of polyQ proteins. a FLAG-ATG14 S29 WT, SA, or SE, Beclin 1-AsRed, and dual myc-Vps34 his-Vps15 plasmid were overexpressed in mCFP-65Q expressing HeLa cells then treated with rapamycin overnight. Cell lysates were separated into soluble and insoluble fractions. PolyQ was measured in the insoluble fraction. b Quantification of polyQ degradation in the insoluble fraction. Degradation was calculated as the ratio of control sample vs. rapamycin sample **p < 0.01 (n = 4). Data are represented as mean +/− SEM. c Myc empty vector, myc-ULK1 WT or KI were overexpressed in mCFP-65Q expressing HeLa cells and cells were lysed 48 h after transfection. Cell lysates were separated into soluble and insoluble fractions. PolyQ was measured in the insoluble fraction. d Quantification of polyQ levels in the insoluble fraction. Values were normalized to empty vector condition. *p < 0.05 n.s. not significant (n = 4). Data are represented as mean +/− SEM

Fig 3: Mutations in the NRBF2 CC domain to render it monomeric lead to weakened binding to PIK3R4/VPS15. (A) Leu-to-Ala mutations in the NRBF2 CC domain targeting the five leucine zipper pairs. Double (L178A_L185A, L201A_L206A), triple (3A) and pentameric (5A) mutants were generated. (B) Co-IP experiments to assess the self-association of NRBF2 mutants. FLAG- and GFP-tagged NRBF2 constructs were co-transfected into HEK293T cells. Their self-association was probed using anti-FLAG M2 magnetic beads for immunoprecipitation (IP), followed by immunoblotting (IB) using anti-GFP antibody. Double mutant (L178A_L185A), 3A and 5A all lost self-association. (C) Co-IP experiments to assess the binding of NRBF2 mutants to endogenous PIK3R4/VPS15. FLAG-tagged NRBF2 constructs were transfected into HEK293T cells. Their binding to endogenous PIK3R4/VPS15 was probed using anti-FLAG M2 magnetic beads for immunoprecipitation (IP), followed by immunoblotting (IB) using anti-PIK3R4/VPS15 antibody. 5A pulled down significantly less of endogenous PIK3R4/VPS15 compared to the wild type. (D) Quantification of PIK3R4/VPS15:FLAG-NRBF2 protein levels from data in (C).

Fig 4: NRBF2 with its CC domain replaced with dimeric or tetrameric Gcn4 module shows strengthened binding to PIK3R4/VPS15. (A) Design scheme to generate NRBF2 constructs with its CC domain replaced by dimeric or tetrameric Gcn4 modules. (B) The oligomeric state of NRBF2_Gcn4_Dimer and _Tetramer is confirmed by dynamic light scattering profiles. The estimated molecular weight matches the expected dimeric or tetrameric state. (C) Co-IP experiments to assess the binding of NRBF2_Gcn4_Dimer and _Tetramer to endogenous PIK3R4/VPS15. FLAG-NRBF2 constructs and UVRAG-GFP were transfected into HEK293T cells. HEK293T cells were untreated or treated with 100 nM rapamycin (2 h) or treated with EBSS medium (2 h) before harvesting. Their binding to endogenous PIK3R4/VPS15 was probed using anti-FLAG M2 magnetic beads for immunoprecipitation (IP), followed by immunoblotting (IB) using anti-PIK3R4/VPS15 antibody. (D) Quantification of PIK3R4/VPS15:FLAG-NRBF2 protein levels from data in (C).

Fig 5: NRBF2 does not have direct interactions with BECN1 or ATG14. (A) ITC profiles to assess the direct interaction between NRBF2 and ATG14. No bindings were detectable. (B) 1H-15N HSQC NMR spectra of 15N-labeled ATG14-1–95 in the absence (black) or presence of NRBF2 MIT domain at 8x molar excess (red). No chemical shifts were observed. (C) Affinity-isolation assay to assess the direct binding between NRBF2 and purified full-length BECN1, ATG14, PIK3C3/VPS34 and the PIK3C3/VPS34-PIK3R4/VPS15 complex. SUMO-tagged NRBF2 only pulled down the PIK3C3/VPS34-PIK3R4/VPS15 complex but not BECN1, ATG14 or PIK3C3/VPS34.