Analysis of acquired mutations in transgenes arising in Ba/F3 transformation assays: findings and recommendations

Abstract

The identification and functional validation of potentially oncogenic mutations in leukemia is an essential step toward a future of personalized targeted therapy. To assess the oncogenic capacity of individual mutations, reliable and scalable in vitro experimental approaches are required. Since 1988, researchers have used the IL-3 dependent Ba/F3 transformation assay to validate the oncogenic potential of mutations to drive factor-independent growth. Here we report a previously unrecognized phenomenon whereby Ba/F3 cells, engineered to express weakly transforming mutations, present with additional acquired mutations in the expressed transgene following factor withdrawal. Using four mutations with known transformative capacity in three cytokine receptors (CSF2RB, CSF3R and IL7R), we demonstrate that the mutated receptors are highly susceptible to acquiring additional mutations. These acquired mutations of unknown functional significance are selected by factor withdrawal but appear to exist prior to the removal of growth factor. This anomaly has the potential to confound efforts to both validate and characterize oncogenic mutations in leukemia, particularly when it is not standard practice to sequence validate cDNAs from transformed Ba/F3 lines. We present specific recommendations to detect and mitigate this phenomenon in future research using Ba/F3 transformation assays, along with methods to make the Ba/F3 assay more quantitative.

Keywords: Ba/F3 transformation assay; functional validation; leukemia; oncogenes; reproducibility in research.

Figure 1. Experimental design schematic

Ba/F3 cells are infected using freshly harvested retrovirus which drives the expression of GFP and the transgene of interest. Infected cells are selected for GFP-expression on day 2, and the first set of limiting dilution plates are created. On day 9 a second set of limiting dilution plates are started, and on day 10 the cells are triple-washed then monitored for IL-3 independent growth over 21 days in two technically replicate flasks. Genomic DNA (gDNA) is harvested from cell lines in IL-3 on day 10 (baseline) and day 31 (cultured). As cells proliferate in the absence of IL-3, gDNA is harvested to detect acquired mutations. This workflow was repeated for every construct to create a minimum of three biologically replicate lines.

Figure 2. Compiled data from all IL-3 withdrawal experiments

A. Outgrowth curves for every replicate in this study, separated by transgene and mutation. The outgrown vector empty line was sequenced and found to be contaminated with IL7R Ins243PPCL. B. Time to outgrowth can be summarized as the number of days to reach a viable cell count 5-times the number of cells initially seeded in each flask at the start of the experiment. Lines that did not grow within the 21-day period are shown above the break. C. Mutations detected in outgrown lines after transformation to factor-independent growth. Every transformed line for CSF2RB WT, R461C and CSF3R WT, W791X are shown, including lines that did not acquire additional mutations (*).

Figure 3. Ba/F3 transformation rates vary by transgene but not by time in culture

A. Transformation rates calculated for each biologically replicate cell line are shown for plates started 2 days and 9 days after retroviral infection. The transformation rate is expressed as 1 in X cells capable of transforming to IL-3 independent growth, therefore a higher transformation rate indicates a weakly transforming cell line. No consistent trend is observed between days post-infection and transformation rate. B. Transformation rate and 95% confidence intervals for every replicate. The median rates across biologically replicate samples are shown below, along with the frequency of transforming cells (inverse of transformation rate). Lines that exhibited no observable transformation are shown with a rate of 1.0×10⁷.

Figure 4. Acquired mutations occur in weak, but not strong, transforming transgenes

A. Cell lines are ordered by transformation rate and colored based on mutational status as observed from sanger sequencing of bulk outgrowth assays of the same lines. B. Cell lines are ordered by days to outgrowth (time to reach a 5-times increase in viable cells over the starting cell number) and colored based on mutational status. BCR-ABL was not fully sequenced due to length and structural complexity.