Fig 1.
Effect of plumbagin treatment on viability and size of cells.
(A) Real-time monitoring of relative cell impedance (showed as a cell index) using the RTCA system. (B) MTT-assessed response to 6h plumbagin treatment. (C) MTT-assessed response to 24h plumbagin treatment. (D) IC50 values according to RTCA and MTT and length of treatment. (E) Time-dependent changes in the quantity of large cells with intact nuclei (SYTO 16++/FSC+) and intact cells (AnnexinV-/PI-) assessed by flow-cytometry. (F) Numbers of large healthy cells depicted as SYTO 16++ /FSC+ (red) cluster at flow-cytometric dot plot at 6h time-point; Forward-scattered light (FSC) is proportional to cell-surface area or size. (G) Granularity of large healthy cells depicted as SYTO 16++ /SSC+ (red) cluster at flow-cytometric dot plot at 6h time-point; Side-scattered light (SSC) is proportional to cell granularity or internal complexity. (H) Morphology of PC-3 cells after 6h plumbagin treatment, 20x magnification, phase contrast microscopy. (I) Numbers of large healthy cells depicted as SYTO 16++/FSC+ (red) cluster at flow-cytometric dot plot at 24h time-point; Forward-scattered light (FSC) is proportional to cell-surface area or size. (J) Granularity of large healthy cells depicted as SYTO 16++/SSC+ (red) cluster at flow-cytometric dot plot at 24h time-point; Side-scattered light (SSC) is proportional to cell granularity or internal complexity. (K) Morphology of PC-3 cells after 24h plumbagin treatment; giant PC-3 cells with polyploid giant cancer cell (PGCCs)-like morphology are highlighted by arrows. 20x magnification, phase contrast microscopy.
Fig 2.
Reactive oxygen species (ROS)-induced mitophagy.
(A) Phase contrast microscopy of PC-3 cell after plumbagin treatment. (B) General accumulation of ROS after plumbagin treatment monitored by confocal microscopy by using CellROX Deep Red Reagent. Areas with ROS accumulation are highlighted by arrows. (C) Mitochondria staining monitored by confocal microscopy using MitoTracker Green; area associated with ROS in Fig 2B are highlighted by arrows. (D) Endoplasmic reticulum (ER) staining monitored by confocal microscopy using ERTracker Red; areas associated with ROS in Fig 2B are highlighted by arrows. (E) Untreated PC-3 cell, cross-section of undamaged mitochondria (highlighted by red arrow); Transmission Electron Microscope (TEM) visualization. (F) plumbagin-treated PC-3 cell, mitochondria coated by ER membrane with ribosomes (highlighted by red arrow); TEM visualization. (G) Plumbagin-treated PC-3 cell, gradual degradation of mitochondria in autophagosomes visualised by TEM (red arrows); Swollen mitochondria as a marker of damage (yellow arrow).
Fig 3.
Time-lapse of cell interactions.
For detailed time-lapse Videos see S1–S4 Videos. (A) Time-lapse imaging of entosis; internalized cell (red arrow) played an active role in its engulfment, which resulted in complete internalization. Both types of cells (engulfing and engulfed) were viable for a long time and lived by about five hours longer than the other observed plumbagin-treated tumour cells. (B) Time-lapse imaging of cell fusion with cannibalism (digestion of engulfed cell); during fusion-cannibalism of living cells, the cannibalic cell (red arrow) came in contact with the target cell (blue arrow). The next step was gradual engulfment of the target cell. The nucleus of the target cell appeared initially unaltered whereas the engulfing cell’s nucleus began to change into a semilunar shape. Bird eye structure was observed as a consequence of target cell vacualisation (see green arrow). (C) Time-lapse imaging of cannibalism without fusion; the dying cell (blue arrow) was attacked and exploited by the cannibalic cell (red arrow). The target cell was dead after the attack. (D) Time-lapse imaging of oncosis; oncotic cells formed typical cytoplasmic blebs that usually lead to necrosis (see red arrow). (E) Time-lapse imaging of reverse oncosis; initial forming of oncotic blebs (see red arrow) did not lead to necrosis; the bleb was absorbed and the cell remained viable. (A-E) Multimodal holographic microscopy, 10x magnification. (F) Communication between PC-3 cells; visualised by TEM (see red arrows). (G) Vesicular transfer between PC-3 cells; visualised by TEM (see red arrows). (H) Speed of the migration of untreated PC-3 cell population; assessed from holographic microscopy data by CellProfiller software by measurement of “distance travelled”parameter. (I) Speed of the migration of PC-3 cell population after 2 μM plumbagin treatment.
