The V-ATPase complex component RNAseK is required for lysosomal hydrolase delivery and autophagosome degradation

Abstract

Autophagy is a finely orchestrated process required for the lysosomal degradation of cytosolic components. The final degradation step is essential for clearing autophagic cargo and recycling macromolecules. Using a CRISPR/Cas9-based screen, we identify RNAseK, a highly conserved transmembrane protein, as a regulator of autophagosome degradation. Analyses of RNAseK knockout cells reveal that, while autophagosome maturation is intact, cargo degradation is severely disrupted. Importantly, lysosomal protease activity and acidification remain intact in the absence of RNAseK suggesting a specificity to autolysosome degradation. Analyses of lysosome fractions show reduced levels of a subset of hydrolases in the absence of RNAseK. Of these, the knockdown of PLD3 leads to a defect in autophagosome clearance. Furthermore, the lysosomal fraction of RNAseK-depleted cells exhibits an accumulation of the ESCRT-III complex component, VPS4a, which is required for the lysosomal targeting of PLD3. Altogether, here we identify a lysosomal hydrolase delivery pathway required for efficient autolysosome degradation.

Fig. 1. Identification of RNAseK in a screen for autophagy regulators.

a Validation of the GFP-LC3 reporter MEF cell line. Cells expressing Cas9 alone (sgControl) or sgRNA targeting ATG7 (sgATG7) were left untreated or amino acid (AA) starved for 2 h in the presence or absence of Baf A1 or VPS34 inhibitor (VPS34i). Cell lysates were subjected to western blot analyses using the indicated antibodies. N = 1. b Work flow for loss of function screen. MEFs stably expressing GFP-LC3 and Cas9 were transduced with sgRNA library and subjected to AA starvation (16 h) to induce autophagy and FACS sorting into GFP positive (autophagy deficient) and GFP negative (autophagy competent) populations. Genomic extraction and amplification was used to identify sgRNA sequences. b was created with BioRender.com released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en. c Deep sequencing analysis of sorted GFP positive cells. Highlighted are top 100 identified hits, including Atg genes and Rnasek. d Western blot analyses of sgControl MEF cells and RNAseK knockout cells using the indicated antibodies. Cells were left untreated or AA starved for 2 h in the presence or absence of Baf A1. N = 3 biologically independent experiments. e Quantification of LC3-II levels normalised to actin in (d). f Quantification of p62 levels normalised to actin in (d). g Quantitative DNA assessment of Rnasek gene editing efficiency in C57BL/6 mice interscapular brown adipose tissue (iBAT). DNA was extracted from iBAT 2 weeks after AAV injection and analysed using online tools. Gene editing in sgRNaseK#3 was normalised to AAV SaCas9 control (sgControl). N = 3 animals. h Western blot analyses of iBAT derived from C57BL/6 mice 2 weeks after administration with the indicated AAV vectors using the indicated antibodies. N = 3 animals per condition. i Quantification of LC3-II and p62 protein levels normalised to GAPDH, as shown in (h). In all panels, mean + SEM is assessed by unpaired two-tailed Student’s t test. Source data are provided with this paper.

Fig. 2. RNAseK knockout results in the accumulation of autolysosomal structures.

a Representative immunofluorescence image of GFP-LC3 and endogenous LAMP1 in sgControl or sgRNAseK MEFs cultured in the absence of AA for 2 h. Scale bar: 10 µm. Quantification of PCC between GFP-LC3 and LAMP1 is shown on the right. N = 30 cells from three independent experiments. b Representative immunofluorescence image of endogenous p62 and LAMP1 in siControl or siRNAseK MEFs treated as in (a). Scale bar: 10 µm. Quantification of PCC between p62 and LAMP1 is shown on the right. N = 30 cells from three independent experiments. c Representative electron microscopy images of sgControl and sgRNAseK U2OS cells treated in the absence of AA for 3 h. Baf A1 is added as indicated. Arrows indicate autophagosomes (white arrows) and autolysosomes (black arrows). Scale bar: 1 µm. N = 1. d Model of studying IAM degradation whereby STX17 association with undegraded or degraded autolysosomes are observed as a lysotracker ring or dots, respectively. d was created with BioRender.com released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en. e Representative images of sgControl and sgRNAseK MEFs stably expressing GFP-STX17 and stained with Lysotracker red. Quantifications of the percentage of STX17 with Lysotracker rings or dots are shown on the right. Scale bar: 10 µm. N = 15 cells from three independent experiments. In all panels, mean + SEM is assessed by unpaired two-tailed Student’s t test. Source data are provided with this paper.

