Analysis of single colonies for mutations in TET2 and JAK2. Mononuclear cells from peripheral blood were grown in methylcellulose in the presence or absence of erythropoietin. Single burst-forming units erythroid (BFU-E), endogenous erythroid colonies (EEC), and colony-forming units granulocyte (CFU-G) were picked and analyzed individually for the presence of TET2 and JAK2-V617F mutations by DNA sequencing and allele-specific polymerase chain reaction (PCR), respectively. Each colony is represented by a dot that is placed into one of 6 quadrangles representing the 6 possible genotypes: wild-type (wt), heterozygous (het), and homozygous (hom) for JAK2-V617F on the vertical axis, and for TET2 mutations on the horizontal axis. The unique patient numbers, the diagnoses (PMF, ET, and PV) and the allelic ratio of JAK2-V617F in purified granulocytes (%T) are shown above the corresponding boxes. Light blue arrows indicate the suggested order of mutation events. (A) Patterns compatible with TET2 mutations occurring before JAK2-V617F. The sequencing chromatograms for patient p226 show the presence of TET2 mutation in DNA from hair follicles, T cells, and granulocytes, demonstrating the germline origin of the mutation. Allele-specific PCR assay for the X-chromosomal gene IDS is shown for p226. The genomic DNA from patient p226 was heterozygous for a C/T single nucleotide polymorphism (not shown). The relative expression of the 2 IDS alleles was determined by comparing the C and T peak intensities obtained by the allele-specific reverse-transcribed PCR assay in T cells and granulocytes. The skewing of expression toward the C-allele is shown for 10 individual colonies (gray area). The inactivated IDS allele is marked with an arrow. (B) Patterns compatible with JAK2 mutations occurring before TET2 was mutated. Patient p021 carries 2 JAK2 mutations, JAK2-V617F and JAK2 N542-E543del, but the TET2 mutation can be only found together with the deletion in exon 12 of JAK2. In patient p191, the sequencing chromatogram for the SNP rs34402524 located in TET2 is shown for one BFU-E colony marked in red. The presence of a heterozygous SNP excludes the possibility that this colony is the product of a mitotic recombination event. (C) Patterns compatible with a biclonal state of the disease. In patient p234, the sequencing chromatogram for the SNP rs6843141 located in TET2 is shown for one BFU-E and one CFU-G colony marked in red. Again, the presence of a heterozygous SNP excludes the possibility that these colonies are the product of a mitotic recombination event. (D) Location of mutations in the Tet2 protein in patients from whom data on single colonies are available. Mutations from this study are shown above the protein strand, and mutations analyzed in previous publications^3,26  are shown below. The gray boxes represent regions conserved between the different TET family members; blue boxes, TET2 mutations that occur before JAK2 mutations; yellow boxes, TET2 mutations that occurred after JAK2; white boxes, TET2 and JAK2 mutations compatible with biclonal disease; and red box, germline TET2 mutation.