Fig 4.
Mechanistic characterization of engulfed and engulfing cells in entosis.
(A) Trajectory travelled of both engulfing and engulfed cell until cell fusion. See differences in the travelled distance and in directionality of individual cells. Directionality describes "purposefulness" of the movement where 0% indicate random movement and 100% indicate straight line trajectory between starting and ending position. Position (0.0, 0.0) indicate place of cell fusion. (B) Changes in cell mass and (C) cell area of engulfed and engulfing cell.
Fig 5.
Autophagy and self-renewal after plumbagin treatment.
(A) Time dependent dynamics of CYTO-ID Green (CYTO-ID+, CYTO-ID++) after 2 μM plumbagin treatment. See autophagy peak at 8h-time point. (B) Time dependent dynamics of cell deaths after 2 μM plumbagin treatment; total CytoID staining (CYTO-ID+ and CYTO-ID++) depicts autophagy, AnnexinV+/PI- depicts apoptosis or early oncosis, and AnnexinV+/PI+ depicts necrosis (raw data and gating strategy in S2 Appendix). (C) Amount of autophagic cells after bafilomycin treatment; red cluster depicts CYTO-ID++ population, blue cluster Cyto-ID+ population. (D) Amount of autophagic cells in control (not-treated population); red cluster depicts CYTO-ID++ population, blue cluster CYTO-ID+ population. (E) Amount of autophagic cells after plumbagin treatment; red cluster depicts CYTO-ID++ population, blue cluster CYTO-ID+ population. (F) Western blot for LC3-I and LC3-II isoforms at 12 time-points after 2 μM plumbagin treatment (raw data in S3 Appendix). (G) Graphic representation of western blot results for LC3-I and LC3-II isoforms at 12 time-points after 2 μM plumbagin treatment. (H) Time dependent dynamics of POU5F, SOX2, NANOG, and BECN1 gene expression after 2 μM plumbagin treatment. (I) Time dependent dynamics of BIRC5, HIF1A, CCL2, and MAP1LC3 gene expression after 2 μM plumbagin treatments. (J) Principal component analysis—projection of variables on the two-factor plane. See distinct clustering of genes with flow-cytometric measurements based on metabolic stress and reprogramming (for details see results). (K) Principal component analysis—projection of time-points on the two-factor plane. The first and the second factor are designated as “metabolic stress high-low” and “non-reprogrammed-reprogrammed”, respectively (for details see results). (L) Cluster analysis of gene expression. The “reprogramming cluster”involved POU5F, SOX2, NANOG, and CCL2; the “autophagic and hypoxia cluster”involved HIF1A, MAP1LC3, and BECN1; the third cluster involved BIRC5 gene. (M) Cluster analysis of gene expression of 12 time-points after 2 μM plumbagin treatment and PGCCs selected by hypoxia and starvation; based on correlations of gene expression patterns, similarity between PGCCs and PC-3 cells after 20h plumbagin treatment was found. (N) Morphology of cells after 20h plumbagin treatment; cells with polyploid giant cancer cell (PGCCs)-like morphology are highlighted by arrows. (O) Morphology of PC-3 cells after 1 month of hypoxia and starvation; cells with PGCCs-like morphology are highlighted by arrows.
Fig 6.
Possible cell fate under oxidative stress.
Prolonged oxidative stress (ROS) leads to severe cell damage and depletion of cell energy reserves. Nevertheless, many cell injuries caused by ROS could be sublethal, especially because PC-3 cells have non-functional p53 and therefore disrupt the triggering of apoptosis. Even if a damaged cell is driven to oncosis, reversion of this process is possible, particularly if the cell is able to restore ATP production. A way to gain enough energy for survival could be autophagy. Similar to autophagy, digestion of the cytoplasm of neighbouring cells can provide a source of amino acids. Retention of a foreign nucleic acid by cannibalistic engulfment could result in aneuploid or polyploid state. Furthermore, reduction in membrane and cell stiffness due to protein catabolism by autophagy could reflect increased entotic activity. Cell in cell structure results in the decreased surface-to-volume ratio, thereby minimizing cell membrane requirements. Furthermore, a live cell internalized by entosis could disrupt host cell division. Subsequently, cytokinesis often fails, which can lead to the formation of polyploid giant cancer cells (PGCCs). PGCCs often dye by apoptosis or senescence, but a small fraction of these cells is able to survive and even produce aneuploid progeny. Senescence, polyploidy and self-renewal seem to be three steps to immortality of cancer cells. Autophagy could play an important role in all of them.