All ETDs from UAB

Advisory Committee Chair

Robert S Welner

Advisory Committee Members

Ravi Bhatia

Christopher A Klug

Stuart J Frank

Adam Wende

Document Type


Date of Award


Degree Name by School

Doctor of Philosophy (PhD) Heersink School of Medicine


Drug persistence is one of the major impediments for cancer therapeutics. There are two known evolutionary models for tumor drug persistence. First, a pre-existing sub-population is inherently resistant to treatment; and second, a sub-population acquires resistance by drug-induced molecular reprograming. One ideal model to study the mechanism of drug persistence is chronic myeloid leukemia (CML), a hematopoietic stem cell (HSC) disorder. HSCs acquire a single translocation between chromosome 9 and 22, forming BCR-ABL, a constitutively active tyrosine kinase. Tyrosine kinase inhibitors (TKI), the standard regimen, have been able to efficaciously control CML and yet many patients relapse upon treatment discontinuation due to persistence of leukemic stem cells (LSCs). This is attributed to the fact that LSCs are heterogeneous. Using single-cell RNA-seq, we and others have shown that CML LSCs have a pre-existent TKI-persistent transcriptional signature. We also show that stem and progenitor cells from CML and TKI-treated CML mice reveals a unique transcriptional signature with upregulation of genes regulated by STAT3. In fact, we and others have reported a BCR-ABL independent STAT3 activation in TKI-persistent CML cells. Inhibition of STAT3 using a small molecule inhibitor, induces apoptosis of TKI-persistent CML cells, but has minimal effect on TKI-sensitive cells. Drug withdrawal from TKI-persistent cells re-sensitizes them to TKI with reduced susceptibility to STAT3 inhibition. Therefore, STAT3 activation seems to confer survival advantage to TKI-persistent cells. Interestingly, ChIP-seq on these stem and progenitor cells revealed localization of phosphorylated STAT3 to unique sites upon TKI treatment, many of which associated with genes regulating metabolic pathways. Subsequent metabolic analysis revealed that unlike TKI-sensitive cells, the TKI-persistent CML cells utilize glycolysis as a source of energy over oxidative phosphorylation. This was accompanied by dysregulated mitochondrial function in the TKI-persistent cells as demonstrated by reduced electron complex chain protein expression as well as activity. Targeting one of the rate-limiting enzymes for glycolysis, PKM2, which was also found to be upregulated in our proteomic screen, specifically eradicated the TKI persistent CML cells. Exploring the mechanism of metabolic and transcriptomic rewiring by STAT3 will provide insight into finding better targets for oncogene independent TKI-persistent leukemic stem cells, the goal of this study.