Detection of the BCR-ABL^35INS mutation in clinical samples does not consistently track with resistance to ABL TKI therapy. Retrospective analysis of the detection of the BCR-ABL^35INS mutation by direct sequencing was performed in the chronologic context of the treatment and response for 20 CML patients at our institution. (A) Summary of patient response status at the time of BCR-ABL^35INS detection. Response status was categorized as follows: responding (improvement or maintenance of previous cytogenetic and molecular response level), intolerant (persistent grade 3 or higher toxicity leading to subsequent change of therapy), early resistance (confirmed increase in BCR-ABL transcripts and/or detection of a BCR-ABL point mutation without loss of level of cytogenetic response), or overt relapse (loss of previous cytogenetic response level). For comparison purposes, the approximate component level of the BCR-ABL^35INS mutation by direct sequencing was designated as minor (≤ 20% of total electropherogram signal) or major (> 20%). Although not strictly quantitative, this illustrates that the majority of patients with evidence of BCR-ABL^35INS by direct sequencing harbored the mutation at low levels that approached the sensitivity limits of this assay.¹⁴ Sequencing primers used by the OHSU Knight Diagnostic Laboratories have been described previously¹⁰; primers used by MolecularMD remain proprietary. ¶An additional intron 4 insertion was detected along with 35INS in patient 7; both represented a minor component of the total signal. *Patient 11 subsequently relapsed while undergoing therapy with dasatinib, with a predominant co-occurring T315I point mutation, whereas 35INS was detected intermittently and always as a minor component. †In patient 13, 35INS was detected intermittently as a minor component while the patient was taking dasatinib but notably was not detected at the time of subsequent cytogenetic relapse. §35INS was detected as a minor component in patient 16 during interruption of dasatinib treatment because of pregnancy; major molecular response was recaptured rapidly after treatment resumed. Representative examples are shown of timing and level of detection of the BCR-ABL^35INS mutation from each of 3 observed clinical treatment/response scenarios: (B) BCR-ABL^35INS was detected while the patient was responding to treatment without signs of resistance; (C) BCR-ABL^35INS was detected at the time of resistance, but with a co-occurring highly predominant BCR-ABL kinase domain point mutation that satisfactorily explained resistance; and (D) BCR-ABL^35INS was the only kinase domain mutation detected at the time of resistance to therapy. Notably, for all patients who showed evidence of the BCR-ABL^35INS mutation at the time of resistance (ie, scenarios C or D above), this mutation was always detected as a minor component by direct sequencing. For more detailed treatment, response, and sequencing information for all patients included in this analysis, see supplemental Table 1. CP indicates chronic phase; CCR, complete cytogenetic response; MMR, major molecular response; MCR, major cytogenetic response; and qPCR, quantitative PCR.

The BCR-ABL^35INS mutant does not confer growth factor independence or imatinib resistance, and ABL^35INS lacks kinase activity. (A) Ba/F3 cells expressing the indicated BCR-ABL construct were cultured in standard media supplemented with IL-3. After IL-3 washout and replating in standard media, cell viability was monitored daily for 4 days with the Guava ViaCount assay (Millipore). Results are shown as the mean viable cells ± SEM from 3 independent experiments. (B) Cell lysates were subjected to SDS-PAGE followed by immunoblot (IB) analysis with the indicated antibodies. Results are representative of 2 independent experiments. The long and short arrows on the right side of the immunoblot indicate the expected positions of BCR-ABL and BCR-ABL^35INS, respectively. (C) Ba/F3 cells stably expressing BCR-ABL were transduced with the indicated BCR-ABL^35INS mutant constructs, sorted for GFP, and cultured in standard media. Cells were plated in the presence of graded concentrations of imatinib (0-3125nM) and evaluated by methanethiosulfonate assay at 72 hours (CellTiter 96 AQ_ueous One Solution Cell Proliferation Assay; Promega) with a Synergy 2 plate reader (BioTek). Results are normalized to those of untreated cells and expressed as the mean ± SEM of 3 independent experiments performed in quadruplicate. (D) Cell lysates were subjected to SDS-PAGE followed by immunoblot (IB) analysis with the indicated antibodies. The long and short arrows on the right side of the immunoblot are provided for reference and indicate the expected positions of BCR-ABL and BCR-ABL^35INS, respectively. The absence of signal corresponding to BCR-ABL^35INS in the second and third panels signifies the lack of the ABL C-terminus and no phosphorylation on Y393, respectively. Lanes 1 and 2 represent nonadjacent lanes from the same gel (indicated by the vertical dashed line). (E) Equimolar amounts of purified, tyrosine-dephosphorylated GST-ABL kinase domain enzymes were incubated with [γ-³²P]-ATP. Samples collected at indicated times were subjected to SDS-PAGE, and signal intensity was measured by autoradiography. Comparable loading was confirmed by ABL immunoblot (IB) analysis with ABL antibody Ab-2 (Oncogene Science). Results are representative of 2 independent experiments. (F) Dephosphorylated GST-ABL enzymes were incubated with a biotinylated peptide substrate (biotin-GGEAIYAAPFKK-amide) and [γ-³²P]-ATP. At the end of the incubation period, reactions were terminated with 7M of guanidine hydrochloride solution, spotted in duplicate on SAM² Biotin Capture Membranes (Promega), washed according to the manufacturer's instructions, and subjected to liquid scintillation counting. GST-ABL reactions were quenched after 3.5 minutes, whereas GST-ABL^35INS and GST-ABL^K271R/35INS reactions were quenched after 1 hour because preliminary experiments at 3.5 minutes demonstrated no detectable activity. Results are shown as the mean enzymatic activity ± SEM normalized to that of GST-ABL. The autophosphorylation and peptide substrate phosphorylation activities of GST-ABL were inhibited by imatinib (data not shown). WT indicates wild-type.