Chemotherapy exposure increases leukemia cell stiffness

Abstract

Deformability of blood cells is known to influence vascular flow and contribute to vascular complications. Medications for hematologic diseases have the potential to modulate these complications if they alter blood cell deformability. Here we report the effect of chemotherapy on leukemia cell mechanical properties. Acute lymphoblastic and acute myeloid leukemia cells were incubated with standard induction chemotherapy, and individual cell stiffness was tracked with atomic force microscopy. When exposed to dexamethasone or daunorubicin, leukemia cell stiffness increased by nearly 2 orders of magnitude, which decreased their passage through microfluidic channels. This stiffness increase occurred before caspase activation and peaked after completion of cell death, and the rate of stiffness increase depended on chemotherapy type. Stiffening with cell death occurred for all cell types investigated and may be due to dynamic changes in the actin cytoskeleton. These observations suggest that chemotherapy itself may increase the risk of vascular complications in acute leukemia.

Figure 1

Chemotherapy-induced cell death increases the stiffness of leukemia cell populations measured by AFM and microfluidic channels. (A) An illustration of the AFM setup (not to scale). A single cell sitting within a microwell is immobilized for force microscopy with an AFM cantilever. A polydimethylsiloxane (PDMS) collar is pressed on the glass to create an open-air chamber. Tubes entering and exiting the chamber continually pass media through, keeping the media fresh over the long time scale of the experiments. The piezoelectric stage moves vertically, causing the cantilever to deflect against the cell. The stage is maintained at 37°C throughout the experiment. (B) An epifluorescence/brightfield overlay of a typical experiment. Seen here are an AFM cantilever tip and 2 dead K562 cells (PI positive, fluorescent), with the left cell immobilized in a microwell. An empty microwell is at the top. Scale bar is 20 μm. (C) Two typical cell indentation acquisitions. As the piezoelectric platform moves the cells up against the cantilever (in the direction of the arrow), the cantilever deflects. When the curves are fit to an elastic Hertzian model, the stiffness of the cells can be determined. The stiffness of a pre-B-ALL cell exposed to 1 μM dexamethasone (red) was 4.3 kPa whereas the stiffness of a control (not exposed to chemotherapy) pre-B-ALL cell (green) from the same patient was 0.2 kPa. (D) Dead (red) lymphoid leukemic cells exposed to 1 μM dexamethasone are significantly stiffer than untreated (green) cells. (E) Dead (red) myeloid leukemic cells exposed to 1 μM daunorubicin are significantly stiffer than untreated (green) cells. Error bars represent standard error. (n > 15, P < .05 for all comparisons of dead/untreated populations). (F) Dual brightfield/epifluorescence microscopy of dexamethasone-exposed pre-B-ALL cells that were passed, from left to right, through PDMS microfluidic channels modeling a branching microvasculature network. Dead (PI⁺) cells (red arrows) were more likely than live (unstained) cells (green arrows) to initiate obstruction and cause cell aggregation in the 5-μm wide by 12-μm tall, capillary-sized channels. Frame from panel G was taken 15 seconds after that seen in panel F, illustrating the relative mobility of 2 live cells, one of which has left the field of view, compared with dead cells that remain fixed in place. Scale bar is 10 μm.

Figure 2

Stiffness of leukemic cells increases with progression of cell death and is attenuated by disruption of the actin cytoskeleton. (A) A typical stiffness trace of a single M5 AML cell exposed to 1 μM daunorubicin (red circles). The apparent stiffness of a typical control cell remains relatively constant (green triangles) and does not undergo apoptosis or cell death during the course of the experiment. Error bars represent standard error. (B) From the same patient sample, the average apparent stiffness of a population of late apoptotic/dead AML cells was significantly stiffer than early apoptotic cells and controls (n = 15, P < .05). (C) Cell stiffness increases faster with 1 μM daunorubicin (DNR, in red) than 1 μM dexamethasone (DEX, in green). Solid and dotted lines represent myeloid and lymphoid leukemia cells, respectively. Transition from open to filled shapes represent onset of cell death (PI-positive staining). (D) Exposure to 2 μM cytochalasin D, an actin polymerization inhibitor, reduces stiffening behavior in HL60 cells exposed to 1 μM daunorubicin. The cells represented by these 3 lines were exposed to daunorubicin at time 0 minutes. The cell represented by the green line was also exposed to cytochalasin D at time 0 minutes (vertical green dashed line) and exhibited little stiffening behavior. The cell represented by the blue line was exposed to cytochalasin D after 45 minutes (vertical blue dashed line) and exhibited little stiffening behavior after exposure. As a positive control, the cell represented by the red line was not exposed to cytochalasin D. (E) HL60 and Jurkat cells were incubated with 1 μM daunorubicin and 2 μM cytochalasin D. The average stiffness of dead HL60 cells (n = 15) exposed to daunorubicin and cytochalasin D (green) was 0.2 kPa ± 0.05 kPa, whereas the average stiffness of dead HL60 cells exposed to daunorubicin alone (red) was 1.2 kPa ± 0.3 kPa (P < .05). Likewise, the average stiffness of dead Jurkat cells (n = 15) exposed to daunorubicin and cytochalasin D (green) was 0.1 kPa ± 0.03 kPa, whereas the average stiffness of dead Jurkat cells exposed to daunorubicin alone (red) was 0.5 kPa ± 0.14 kPa (P < .05).