Fig. 3. RNAseK knockout does not affect general lysosomal acidification or proteolytic activities.

a Quantification of LysoSensor Green signal fold change in control (sgControl) or RNAseK knockout (sgRNAseK) cells relative to signal in Control cells. MEF cells were left untreated or AA starved for 2 h and incubated with LysoSensor Green for 30 min followed by FACS analyses of fluorescence intensity. Mean + SEM is assessed by paired two-tailed Student’s t test. N = 3 biologically independent experiments. b Ratiometric pH quantification using Oregon Green 488-dextran. Control or RNAseK knockout cells were incubated with Oregon Green 488-dextran for 24 h, followed by assessment of fluorescence emission at 520 nm upon excitation at either 440 nm (pH insensitive) or 484 nm (pH sensitive). Calibration using standard buffers was used to determine lysosomal pH. Baf A1 was included as indicated. N = 30 cells examined over three independent experiments. c Western blot analyses of Cathepsin B in sgControl or sgRNAseK cells. Cells were untreated or AA starved in the presence or absence of Baf A1. d Quantification of mature and pro-Cathepsin B relative to sgControl cells in (c). N = 3 biologically independent experiments. e Schematic diagram of Cathepsin B activity assay where cleavage of Cathepsin B substrate results in the fluorescence of the cleaved product. e was created with BioRender.com released under a Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International license https://creativecommons.org/licenses/by-nc-nd/4.0/deed.en. f Quantification of Cathepsin B substrate fluorescence in the indicated cells. Signal was normalised relative to sgControl cells. MEF cells were left untreated or AA starved for 2 h and incubated with the Cathepsin B substrate for 30 min followed by FACS analyses. Mean + SEM is assessed by paired two-tailed Student’s t test. N = 3 biologically independent experiments. g Western blot analyses of EGFR levels in sgControl and sgRNAseK cells. Cells were cultured without serum for 4 h, followed by stimulation with EGF (20 ng/mL) for the indicated times. N = 3 biologically independent experiments. h Quantification of EGFR levels following EGF stimulation in (g) expressed as a percentage of the EGFR levels at time 0 in the relative cell line. N = 3 biologically independent experiments. In all panels, mean + SEM is assessed by unpaired two-tailed Student’s t-test, unless stated otherwise. Source data are provided with this paper.

Fig. 4. Analyses of the subcellular localisation of RNAseK.

a–c Representative immunofluorescence images of wild type or endogenously tagged RNAseK-MYC MEFs cultured in AA-free media for 2 h. Cells were fixed and stained against MYC and the indicated endogenous proteins. Scale bar: 10 µm. The pearson’s colocalisation coefficient (PCC) values for the following colocalisations are: MYC-LAMP1 (PCC = 0.63), MYC-RAB7 (PCC = 0.42), and MYC-GM130 (PCC = 0.22). N = 30 cells from three independent experiments. d Volcano plot analyses of RNAseK-TurboID (TiD) proximity labelling hits identified by MS presented as relative values to hits obtained from MYC-TiD control cells. Highlighted hits are represented in magenta. The statistical significance was calculated by unpaired two-tailed Student’s t test. N = 1 including three technical repeats. e Top ten biological processes (left) and cellular components (right) of hits obtained in (d). GO enrichment analysis plotted according to −log₁₀ False Discovery Rate (FDR). The statistical significance was calculated by a one-way Fisher’s exact test. Source data are provided with this paper.

