Activating PAX gene family paralogs to complement PAX5 leukemia driver mutations

Two separate transfections of PAX2, PAX5, PAX8, or empty vector (pRRL-CMV-IRES-hrGFPII) control were performed via electroporation in 400μL OPTI-MEM on 1.5×106 Reh cells per treatment condition (BioRad GenePulser Xcell, Square wave, 210V, 15ms, 2x pulse, 0.1sec gap). The transfected cells were incubated for 24 hours post transfection in normal growth media (see above). Cells were washed twice in PBS plus 1% FBS, and 4–6×105 GFP (+) cells per treatment condition were harvested via FACS for RNA. Total RNA was isolated from sorted cells using the Qiagen RNeasy Plus kit.

1×106 Reh cells, cultured in RPMI supplemented with 10% FBS, were exposed to either K-gluconate or CaCl2 (80mM concentration added to media) for 24 hours. The cells were then washed twice with PBS, and total RNA was isolated using the Qiagen RNeasy Plus kit. Each condition, along with a no treatment control, was repeated for a total of 3 samples.

mRNA libraries were prepared from total RNA (~1 μg per sample) using the Illumina TruSeq Stranded mRNA Library Prep kit. Individual RNA samples were created by pooling equal amounts of total RNA from multiple experiments (see above). Prepared libraries were sequenced on an Illumina NextSeq 500 instrument (1×75 bp read). Reads were aligned against GRCh38 using the HISAT2 aligner. Reads mapped to exons defined by GRCh38.88 were counted, and transcripts per million (TPM) values for each gene were calculated using StringTie software (Johns Hopkins Center for Computational Biology). An average of 53 million reads were aligned per sample.

Bar graphs represent combined mean values of 2-ΔΔCT for included experimental replicates, relative to controls indicated in the figure legend and graph axis labels. Error bars represent standard deviation. Significance was determined using one-way t-test method for deviation from a fixed value (i.e., value of control sample). p-values * <0.05, ** <0.005, *** <0.0005.

Bar graphs for antibodies and cell size (FSC-A) represent mean fluorescence intensity. Values are averaged across several experimental replicates, as indicated, above. Error bars represent standard deviation. Significance was determined using one-way t-test method for deviation from a fixed value (i.e., normalized value of control sample). p-values * <0.05, ** <0.005, *** <0.0005. Bar graphs for cell cycle phase (S5 Fig) were determined from percentages of cells in G1, S, and G2 phase, based on DAPI staining, and assessed by the “Cell cycle” function in FlowJo, vX.

Normalization and differential expression calculations were performed using the R package DESeq2 [73] based on TPM data. Clustering and heatmap creation were performed using the heatmap.2 package (dist = Euclidean and method = complete).

Expressed genes in each sample were ranked based on their log2 fold change in mRNA levels when compared to the appropriate control. GSEA was conducted using GSEA Desktop 3.0 software (Broad Institute). Gene sets analyzed (Molecular Signatures Database v6.1) include the biological process group from Gene Ontology (GO:BP), the transcription factor targets (TFT) group, and a custom PAX5 related group of human genes based on genes differentially expressed at various stages of B cell development in mice that have either normal levels of PAX5 or are deficient [13,43]. Briefly, the pro and mature B cell sets are comprised of genes that are differentially expressed compared to the appropriate controls and have predicted PAX5 binding sites in their promoter region while the pro-to-mature B cell sets contain all genes differentially expressed when comparing mature B cells to pro-B cells in the presence or absence of PAX5. Analysis was conducted using the GSEAPreranked tool to calculate a classic enrichment score for each set. Gene sets with a false discovery q-value (FDR) of <0.05 were selected for further analysis.

We thank members of the Horwitz and Miller labs and Dr. David Wu for their suggestions and support. We thank Dr. Karol Bomsztyk for suggesting use of mannitol. TN and MSH wish to dedicate this paper to the memory of Dr. George Stamatoyannopoulos.

All RNA-seq raw data files are available from the Gene Expression Omnibus (GEO) database (accession number GSE109860), https://www.ncbi.nlm.nih.gov/geo/.

MRH was supported by the University of Washington, Division of Hematology Training Grant, National Institutes of Health (https://www.nih.gov/) T32HL007093. MSH and ML gratefully acknowledge the support of the Allen Discovery Center program through the Paul G. Allen Frontiers Group (https://www.alleninstitute.org/what-we-do/frontiers-group/). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

All RNA-seq raw data files are available from the Gene Expression Omnibus (GEO) database (accession number GSE109860), https://www.ncbi.nlm.nih.gov/geo/.