Use of Ubiquitin-Proteasome System Profiling for Differentiating Between Various Leukemic Processes.

Abstract

Abstract 4712The ubiquitin-proteasome system (UPS) plays a major role in cell homeostasis in normal and neoplastic states. Expression and function of the UPS system vary with the specific characteristics of individual cell types, suggesting that determination of UPS “signatures” could be useful in identifying various cell populations. Since direct analysis of cancer cells is often problematic, even in hematologic diseases, we explored the potential of using UPS signatures in plasma to differentiate between various leukemias. We first analyzed plasma UPS profiles of patients with acute myeloid leukemia (AML; n=111), acute lymphoblastic leukemia (ALL; n=29), advanced myelodysplastic syndrome (MDS; n=20), chronic lymphocytic leukemia (CLL; n=118), or chronic myeloid leukemia (CML; n=128; 46 in accelerated/blast crisis [ACC/BL], 82 in chronic phase), and 85 healthy control subjects. Plasma levels of proteasome, ubiquitin (poly-ubiquitin), and the 3 proteasome enzymatic activities (chymotrypsin-like [Ch-L], caspase-like [Cas-L], trypsin-like [Tr-L]) were measured. Specific activities were calculated by normalizing each of the 3 enzyme activities to the levels of proteasome protein in plasma (Ch-L/p, Cas-L/p, and Tr-L/p). These 8 variables were used in multivariate logistic regression models to differentiate between leukemic processes. UPS signatures provided clear differentiation between patients with a leukemic process and normal controls (AUC=0.991), using 6 different variables (Tr-L/P, Ch-L, Ch-L/p, Cas-L, Cas-L/P, ubiquitin). Distinguishing between acute (AML, ALL, MDS) and chronic (CML, CLL) processes was less efficient (AUC=0.853 using Tr-L, Tr-L/P, Cas-L/P, Ch-L/P, proteasome, Ch-L), likely due to the high proportion (36%) of CML patients in ACC/BL phase. However, UPS signatures generally yielded powerful differentiation between individual leukemias (Table). MDS was not well differentiated from AML (AUC=0.791), reflecting the significant biological overlap of these diseases. These data support the potential usefulness of the UPS profile to aid in the differential diagnosis of various leukemias.ComparisonAnalytesN (Case/Control)AUCMDS vsNormalPS, UBT, C-L, Tr-L, C-LP Ch-P105 (20/85)0.98ALLPS, UBT, Tr-L, C-LP49 (20/29)0.9845AMLPS, UBT, Ch-L, Tr-L, C-LP, Ch-LP, Tr-L/P131 (20/111)0.791CLLC-L/P, Ch-LP163 (27/136)0.9909CMLUBT, C-LP, Tr-L147 (20/127)0.9496CML (Chronic)UBT, C-LP, Tr-L, Tr-L/P101 (20/81)0.9608CML (ACC/BL)UBT, C-LP, Tr-L, Tr-LP, Ch-L66 (20/46)0.9516AML vsNormalUBT, C-L, Tr-L, Tr-LP, Ch-LP196 (111/85)0.9884ALLC-LP, Ch-LP, UBT, Ch-LP, PS140 (111/29)0.9503CLLC-LP, Ch-LP283 (147/136)0.9835CMLTr-LP, Tr-L, UBT, C-LP, C-L, Ch-LP238 (111/127)0.8301CML (Chronic)Tr-LP, Tr-L, UBT, C-LP, Ch-LP, PS192 (111/81)0.8396CML (ACC/BL)Tr-LP, Tr-L, UBT, C-LP, Ch-LP, C-L157 (111/46)0.8239ALL vsNormalUBT, Tr-L114 (29/85)0.9984CLLUBT, Tr-L, Tr-LP, Ch-L, Ch-LP, C-LP147 (29/118)0.9031CMLno Tr-LP156 (29/127)0.9005CML (Chronic)no Tr-L110 (29/81)0.9264CML (ACC/BL)no Tr-L75 (29/46)0.8909CLL vsNormalC-L/P, Ch-L, UBT203 (118/85)0.9978CMLC-LP, C-L, Ch-LP, Tr-L, UBT245 (118/127)0.9746CML (Chronic)Cas-LP, C-L, Ch-LP, Tr-L UBT199 (118/81)0.9749CML (ACC/BL)C-LP, C-L, Ch-LP, Tr-L, UBT164 (118/46)0.9746CML vs NormalUBT, Tr-L212 (127/85)0.9986Abbreviation: PS=proteasome protein Disclosures:No relevant conflicts of interest to declare.