Tumor cells express the glycolytic regulator pyruvate kinase subtype M2 (M2-PK), which can occur in a tetrameric form with high affinity to its substrate phosphoenolpyruvate (PEP) and a dimeric form with a low PEP affinity. The transition between both conformations contributes to the control of glycolysis and is important for tumor cell proliferation and survival. Here we targeted M2-PK by synthetic peptide aptamers, which specifically bind to M2-PK and shift the isoenzyme into its low affinity dimeric conformation. The aptamer-induced dimerization and inactivation of M2-PK led to a significant decrease in the PK mass-action ratio as well as ATP:ADP ratio in the target cells. Furthermore, the expression of M2-PK-binding peptide aptamers moderately reduced the growth of immortalized NIH3T3 cell populations by decelerating cell proliferation, but without affecting apoptotic cell death. Moreover, the M2-PK-binding peptide aptamers also reduced the proliferation rate of human U-2 OS osteosarcoma cells. In the present study, we developed the first specific inhibitors of the pyruvate kinase isoenzyme type M2 and present evidence that these inhibitors moderately decelerate tumor cell proliferation. ' 2008 Wiley-Liss, Inc.Key words: tumor metabolism; peptide aptamer; proliferation; pyruvate kinase Tumorigenesis is characterized by an increase in glycolytic enzyme activities as well as distinct changes in the glycolytic isoenzyme equipment. 1-3 A key glycolytic enzyme which is consistently altered in tumor cells is pyruvate kinase (EC 2.7.1.40), catalyzing the dephosphorylation of phosphoenolpyruvate (PEP) to pyruvate, which significantly contributes to net adenosine triphosphate (ATP) production within the glycolytic pathway. Depending on the metabolic functions of various tissues, different isoenzymes of pyruvate kinase are expressed. 4 The pyruvate kinase isoenzyme M1 (M1-PK) is characteristic for tissues with high energy turnover (muscle and brain), whereas the pyruvate kinase isoform L (L-PK) is found in tissues with gluconeogenesis, such as liver. When quiescent cells re-enter the cell cycle, as during tumorigenesis, the tissue-specific isoenzymes disappear and the pyruvate kinase isoenzyme type M2 (M2-PK) is expressed. 3,4 According to the different metabolic functions, M1-PK has the highest affinity to its substrate PEP, whereas the PEP affinity of M2-PK depends on its quaternary structure. M2-PK occurs as a tetrameric form characterized by a high affinity to its substrate PEP which is highly active at physiological PEP concentrations and as a dimeric form with low PEP affinity which is nearly inactive under physiological conditions. 3 The tetrameric but not the dimeric form of M2-PK is associated with other glycolytic enzymes as well as with lactate dehydrogenase (LDH), nucleotide diphosphate kinase and adenylate kinase, within a large complex, referred to as the glycolytic enzyme complex. 5-7 A high amount of the tetrameric form of M2-PK correlates with a relative high PK mass-action ratio ([pyruvate] AE[ATP])...

Pyruvate kinase M2 (M2-PK) controls the rate-limiting step at the end of the glycolytic pathway in normal proliferating and tumor cells. Other functions of M2-PK in addition to its role in glycolysis are little understood. The aim of this study was to identify new cellular interaction partners of M2-PK in order to discover novel links between M2-PK and cellular functions. Here we show that the SUMO-E3 ligase protein PIAS3 (inhibitor of activated STAT3) physically interacts with M2-PK and its isoenzyme M1-PK. Moreover, we demonstrate that endogenous SUMO-1-M2-PK conjugates exist in mammalian cells. Furthermore, we show that transient expression of PIAS3 but not the RING domain mutant PIAS3 (C299S, H301A) is consistent with nuclear localization of M2-PK and PIAS3 and M2-PK partially co-localize in the nucleus of these cells. This study suggests a link between PIAS3 and nuclear pyruvate kinase.

Insulin-like growth factor binding protein-3 (IGFBP-3), the product of a tumor suppressor target gene, can modulate cell proliferation and apoptosis by IGF-I-dependent and IGF-Iindependent mechanisms. IGFBP-3 controls the bioavailability of IGFs in the extracellular environment and is known to be subject to degradation by various extracellular proteases. Although nuclear localization and functions of IGFBP-3 have been described in the past, we show as the novel features of this study that the abundance of nuclear IGFBP-3 is directly regulated by ubiquitin/proteasome-dependent proteolysis. We show that IGFBP-3 degradation depends on an active ubiquitin-E1 ligase, specific 26S proteasome inhibitors can efficiently stabilize nuclear IGFBP-3, and the metabolic halflife of nuclear IGFBP-3 is strongly reduced relative to cytoplasmic IGFBP-3. Nuclear IGFBP-3 is highly polyubiquitinated at multiple lysine residues in its conserved COOHterminal domain and stabilized through mutation of two COOH-terminal lysine residues. Moreover, we show that IGFBP-3, if ectopically expressed in the nucleus, can induce apoptotic cell death. These results suggest that ubiquitin/proteasomemediated proteolysis of IGFBP-3 may contribute to downregulation of apoptosis. (Cancer Res 2006; 66(6): 3024-33)