Volume 24, Issue 11, Pages 3033-3044 (September 2018) TLR Crosstalk Activates LRP1 to Recruit Rab8a and PI3Kγ for Suppression of Inflammatory Responses  Lin Luo, Adam A. Wall, Samuel J. Tong, Yu Hung, Zhijian Xiao, Abdullah A. Tarique, Peter D. Sly, Emmanuelle Fantino, María-Paz Marzolo, Jennifer L. Stow  Cell Reports  Volume 24, Issue 11, Pages 3033-3044 (September 2018) DOI: 10.1016/j.celrep.2018.08.028 Copyright © 2018 The University of Queensland Terms and Conditions

Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 1 PI3Kγ Is Recruited by Active Rab8 and Its Adaptor 101 during TLR4 Activation (A) Co-immunoprecipitation of PI3Kγ adaptors p101 or p87 by myc-Rab8a in unstimulated or LPS-induced Rab8a-deficient RAW 264.7 cells stably expressing myc-Rab8a. 1% of lysates and 20% of eluates were loaded. (B) Co-immunoprecipitation of TLR4, p101, or p87 by PI3Kγ in unstimulated or LPS-induced cells. 1% of lysates and 50% of eluates were loaded on the gel. The gels (A and B) are representative of three independent experiments (n = 3). (C) RAW 264.7 cells transiently coexpressing GFP-Rab8a and Rab5-mcherry were either unstimulated or pre-treated with LPS for 30 min and cells labeled for 15 min with dextran-647. (D) RAW 264.7 cells co-transfected with GFP-Rab8a and TagRFP-T-Akt1 were examined by live-cell spinning-disc confocal microscopy. Cells were imaged over 10 min at 5-s intervals, and in some cells, LPS (10 ng/mL) was added and imaging was then continued. Insets show examples of colabeled macropinosomes. GFP and RFP pixel intensities were measured for regions of interest (ROIs) before and after LPS. The scale bars represent 10 μm. For statistical analysis, a p value < 0.05 was considered statistically significant (∗p < 0.05). Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 2 Identification of LRP1 as the Crosstalk Receptor for Rab8a in Macrophages (A) GST-Rab8a coupled to GSH-Sepharose was used for pull-downs from LPS-activated RAW 264.7 cell extracts. Bound proteins were separated by SDS-PAGE after PreScission protease elution method. Excised bands were identified by mass spectrometry. Major bands at 87 kDa and >250 kDa, absent from the GST control, were identified as LRP1. (B) A selection of candidate known and unknown binding partners identified in the pull-down are tabulated and ranked according to their mass spectrometry scores, as detailed in Experimental Procedures. Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 3 Rab8a/PI3Kγ Is Recruited by LRP1 in a LPS-Dependent but Nucleotide-Independent Manner (A) Co-immunoprecipitation of Rab8 by LRP1 in the presence or absence of LPS (30 min)-treated RAW 264.7 cells. (B) Pull-down using GST-tagged LRP1 cytoplasmic domain (amino acids 4,442–4,545) from extracts of LPS-activated cells. (C) Nucleotide exchange of GST-Rab8a with GTP or GDP followed by pull-down of LRP1 and PI3Kγ from activated cell lysates. Gels (A–C) are representative of three independent experiments (n = 3). Graphs represent mean ± SEM. 1% of lysates and 20% of eluates were loaded on each gel. (D) Confocal z-slice from RAW264.7 macrophages transiently expressing mCherry-2xFYVE and labeled for 15 min with 10 μg/mL 70K MW Oregon Green Dextran. Fixed cells were stained with anti-LRP1. Insets highlight an example of a triple-labeled peripheral macropinosome. The scale bar represents 10 μm. (E) RAW 264.7 macrophages transiently expressing V5-tagged Rab8a (green) were treated or not with LPS and fixed for staining with anti-LRP1 (red). z-slices from DeltaVision deconvolution microscopy and insets highlight macropinosomes with LRP1 and Rab8a labeling. The scale bars represent 10 μm. For statistical analysis, a p value < 0.05 was considered statistically significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001). Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 4 TLR4-Induced LRP1 Phosphorylation at Y4507 Regulates Rab8a Recruitment (A) Total LRP1 tyrosine phosphorylation was assessed by immunoprecipitation of LRP1, followed by pY immunoblotting in unstimulated or LPS-induced (30 min) macrophages. (B) RAW264.7 cells were subjected to a 30-min time course of LPS. SH2 phosphotyrosine probes were used to pull-down tyrosine-phosphorylated LRP1 at Y4507 from cell lysates. (C) LRP1 phosphorylation at Y4507 in WT and TLR4−/− bone-marrow-derived macrophages detected with SH2 phosphotyrosine probes. (D) LRP1 phosphorylation at Y4507 in WT, MAL−/−, and TRIF−/− immortalized bone-marrow-derived macrophages detected with SH2 phosphotyrosine probes. (E) GST pull-down using WT, phosphorylation-deficient mutants Y4474F, Y4507F, or DFM of GST-tagged LRP1 cytoplasmic domain with wild-type Rab8a from LPS-activated macrophage lysates. GST alone was used as a negative control. Quantification of western blots was performed using densitometry from three independent experiments (n = 3). Graphs represent mean ± SEM. The gels (A–E) are representative of three independent experiments (n = 3). (A and E) 1% of lysates and 20% of eluates were loaded on each gel. For statistical analysis, a p value < 0.05 was considered statistically significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001). Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 5 LRP1 Regulates Rab8 Activation during Activation of Multiple TLRs (A–C) RAW264.7 cells were subjected to a 30-min time course of Pam3CSK4 (A), Poly: IC (B), and CpG (C). SH2 phosphotyrosine probes were used to pull-down LRP1 phosphorylated at Y4507 from cell lysates. (D–G) Representative immunoblots of active Rab8 (GTP-loaded Rab8) in TLR2 (Pam3CYSK4; D)-, TLR3 (Poly IC; E)-, TLR4 (LPS; F)-, and TLR9 (CpG; G)-activated WT and LRP1-deficient mouse embryonic fibroblasts (MEFs). The gels (A–G) are representative of three independent experiments. 1% of lysates and 20% of eluates were loaded on each gel. Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 6 LRP1 Is a Potent Regulator of Akt/mTOR in Inflammation Densitometric quantification of phosphorylation in WT and CRISPR-Cas9-mediated LRP1 knockout macrophages. Cells were subjected to a 60-min time course of LPS. Cell extracts were analyzed for activation of Akt, Erk1/2 (A), mTOR (B), and mTOR substrates (C) p70S6K, PRAS40 phosphorylation. Densitometric quantification of phosphorylation is at each time point. Phosphoprotein levels of Akt, Erk1/2, mTOR, p70S6K, and PRAS40 were ratioed against total proteins. Graphs represent mean ± SEM (n = 3). For statistical significance, a p value < 0.05 was considered statistically significant (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, and ∗∗∗∗p < 0.0001). Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions

Figure 7 LRP1 Biases Cytokine Profiles to Constrain Inflammation (A) BMMs derived from WT and PI3Kγ−/− animals were subjected to an 8-hr treatment of LPS. Cytokine concentrations of IL-6, IL-12, and IL-10 in medium were measured by ELISA post-LPS. Statistical analyses were performed using Student’s t test. Each data point is the mean ± SEM (n = 5). (B) Control and LRP1-deficient RAW 264.7 cell lines were analyzed by qRT-PCR for transcriptional response of the cytokines IL-6, IL-12, and IL-10. Graphs represent mean ± SEM (n = 4 each). Time course profiles were analyzed by two-way ANOVA. Two-way ANOVA was analyzed by Sidak’s post-test for multiple comparisons (∗p < 0.05%; ∗∗p < 0.01; ∗∗∗p < 0.001; ∗∗∗∗p < 0.0001). Cell Reports 2018 24, 3033-3044DOI: (10.1016/j.celrep.2018.08.028) Copyright © 2018 The University of Queensland Terms and Conditions