A Subpopulation of Blasts with Attenuated p38^MAPK Response Is Seen in Virtually All AML Patients and AML Cell Lines and Is Defined By Cells with Augmented Lipid-Associated Anti-Oxidant Defense

The serine/ threonine kinase, p38^MAPK is activated by phosphorylation in response to a variety of cellular stresses including oxidative stress. Prolonged p38^MAPK activation drives cell-cycle arrest and apoptosis; and in HSC activation of p38^MAPK leads to a loss of reconstituting capacity (Ito et al, Nat.Med. 2006;12:446-451). In cancer, p38^MAPK responses are often attenuated and cancer models suggest that this is a necessary adaptation for transformation (Dolado et al, Cancer Cell 2007;11:191-205). Previously we have shown that 60% of acute myeloid leukemia (AML) patients constitutively generate significantly more extracellular reactive oxygen species (ROS) than normal hematopoietic CD34⁺ cells (Hole et al, Blood 2013;122:3322-3330). Despite this, AML blasts showed low or absent p38^MAPK phosphorylation; even in patients generating high levels of ROS. Here we examine p38^MAPK activation at the single cell level in primary AML blasts using flow cytometry.

We challenged AML blasts with a dose of hydrogen peroxide (H₂O₂) sufficient to completely activate p38^MAPK in normal CD34⁺ cells (1 mM for 30 min), where the threshold for activation was defined as the 95^th percentile of basal p38^MAPK activation in unstimulated cells. Attenuated responses to H₂O₂ were seen in 14/15 (93%) of patients; where 16-95% of the total blast population failed to activate p38^MAPK_. These non-responding cells are hereafter termed “Δpp38^MAPK cells” and were absent in normal CD34⁺ cells (p < 0.01; Figure 1). Examination of a panel of 6 AML cell lines showed that each of the lines contained Δpp38^MAPK cells at different frequencies: MV4-11 (10%); HL60 (10%); KG-1 (15%), U937 (30%), NB4 (50%), THP-1 (65%). Further analysis showed that Δpp38^MAPK cells were not distinguished by cell cycle phase, immunophenotype or reduced viability in either cell lines or AML blasts. These data suggest that nearly all AML patients harbor a population of blasts which have developed resistance to p38^MAPK activation. We reasoned that failure to respond could arise either through defective p38^MAPK signaling or because of enhanced anti-oxidative protection in a subpopulation of cells. To investigate the latter, we labelled cells with the lipophilic oxidation probe, C₁₁ -BODIPY or the cytosolic oxidant probe, CM-DCFDA and monitored the oxidative response to H₂O₂ at the single cell level in the AML cell lines: KG-1, MV4-11 and THP-1. In each case C₁₁ -BODIPY oxidation exactly matched the heterogeneous profile of p38^MAPK activation in these cells, whereas CM-DCFDA showed only homogeneous responses to H₂O₂ induction. These data show that Δpp38^MAPK cells are defined by an enhanced membrane-associated anti-oxidant capacity and we are currently analyzing this resistant subpopulation to identify the molecules responsible.

To examine whether p38^MAPK responsiveness influenced responsiveness to pro-oxidant drugs, we selected KG-1 and THP-1 cells (as representative examples of strong and weak p38^MAPK responses respectively) and tested their sensitivity to the pro-oxidant drugs, phenethyl isothiocyanate (PEITC) and buthionine sulfoximine (BSO). We found that the IC₅₀ was higher for THP-1 for both PEITC (KG-1 = 0.6µM; THP-1 = 7.5µM) and BSO (KG-1 = 50µM; THP-1 = 70µM), indicating that the p38^MAPK responsiveness limits the effectiveness of pro-oxidant drugs. We next examined whether promoting p38^MAPK activation could augment the effects of these pro-oxidants. We used the p38^MAPK activator 2-benzylidene-3-(cyclohexylamino)-1-indanone (BCI), which promotes activation of p38^MAPK via inhibition of a p38^MAPK phosphatase, DUSP1. This compound weakly promoted phosphorylation of p38^MAPK in THP-1 cells and consistent with this, had no effect on the efficacy of these compounds in these cells. However, BCI potently activated p38^MAPK in KG-1 cells and showed synergy with BSO in this context (CI = 0.3; Figure 2) indicating that where BCI is effective in activating p38^MAPK it can promote the effectiveness of pro-oxidant drugs.

In summary, we show for the first time that AML patients almost universally display attenuated p38^MAPK responses in all or part of the blast population and we suggest that this trait may be selected for to maintain self-renewing potential under the pro-oxidative conditions found in the leukemic marrow. Further we show that by manipulating p38^MAPK activity, we can augment the potency of the pro-oxidant compound BSO.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal

Figure 2

View largeDownload slide

Figure 2

View largeDownload slide

Close modal