There is an ongoing controversy as to whether cancer is always maintained by a rare population of highly specialized cancer stem cells or whether cancer-propagating cells may be more abundant in some cancer types. We have previously shown that in a heterogeneous group of childhood ALL different blast populations, regardless of their expression of the progenitor/stem cell marker CD34 or the lymphoid differentiation antigen CD19, contain leukemia-initiating activity (Cancer Cell 2008, 14(1), 47–58). By profiling B cell transcription factor expression, these different populations appeared to mirror stages of normal B cell development. Here we extend our experiments to another lymphoid differentiation marker, CD20, to provide further evidence that ALL blasts at different stages of maturation possess the ability to re-initiate the leukemia.

PatientTransplantMicePopulationCell doseEngrafted
L736 Secondary CD20High 10.000–100.000 
  11 CD20−/Low  10 
 Tertiary CD20High 10.000 
  CD20−/Low  
L754 Primary CD20High 100.000 
  CD20−/Low  
 Secondary 11 CD20High 5.000-100.000 
  11 CD20−/Low  
A67 Secondary CD20High 9.000-20.000 
  CD20−/Low  
PatientTransplantMicePopulationCell doseEngrafted
L736 Secondary CD20High 10.000–100.000 
  11 CD20−/Low  10 
 Tertiary CD20High 10.000 
  CD20−/Low  
L754 Primary CD20High 100.000 
  CD20−/Low  
 Secondary 11 CD20High 5.000-100.000 
  11 CD20−/Low  
A67 Secondary CD20High 9.000-20.000 
  CD20−/Low  

Unsorted bone marrow cells from 3 different ALL patients (L736, L754 and A67) were transplanted into 12 primary mice. Bone marrow was harvested from leukemic mice and flow sorted candidate populations (CD19+CD20−/Low and CD19+CD20High) were re-transplanted into 52 secondary and 8 tertiary mice. As expected from our previous experiments both CD19+CD20−/Low and CD19+CD20High cells were able to re-establish the disease in unconditioned NOD/scid y−/− mice (see table). Leukemic engraftment ranged from 0.5 to 73% as determined by flow cytometry on bone marrow aspirates. Both populations re-established the complete phenotype of the original leukemia including CD20−/Low and CD20High blasts. These results were confirmed by directly sorting primary ALL blasts from L754 without prior passage in the mice. Cell purity after flow sorting was high (81–99%) and low numbers of cells engrafted (5000 for both CD19+CD20−/Low and CD19+CD20High). The three patients reflected different ALL subtypes (L736 intermediate risk ALL: high WBC/MRD low risk, L754:high hyperdiploid/MRD high risk; and A67: high risk ALL/t(9;22)).

In conclusion, these results confirm our previous observation that ALL blasts irrespective of the expression of lymphoid differentiation markers are able to engraft immune-deficient mice. Therefore, leukemia-propagating cells in childhood ALL may be more abundant than previously thought.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author

Sign in via your Institution