Autophagy-dependent cancer cells circumvent loss of the upstream regulator RB1CC1/FIP200 and loss of LC3 conjugation by similar mechanisms

Abstract

Macroautophagy/autophagy degrades proteins and organelles to generate macromolecular building blocks. As such, some cancer cells are particularly dependent on autophagy. In a previous paper, we found that even highly autophagy-dependent cancer cells can adapt to circumvent autophagy inhibition. However, it remains unclear if autophagy-dependent cancer cells could survive the complete elimination of autophagosome formation. We extended our previous findings to show that knockout (KO) of both the upstream autophagy regulator RB1CC1/FIP200 and the downstream regulator and mediator of LC3 conjugation, ATG7, strongly inhibits growth in highly autophagy-dependent cells within one week of editing. However, rare clones survived the loss of ATG7 or RB1CC1 and maintained growth even under autophagy-inducing conditions. Autophagy-dependent cells circumvent the complete loss of autophagy that is mediated by RB1CC1 KO, similar to the loss of ATG7, by upregulating NFE2L2/NRF2 signaling. These results indicate that cancer cell lines could adapt to the complete loss of autophagy by changing their biology to adopt alternative ways of dealing with autophagy-mediated cellular functions.

Abbreviations: CGS: CRISPR growth score; CQ: chloroquine; CRISPR: clustered regularly interspaced short palindromic repeats; EBSS: Earl's balanced salt solution; EEF2: eukaryotic translation elongation factor 2; FOXO3/FOXO3a: forkhead box O3; GFP: green fluorescent protein; KEAP1: kelch Like ECH associated protein 1; KO: knockout; MAP1LC3B/LC3B: microtubule associated protein 1 light chain 3 beta; MEFs: mouse embryonic fibroblasts; NFE2L2/NRF2: nuclear factor, erythroid 2 like 2; NLS: nuclear localization signal; PCNA: proliferating cell nuclear antigen; PE: phosphatidylethanolamine; POLR2A: RNA polymerase II subunit A; PTEN: phosphatase and tensin homolog; ROS: reactive oxygen species; SNARE: soluble NSF attachment protein receptor; SQSTM1: sequestosome 1; STX17: syntaxin 17; TBHP: tert-butyl hydroperoxide; ULK1: unc-51 like autophagy activating kinase 1; ULK2: unc-51 like autophagy activating kinase 2; WT: wild type.

Keywords: ATG7; Adaptation; CRISPR-Cas9; NFE2L2/NRF2; ROS; STX17.

Figure 1.

The generation of ATG7 and RB1CC1 KO clones from autophagy-dependent, BT549, cells. (A) Schematic of the acute CRISPR assay showing double-labeled cells transfected with gRNAs targeting both GFP and a gene-of-interest and the mCherry⁺ GFP⁻ population can be tracked with live-cell Incucyte imaging for 7 d immediately following transfection. Extensive proof of principle studies were conducted to confirm the efficiency and accuracy of the assay. Next, the assay was used to identify autophagy-dependent and independent cell lines by targeting 12 core autophagy proteins in an arrayed assay compared to known essential and non-essential genes. Normalized CRISPR growth scores were generated for each gene to quantify the data [2]. (B) The acute CRISPR assay performed in mCherry⁺ GFP⁺ BT549 cells with gRNAs targeting GFP along with either PTEN, PCNA, ATG7, or RB1CC1. The mCherry⁺ GFP⁻ population is graphed over time. The data are represented as mean ± SEM for technical replicates (N of 3), and the graphs shown are representative of 3 individual experiments. Statistical analysis: 2-way ANOVA and the significance at the last time point is shown. **** p$\le$0.0001. (C-E) Western blot analysis in clones of BT549 cells and in (E) the cells were either starved for 24 h in EBSS with or without bafilomycin-A1 (10 nM) for 2 h. (F) Flow cytometry in cells with stable expression of GFP-mCherry-LC3 and starved with EBSS for 24 h. Representative histograms of the derived ratio of mCherry/GFP is shown (representative of 3 experiments).(G) Left: Confocal imaging of mCherry-NLS-labeled BT549 WT, ATG7 KO, or RB1CC1 KO cells retrovirally transduced with STX17TM-GFP after treatment with EBSS for 24 h. Scale bars: 2 μm. Right: Quantification of STX17TM-GFP puncta per cell. 5 fields were counted per condition and each bullet point represents a cell. Data are representative of >3 experiments.

