Detection of Proteins Related to Therapeutic Outcome, Including Drug Resistance, in Acute Myeloid Leukemia Using Mass Spectrometry and Gel Based Proteomic Profiling.

To further improve and individualize curative treatment approaches for patients with AML, novel proteins and protein functions need to be discovered. In particular, the mechanisms responsible for chemotherapy resistance appear essential to unravel. Clinical proteomics may help to provide such information. We have used a chip-based technique, surface enhanced laser desorption/ionization mass spectrometry (SELDI-MS), to compare protein expression profiles in AML. With this technique proteins in the molecular weight area of 2–40 kDa can be spotted. For selected samples, a two-dimensional electrophoresis (2DE) method was employed, allowing detection of proteins between 15–120kDa. In the first part of the study, mononuclear peripheral blood cells were obtained from 8 AML patients clinically sensitive to and 7 patients clinically resistant to doxorubicin-containing induction chemotherapy. The clinical outcome correlated to results from parallel in vitro drug testing using the FMCA method. In addition, two variants of the human leukemic cell line CEM, one sensitive and one resistant to teniposide (CEM/VMI), were analysed. In the second part of the study, global proteomic profiles from 24 selected AML patients with good vs. poor clinical outcome (remission duration) were investigated. A total of 812 proteins were detected in the mononuclear cells obtained from the 15 doxorubicin-treated patients. Twenty of these proteins were distinguished as different between the sensitive vs. resistant patient groups. Assessing the CEM cell lines by utilizing four different affinity surfaces on ProteinChip™ arrays, approximately 450 peptides and proteins were detected. Between 29 and up to 65 proteins were detected as differentially expressed between the sensitive and the resistant CEM variants. Using the 2DE technology, approximately 1000 spots were matched on the CEM cell lines, with 47 clear expression level differences between the sub cell lines. In clustering analysis using the protein profile data we could demonstrate clear clustering of sensitive and resistant patient and cell line cells according to sensitivity measurements (IC50). Using these methods we have selected a panel of putative resistance related proteins, which are currently being identified and validated. Data from the second phase of the study, derived from 24 selected AML cell populations, are presently being analyzed. In conclusion, significant differences in protein profile expression were observed between cells obtained from AML patients with clinically divergent treatment responses. Further studies aim at identifying key proteins and protein functions to allow for improved therapeutic efficacy.