Page 113 - 2019_07 resto del Mondo-web

SP enriches for LPC and Wnt expression in zebrafish ALL
experiments, there was a very significant increase in LPC frequency in the SP groups as compared to the non-SP groups (Figure 3B). These experiments showed both an enrichment of LPC activity in the SP groups and a deple- tion within the non-SP groups, including one tumor whose non-SP did not have engraftment of a single LPC. These results, in combination with the previous correla- tive data, strongly demonstrate that the SP in this zebrafish ALL model is enriched for LPC.
Zebrafish acute lymphoblastic leukemia side population cells have a common expression profile with increased expression of stem cell- and Wnt pathway-associated genes
Once we had validated the SP assay as a method to enrich for LPC in our model, we did bulk RNA sequenc- ing on sorted SP and non-SP cells from three distinct tumors (Figure 4). This produced 272 genes that were dif- ferentially expressed in the SP of each individual tumor (Figure 4A and Online Supplementary Table S4) and 761 genes whose mean expression across all three tumors (Figure 4B and Online Supplementary Table S5) was differ- entially expressed between SP and non-SP cells. Overall,
AB
the common expression profile showed a bias towards genes with increased expression in the SP, without a strong shared pattern of expression in the non-SP. This is not surprising as despite having been derived from genet- ically identical fish, individual tumors in this model are quite heterogeneous and have varied progression and functional parameters. PANTHER pathway analysis34 of the two differentially expressed gene sets identified Wnt signaling as the most overrepresented pathway (Figure 4C and Online Supplementary Table S6). The Wnt pathway has been implicated in regulating stem cells in normal and malignant tissues,35 shown to be necessary for self-renew- al of LSC in acute myelogenous leukemia,36 and also been proposed to play a similar role in ALL.37,38
We further analyzed the entire RNA-sequencing dataset with gene set enrichment analysis (GSEA) using the Gene Ontology: Biological Processes gene set. This analysis identified 464 significantly upregulated path- ways (Online Supplementary Table S7) including canonical Wnt signaling, and stem cell differentiation as the highest ranked gene set (Figure 5A). The majority of the 185 sig- nificantly downregulated pathways related to metabolic processes, including cell cycle transitions, protein transla-
C
Figure 2. Side population size undergoes clonal evolution in tan- dem with leukemia-propagating cell frequency. (A) Graphic depicting the workflow of serial limiting-dilution transplants. (B) Representative lineage of a primary tumor with leukemia-propa- gating cell (LPC) frequency, percent side population (SP), and latency to >50% tumor burden. Dpi: days post-injection; dpt: days post-transplant. The statistical significance of the increase in LPC% from the parent tumor was determined by pairwise tests using Extreme Limiting Dilution Analysis software *P<0.01. (C) Pooled data from multiple lineages showing percent SP (left y- axis, red) and LPC frequency (right y-axis, blue) (n=14).
haematologica | 2019; 104(7)
1391