Figure 2.

BT549 ATG7 KO and RB1CC1 KO clones are not more sensitive to nutrient deprivation but are more dependent on NFE2L2 signaling. (A) Incucyte mCherry⁺ growth curves of BT549 WT, ATG7 KO, RB1CC1 KO clones in full growth media. The data are represented as mean± SEM for technical replicates (N of 3), and the graphs shown are representative of >3 individual experiments. Statistical analysis: 2-way ANOVA and the significance at the last time point is shown. (B) The area under the curve (AUC) of mCherry⁺ cell count of cells treated with EBSS for 4 d graphed as a fold-change compared to the AUC of the mCherry⁺ cell count in full growth media. The data are graphed as mean± SEM for two combined Incucyte experiments. Statistical analysis: 1-way ANOVA. (C) The AUC of mCherry⁺ cell count of cells grown in galactose-containing media for 6 d graphed as a fold-change compared to the AUC of mCherry⁺ cell counts in glucose-containing media. The data are graphed as mean± SEM for two combined Incucyte experiments. Statistical analysis: 1-way ANOVA. (D) Incucyte scans of BT549 WT, ATG7 KO, RB1CC1 KO clones were used to calculate the AUC of mCherry⁺ cell counts 48 h after incubation in 1% O₂ and shown as a fold-change over the AUC of mCherry+ cell counts from cells grown in normoxic, 5% O_2, conditions. The data are graphed as mean± SEM for two combined Incucyte experiments. Statistical analysis: 1-way ANOVA. (E) Incucyte mCherry⁺ growth curves of BT549 WT, ATG7 KO, RB1CC1 KO clones treated with chloroquine (50 μM) graphed as a fold-change over DMSO-treated cells for each clone. The data are represented as mean± SEM for technical replicates (N of 3) and the graphs shown are representative of 3 individual experiments. Statistical analysis: 2-way ANOVA and the significance at the last time point is shown. (F) Western blot analysis in WT, ATG7 KO, and RB1CC1 KO clones showing an increase in NFE2L2 protein expression. Dotted line indicates where unnecessary lanes that were not full RB1CC1 KO clones were eliminated. (G) Flow cytometry analysis of cellROX green was measured to identify levels of ROS. After gating live singlets based on forward and side scatter, a cellROX⁺ gate was drawn such that 10% of the untreated WT cells were counted as positive and used to quantify the percent of ATG7 and RB1CC1 KO cells within that gate. The gate is depicted on each graph and color-coded in green to show the gated cells. Right: The graphs shown on the left for each clone are overlaid together. The graph shown is representative of 2–3 individual experiments. (H) Western blot analysis in BT549 WT, ATG7 KO, RB1CC1 KO clones 5 d after transduction with either non-silencing shRNAs or NFE2L2-targeted shRNAs. The blots shown are representative of 2–4 experiments. (I) qRT-PCR for mRNA expression of NQO1 normalized to 18 S rRNA expression 4 d after viral transduction. The data are represented as mean± SD (N of 2) and the graph shown is representative of 2–3 experiments. Statistical analysis: One-way ANOVA. (J) In BT549 WT, ATG7 KO, RB1CC1 KO cells treated with TBHP (50 μM), CellEvent Caspase-3/7 Green Detection Reagent cell event counts are shown normalized to mCherry⁺ cell counts. The data are represented as mean± SEM for technical replicates (N of 3) and the graphs shown are representative of 3 individual experiments. Statistical analysis: 2-way ANOVA and the significance at the last time point is shown. (K and L) Live cell incucyte analysis of BT549 WT, ATG7 KO, RB1CC1 KO clones 4 d after transduction with either non-silencing shRNAs or NFE2L2-targeted shRNAs. (K) Incucyte mCherry⁺ growth curves are shown normalized to TP 0. The data are represented as mean ± SEM for technical replicates (N of 3) and the graphs shown are representative of 2–4 individual experiments. Statistical analysis: 2-way ANOVA and the significance at the last time point is shown. (L) CellEvent Caspase-3/7 Green Detection Reagent cell event counts are shown normalized to mCherry⁺ cell counts. The data are represented as mean± SEM for technical replicates (N of 3) and the graphs shown are representative of 2–4 individual experiments. Statistical analysis: 2-way ANOVA and the significance at the last time point is shown. *p$\le$0.05, **p$\le$0.01, *** p$\le$0.01, **** p$\le$0.001**** p$\le$0.0001.