Pharmacological Inhibition of Nemo-like Kinase Rescues mTOR-Mediated Translation and Primes Progenitors for Leucine-Stimulated Erythroid Expansion in Pre-Clinical Models of Diamond Blackfan Anemia

Diamond Blackfan Anemia (DBA) is associated with anemia, congenital abnormalities, and cancer. Current therapies for DBA have undesirable side effects, including iron overload from repeated red cell transfusions or infections from immunosuppressive drugs and hematopoietic stem cell transplantation. Human hematopoietic stem and progenitor cells (HSPCs) from cord blood were transduced with lentiviral shRNA against a number of ribosomal genes associated with DBA, reducing the specific ribosomal protein expression by approximately 50%. During differentiation, these cells demonstrated a DBA-like phenotype with significantly reduced differentiation of erythroid progenitors (over 80%), yet only modest (15-30%) reduction of other hematopoietic lineages. NLK was immunopurifed from differentiating HSPCs and activity was assessed by the extent of in vitro phosphorylation of 3 known NLK substrates NLK, c-Myb and Raptor. As NLK activation requires phosphorylation at Thr298, we also showed that in vitro activity correlated with intracellular NLK phosphorylation by Western blot analysis.

Nemo-like Kinase (NLK) was hyperactivated in the erythroid progenitors (but not other lineages), irrespective of the type of ribosomal gene insufficiency. We extended these studies using other sources of HSPCs (fetal liver, whole blood and bone marrow), along with RPS19- and RPL11-insufficient mouse models of the disease, as well as DBA patient samples. NLK was hyperactivated in erythroid progenitors from mice (5.3- and 7.2-fold increase in Raptor phosphorylation in RPS19- and RPL-11 insufficiency respectively) and from humans (7.3- and 9.0-fold in RPS19- and RPL11-insufficiency respectively) as well as HSPCs from three DBA patient (4.8-, 4.1- and 4.2-fold increase above controls). In RPS19-insufficient human HSPCs, genetic silencing of NLK increased erythroid expansion by 2.2-fold (p=0.0065), indicating that aberrant NLK activation contributes to disease pathogenesis.

Furthermore, a high-throughput inhibitor screen identified a compound that inhibits NLK (IC50:440nM) and increases erythroid expansion in murine (5.4-fold) and human (6.3-fold) models of DBA without effects on normal erythropoiesis (EC50: 0.7 µM). Identical results were observed in bone marrow CD34+ progenitors from three DBA patients with a 2.3 (p=0.0009), 1.9 (p=0.0007) and 2.1-fold (p=0.0001) increase in CD235+ erythroid progenitor population following NLK inhibition.

In erythroid progenitors, RPS19-insufficiency increased phosphorylation of the mTORC1 component Raptor, reducing mTOR in vitro activity by 82%. This was restored close to basal levels (93.8% of healthy control) upon inhibition of NLK. To compensate for a reduction in ribosomes, stimulating mTOR activity with leucine has been proposed to increase translational efficiency in DBA patients. In early clinical trials, not all DBA patients have responded to leucine therapy. We hypothesize that one of the reasons might be due to NLK phosphorylation of Raptor. While leucine treatment increased mTOR activity in both RPS19-insufficient and control cells (164% of healthy controls: p=0.007 and 24% to 42% of healthy controls: p=0.0064), combining leucine with NLK inhibition increased mTOR activity in RPS19-insufficiency from 24% to 142% of control (p=0.0012). This translated to improvements in erythroid expansion of RPS19-insufficient HSPCs from 8.4% to 16.3% with leucine treatment alone, 28.4% with NLK inhibition alone, but 68.6% when leucine and NLK inhibition were combined. This 8.2-fold improvement in erythroid progenitor production indicates that identification of aberrantly activated enzymes, such as NLK, offer therapeutic promise used alone, or in combination with existing therapies, as druggable targets in the clinical management of DBA.