Fig. 5. PLD3 is required for autophagy.

a Quantitative proteomic analysis of enriched lysosomal fractions derived from sgControl or sgRNAseK MEFs. The statistical significance was calculated by unpaired two-tailed Student’s t test. b MS analyses of PLD3 levels in cell media derived from sgControl or sgRNAseK MEFs. Baf A1 treatment was added as indicated (24 h). Absolute PLD3 intensity is shown as label-free quantitation (LFQ) values. N = 1 including three technical repeats. c qRT-PCR analyses of mRNA from MEFs transfected with non-targetting siRNA (siControl) or siPLD3. Pld3 mRNA levels were normalised to Actin. Fold change in expression was compared to siControl cells. N = 3 biologically independent experiments. d Western blot analyses of cell lysates derived from siControl MEFs or cells transfected with siRNA sequences targeting PLD3. Baf A1 treatment (2 h) was included as indicated. Quantification of p62 and LC3-II band intensity normalised relative to siControl (lane 1) is shown. N = 3. e Western blot analyses of lysates derived from SH-SY5Y cells transfected with siControl or pool siRNAs targeting RNAseK or PLD3. Baf A1 treatment (3 h) was included as indicated. Quantification of LC3-II band intensity normalised relative to siControl (lane 1) is shown. N = 3. f Representative electron microscopy images of siControl and siPLD3 MEFs treated in the absence of AA (3 h). Arrows indicate autophagosomes (white) and autolysosomes (black). Scale bar: 1 µm. N = 1. g Representative immunofluorescence images of MEFs transfected with siControl or siPLD3. Cells were cultured in the absence of AA (2 h). Scale bar: 10 µm. Quantification of PCC between p62 and LAMP1 in siControl and siPLD3 is shown on the right. N = 30 cells from three independent experiments. h Representative images of sgControl and sgPLD3 MEFs expressing GFP-STX17 and stained with Lysotracker red. Quantifications of the percentage of STX17 with Lysotracker rings or dots are shown on the right. Scale bar: 10 µm. N = 15 cells from three independent experiments. i Western blot analyses of EGFR levels in siControl and siPLD3 MEF cells. Cells were cultured without serum (4 h), followed by stimulation with EGF (20 ng/mL). Quantification of EGFR levels as a percentage of time 0 is shown on the right. N = 3 biologically independent experiments. In all panels, mean + SEM is assessed by unpaired two-tailed Student’s t test. Source data are provided with this paper.

Fig. 6. VPS4a accumulates on lysosomal membranes in the absence of RNAseK.

a Representative confocal microscopy images of GFP-VPS4a expressed in sgControl or sgRNAseK U2OS cells and cultured in AA starvation media for 3 h. Baf A1 was added as indicated. Scale bar: 10 µm. Quantification of VPS4a puncta per cell is shown on the right. N = 30 cells from three independent experiments. b–d Representative immunofluorescence images of GFP-VPS4a expressed in U2OS cells as in (a) and stained using antibodies against the indicated endogenous proteins. Scale bar: 10 µm. Quantifications of PCC between GFP-VPS4a^WT and the indicated markers are shown on the right. N = 30 cells from 3 independent experiments. e Volcano plot of interacting proteins identified by MS following GFP-TRAP pulldown of GFP-VPS4a transfected into sgControl or sgRNAseK MEF cells. Selected hits showing significant association with GFP-VPS4a in sgRNAseK cells relative to sgControl cells are highlighted in magenta. The statistical significance was calculated by unpaired two-tailed Student’s t test. N = 1 including three technical repeats. f Western blot analyses of U2OS cells transiently expressing the indicated constructs. Cells were AA starved in the presence or absence of Baf A1 for 3 h. N = 1. g HEK293T cell lysates transfected with GFP or GFP-RNAseK plasmids were mixed with lysates expressing mCherry-VPS4a followed by GFP-TRAP pulldown (PD) and western blotting analyses using the indicated antibodies. N = 3. Quantification of VPS4a pulldown relative to control is shown on the right. h HEK293T cell lysates transfected with GFP or GFP-VPS4a plasmids were mixed with lysates expressing RNAseK-MYC-TurboID. Cell lysates were treated as in (g). N = 3. Quantification of RNAseK pulldown relative to control is shown below. i Representative immunofluorescence images of endogenously tagged RNAseK-MYC MEF cells transfected with GFP-VPS4a^WT (top panel) or GFP-VPS4a^E228Q (bottom panel) and cultured in the absence of AA for 2 h. Cells were fixed and stained using antibodies against MYC. Scale bar: 10 µm. Quantification of PCC between RNAseK-MYC and GFP-VPS4a is shown on the right. N = 30 cells from three independent experiments. In all panels, mean + SEM is assessed by unpaired two-tailed Student’s t test. Source data are provided with this paper.