The human bone marrow (BM) compartment consists of a heterogeneous, multi-functional network of cells and extracellular matrix that interact with hematopoietic stem cells. Growing evidence suggests that this microenvironment also likely adds to the resistance of chronic myeloid leukemia (CML) stem cells to tyrosine kinase inhibitor (TKI) therapy in vivo. Integrin-linked kinase (ILK) is a serine/threonine kinase that is an important constituent of focal adhesions, a regulator of spindle formation during mitosis, and a key mediator of multiple signaling pathways. However, its potential role in the regulation of primitive CML cells and their response to TKIs is unknown.

Our RNA-seq analysis revealed that ILK expression and its downstream targets, such as AKT, are highly upregulated in pre-treatment CD34+ CML stem/progenitor cells (n=6) compared to normal BM controls (n=3, p<0.05). We confirmed this finding by qRT-PCR analysis of CD34+ cells obtained from 28 CML patients (including some who subsequently became clinically resistant to imatinib mesylate (IM) treatment) compared to CD34+ BM cells from 9 normal adults (p<0.05). Elevated expression of ILK protein in CD34+ CML cells was also demonstrated by Western blot analysis (threefold, n=4, p<0.05). Interestingly, we also found ILK transcript levels to be much higher in the more primitive and rarer CD38- CML stem cell-enriched subset of the lin-CD34+ population as compared to the more mature and prevalent CD38+ subset or the terminally differentiated lin+CD34- cells (p<0.01). Both ILK transcripts and intracellular protein levels were elevated in CD34+ CML cells, but not in CD34- differentiated cells, in response to IM treatment in vitro, especially in the presence of expanded BM stromal cells, suggesting differential regulation of ILK expression in primitive CML cells and their more mature counterparts.

We next asked whether ILK influences primitive CML cell responses to TKIs by using a clinically validated and selective ILK inhibitor (QLT0267) to suppress ILK activity in colony assays of CD34+ CML cells. QLT0267 plus TKIs significantly reduced the yield of colonies obtained compared to any single agent or combination of TKIs and this enhanced cell killing was most pronounced on cells from IM non-responders (n=6; p<0.01). Analysis of the lineages affected showed that the combination of IM or dasatinib (DA) with QLT0267 had a more pronounced effect on myeloid colony formation (CFU-GM), at concentrations where either or both TKIs had little effect (i.e. 85-90% inhibition vs. 30-45%). Moreover, the simultaneous suppression of ILK and BCR-ABL activities also effectively inhibited the growth of more primitive CML cells (from IM non-responder patients) when these were co-cultured with stromal cells in 6-week long-term cultures (LTC), in contrast to the lack of these effects in the presence of single TKIs or TKI combinations.

Mechanistically, we found that the combination of QLT0267 with a TKI enhanced the induction of apoptosis of CD34+ CML cells in suspension cultures within the first three days (from 10-15% to 30-45% apoptotic cells relative to untreated controls) and CFSE (carboxy-fluorescein diacetate succinimidyl diester) tracking of cell division showed that QLT0267 specifically targets quiescent CML stem cells from IM-resistant patients (n=4; p<0.05). In addition, treatment with the ILK inhibitor alone or in combination with a TKI enhanced apoptosis of primitive CML cells in vitro by abolishing the protective effect of BM stromal cells observed under TKI treatment alone (25-30% apoptotic cells vs. 3-8%) and combination treatments confirmed strong synergy between DA and QLT0267 (CI value <0.3). Importantly, QLT0267 (up to 10 μM) was not toxic to normal CD34+BM cells in either short- or long-term culture systems (with and without stromal cells).

Together, these findings indicate that ILK is a critical player in the regulation of TKI response/resistance of primitive CML cells. The enhanced and selective effect obtained by dual inhibition of both targets, particularly in the presence of protective stromal cells to mimic their response within the BM microenvironment, may offer an important new therapeutic possibility.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution