Suppression of ADAM17-mediated Lyn/Akt Pathways Induces Apoptosis of Human Leukemia U937 Cells

Cell surface proteases have been demonstrated to play an important role in facilitating cell invasion into the extracellular matrix and may contribute significantly to extracellular matrix degradation by metastatic cancer cells. Abundant expression of these enzymes is associated with poor prognosis. Thus, protease inhibitors that repress cell surface proteases may be applicable to cancer therapy. Because soybean Kunitz-type trypsin inhibitor has been found to induce apoptotic death of human leukemia Jurkat cells, anti-leukemia activity of Bungarus multicinctus protease inhibitor-like protein-1 (PILP-1) is thus examined. PILP-1 induced apoptosis of human leukemia U937 cells, characteristic of loss of mitochondrial membrane potential, degradation of procaspase-8, and production of t-Bid. FADD down-regulation neither restored viability of PILP-1-treated cells nor attenuated production of active caspase-8 and t-Bid in PILP-1-treated cells, suggesting that the death receptor-mediated pathway was not involved in the cytotoxicity of PILP-1. It was found that PILP-1-evoked p38 MAPK activation and ERK inactivation led to PILP-1-induced cell death and down-regulation of ADAM17. Knockdown of ADAM17 by siRNA induced death of U937 cells and inactivation of Lyn and Akt. Immunoprecipitation suggested that ADAM17 and Lyn form complexes. Overexpression of ADAM17, LynY507F (gain of function), and constitutively active Akt suppressed the cytotoxic effects of PILP-1. PILP-1-elicited inactivation of Lyn and Akt was abrogated in cells with overexpressed ADAM17 or LynY507F. Taken together, our data indicate that ADAM17-mediated activation of Lyn/Akt maintains the viability of U937 cells and that suppression of the pathway is responsible for PILP-1-induced apoptosis.

Protease inhibitors play a critical role in the regulation of several biological processes such as blood coagulation, complement fixation, fibrinolysis, fertilization, and embryogenesis (1). Dysregulation of proteinases leads to many pathophysiological conditions that include cancer, atherosclerosis, and inflammation. In particular, cell surface proteases, including meprin, matrix metalloproteinase, dipeptidyl peptidase IV, and seprase, have been demonstrated to play an important role in facilitating cell invasion into extracellular matrix and may contribute significantly to extracellular matrix degradation by metastatic cancer cells (2). Abundant expression of these enzymes is associated with poor prognosis. Thus, protease inhibitors that repress cell surface proteases may be applicable to cancer therapy. Protease inhibitors are grouped into a number of families, including the Kunitz, Kazal, Serpin, and mucous families (3). Several Kunitz-type protease inhibitors, including bikunin, hepatocyte growth factor activator inhibitor-2, and tissue factor pathway inhibitor-2, are found to suppress tumor invasion and metastasis (4 -7). It was suggested that bikunin and tissue factor pathway inhibitor-2 exerted their biology activities through a cell surface receptor-mediated process (3,4). Moreover, tissue factor pathway inhibitor-2 elicits pro-apoptotic signaling pathway in the human fibrosarcoma cell line (8).
Snake venoms are complex mixtures of pharmacologically active polypeptide toxins that are believed to have evolved to alter functionally the physiological activities along with predator-prey interaction (9 -11). In addition to enzymes and toxins, snake venom also contains serine protease inhibitors. Several Kunitz/bovine pancreatic trypsin inhibitors from the venom of Viperidae and Elapidae snakes have been isolated and sequenced (12)(13)(14)(15). These snake venom Kunitztype protease inhibitors have been demonstrated to specifically inhibit the proteolytic activity of trypsin or chymotrypsin. Nevertheless, their physiological roles in the regulatory mechanisms that influence the proteases in coagulation, fibrinolysis, and inflammation have been rarely considered. Three protease inhibitor-like protein genes have been cloned from Bungarus multicinctus genome in our laboratory (16). The deduced protein sequences of protease inhibitor-like proteins are highly homologous with those of Kunitz-type protease inhibitors. However, their biological activities remain elusive. Because soybean Kunitz-type trypsin inhibitor has been found to induce apoptotic death of human leukemia Jurkat cells (17), anti-leukemia activity of B. multicinctus protease inhibitor-like proteins is thus examined. In this study, human leukemia U937 cells were treated with B. multicinctus protease inhibitor like protein-1 (PILP-1). It was found that PILP-1-induced downregulation of a disintegrin and metalloprotease 17 (ADAM17) led to inactivation of Lyn/Akt pathways. The signaling pathways further triggered apoptosis of U937 cells through the mitochondrion-mediated death pathway. Collectively, our data elucidate a novel ADAM17/Lyn/Akt signaling pathway in maintaining the viability of leukemia cells and suggest a strategy in improving leukemia therapy through suppression of ADAM17 protein expression.
RNA Preparation and RT-PCR-Total RNA was isolated from untreated control cells or PILP-1-treated cells using the RNeasy minikit (Qiagen Inc., Valencia, CA) according to the manufacturer's instructions. Reverse transcriptase reaction was performed with 2 g of total RNA using Moloney murine leukemia virus reverse transcriptase (Promega) as recommended by the manufacturer. A reaction without reverse transcriptase was performed in parallel to ensure the absence of genomic DNA contamination. After initial denaturation at 95°C for 10 min, PCR amplification was performed using GoTaq Flexi DNA polymerase (Promega) followed by 30 cycles at 94°C for 60 s, 55°C for 60 s, and 72°C for 60 s. After a final extension at 72°C for 5 min, PCR products were resolved on 2% agarose gels and visualized by ethidium bromide transillumination under UV light. Primer sequences were as follows: TNFR2 (forward), 5Ј-ACATCAGACGTGGTGTGCAA-3Ј, and TNFR2 (reverse), 5Ј-CCAACTGGAAGAGCGAAGTC-3Ј; ADAM17 (forward), 5Ј-CAGCACAGCTGCCAAGTCATT-3Ј and ADAM17 (reverse), 5Ј-CCAGCATCTGCTAAGTCACTTCC-3Ј. The PCR yielded PCR products of 323 and 235 bp for TNFR2 and ADAM17, respectively. Each reverse-transcribed mRNA product was internally controlled by glyceraldehyde-3-phosphate dehydrogenase (GAPDH) PCR using primers 5Ј-GAGTCAAC-GGATTTGGTCGT-3Ј (forward) and 5Ј-TGTGGTCATGAG-TCCTTCCA-3Ј (reverse), yielding a 512-bp PCR product. The TNFR2 and ADAM17 reverse transcriptase-PCR products were subsequently confirmed by direct sequencing.
Cloning of Luciferase Reporter Plasmid of ADAM17 Promoter and Luciferase Assay-DNA segment containing nucleotides Ϫ938 to ϩ235 of the human ADAM17 gene was amplified by PCR from human genomic DNA. The PCR-amplified genomic DNA was subcloned into the firefly luciferase reporter vector pGL3-basic (Promega) between KpnI and XhoI sites. The nucleotide sequences of constructs were identified by DNA sequencing. The resulting pGL-ADAM17 was used for promoter activity assay, and luciferase assay was performed with the luciferase reporter assay system (Promega).
DNA Transfection-The pCMV-MEK1 (expressed the constitutively active MEK1) and constitutively activated, myristoylated Akt vector were generous gifts from Dr. W. C. Hung (National Sun Yat-Sen University, Taiwan). Human ADAM17 expression plasmid, pME18S-ADAM17, was kindly provided by Dr. E. Nishi (Kyoto University, Japan), and pMX-IRES-GFP-Lyn expression vector was obtained from Dr. N. J. Donato (University of Michigan Comprehensive Cancer Center). LynY507F (gain of function mutant) cDNA was prepared from pMX-IRES-GFP-Lyn expression vector using the PCR method and subcloned into pcDNA3 expression vector. The plasmids were transfected into U937 cells using the pipettetype electroporator (MicroPorator-MP100, Digital Bio Technology Co., Korea).
Co-immunoprecipitation of ADAM17 and Lyn-After treatment with or without 10 M PILP-1 for 24 h, U937 cells were harvested and washed with cold PBS. The cells were then incubated with lysis buffer (50 mM NaCl, 50 mM Tris-HCl, pH 8.0, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.1 mM PMSF, 50 mM NaF, and 1 mM Na 2 VO 4 ) in ice for 10 min. After centrifugation step at 15,000 ϫ g for 20 min, the supernatant was mixed with protein G Plus/protein A-agarose suspension (Calbiochem) and incubated for 1 h at 4°C (preclearing). After removal of agarose, the cell lysate was incubated with antibodies (anti-ADAM17 or anti-Lyn) overnight at 4°C on a rotating plate. Then protein G Plus/protein A-agarose suspension (Calbiochem) was added to each sample. Following an additional 2 h of incubation at 4°C, immunoprecipitates were washed three times with lysis buffer and eluted by SDS-gel loading buffer for SDS-PAGE and Western blot analyses.
RNA Interference-ADAM17 siRNA (catalogue number sc-36604) and negative control siRNA (catalogue number sc-37007) were purchased from Santa Cruz Biotechnology, Inc. For the transfection procedure, cells were grown to 60% confluence, and 100 nM siRNA were transfected using Lipofectamine TM 2000 (Invitrogen) according to the manufacturer's instructions.
Separation of Cytosolic and Mitochondrial Fractions-Following specific treatment, cytosolic and pellet (mitochondrial) fractions were generated using a digitonin-based subcellular fractionation technique (18). Cytochrome c and proteins of the Bcl-2 family were detected by Western blot analysis.
Determination of Soluble TNF-␣ Receptor 2 (TNFR2) by ELISA-After treatment with PILP-1 for 24 h, the culture media of U937 cells were collected and centrifuged at 12,000 rpm for 10 min, and the clarified supernatants were collected. ELISA for soluble TNFR2 (sTNFR2) was performed according to the manufacturer's protocol (R & D Systems, Inc.). Developed assay plates were read at wavelength 450 nm using a plate reader, and the results were calculated using a standard curve generated each time an assay was performed.
Flow Cytometry Analyses of Cell Surface Expression of ADAM9, ADAM10, and ADAM17-After specific treatment, nonspecific antibody-binding sites were blocked by incubation with PBS containing 0.01% human IgG. One fluorescent parameter flow cytometry was performed by staining cells with monoclonal anti-human ADAM17fluorescein according to the manufacturer's protocol (R & D Systems). The stained cells were analyzed by a Beckman Coulter Epics XL flow cytometer. For detection of ADAM10 and ADAM9 protein expression on the cell surface, cells were incubated with anti-ADAM9 (AP7437a) (Abgent, San Diego) and anti-ADAM10 (MAB1427) (R & D Systems) antibodies at 4°C for 30 min. After washing, cells were incubated with fluorescein isothiocyanate-conjugated goat anti-mouse or anti-rabbit IgG (HϩL) (Anaspec, Fremont, CA) and subjected to flow cytometric analysis.
Measurement of ADAM17 Activity-ADAM17 activity was measured according to the manufacturer's protocol (Calbiochem). Its fluorescence-related enzymatic cleavage was monitored at 320 nm excitation and 405 nm emission wavelength using a microplate fluorescence reader.
Statistical Analysis-All data are presented as means Ϯ S.D. Significant differences among the groups were determined using the unpaired Student's t test. A value of p Ͻ 0.05 was taken as an indication of statistical significance. All the figures shown in this study were obtained from at least three independent experiments with similar results.
Other Tests-Cell viability assay, annexin V/propidium iodide staining, DNA content analysis, measurement of mitochondrial membrane potential, caspase-3 and -9 activity assay, separation of human peripheral blood mononuclear cells (PBMCs), and Western blot analysis were performed in essentially the same manner as described previously (18). Results of Western blots were quantified by a scanning densitometer. All bands were normalized to ␤-actin expression, and fold changes in protein expression were determined on the basis of ␤-actin loading control.
Increasing evidence suggests that altered mitochondrial function is linked to apoptosis, and a decreasing mitochondrial transmembrane potential (⌬⌿ m ) is associated with mitochondrial dysfunction (19). As shown in supplemental Fig. 2A, the increasing population of U937 cells exhibited the loss of ⌬⌿ m after PILP-1 treatment. As seen in supplemental Fig. 2B, a timedependent release of cytochrome c into cytosol was detected relative to gradual decrease in mitochondrial cytochrome c. Moreover, down-regulation of Bcl-2 and production of t-Bid were notably observed after PILP-1 treatment for 24 h. In the mitochondrial pathway, caspase-8 converts the Bid from the inactive form (22 kDa) to the active form (15 kDa), which is called truncated Bid (tBid). tBid associates with the mitochondria outer membrane, disrupts mitochondrial membrane potential (⌬⌿ m ), and releases cytochrome c into the cytoplasm. Apoptotic signals converge on mitochondria to trigger the release of cytochrome c into the cytosol, causing caspase-9 and -3 activation and cell death (20). Consistent with this result, procaspase-9 degradation was noted with PILP-1-treated cells. Moreover, PILP-1 treatment induced an increase in activities of caspase-3 and caspase-9 in U937 cells (supplemental Fig. 2C). Pretreatment with caspase-8 inhibitor abolished the production of active caspase-3 in PILP-1-treated cells, whereas caspase-3 inhibitor was unable to block PILP-1-elicited procaspase-8 degradation (supplemental Fig. 2D). It suggested that caspase-8 was located on the upstream position for procaspase-3 degradation. Moreover, Z-IETD-fmk (caspase-8 inhibitor) abolished PILP-1-induced loss of ⌬⌿ m (supplemental Fig. 2A). Taken together, the data indicate that PILP-1-induced apoptosis of U937 cells was mediated via the caspase-8/mitochondrial pathway.
Death receptors of the tumor necrosis factor (TNF) family such as Fas and tumor necrosis factor-␣ receptor 1 (TNFR1) are the best understood death pathways that recruit FADD and procaspase-8 to the receptor (22). Recruitment of procaspase-8 through FADD leads to its auto-cleavage and activation, and in turn it activates effector caspases such as caspase-3 or initial apoptosis through the mitochondrial pathway (23). Because up-regulation of death receptors evokes caspase-8 activation, protein expression of TNFR1, TNFR2, Fas, and FasL in PILP-1treated cells was thus examined. Fig. 2A shows that, unlike that  OCTOBER  of Fas, FasL, and TNFR1, protein expression of TNFR2 increased markedly in PILP-1-treated cells. Nevertheless, transcription of TNFR2 mRNA did not significantly change in PILP-1-treated cells as revealed by reverse transcription-PCR amplification (Fig. 2B). ELISA revealed that PILP-1 treatment led to a reduction in detectable soluble TNFR2 (sTNFR2) concentration in culture medium of U937 cells (Fig. 2C). Given that ectodomain shedding of TNFR2 resulted in the production of sTNFR2, our data suggested that PILP-1-induced up-regulation of TNFR2 arose from reduction in TNFR2 shedding. Knockdown of FADD did not significantly affect PILP-1-elicited procaspase-8 degradation and rescue viability of PILP-1treated cells (Fig. 2D), reflecting that the death receptor-mediated pathway was not involved in PILP-1-induced cell death.

Suppression of ADAM17-mediated Lyn/Akt Pathways
Given that ADAM17 is involved in the release of the ectodomain of TNFR2 (24), protein expression of ADAM17 in PILP-1-treated cells was examined. Fig. 3A shows that protein expression of pro-ADAM17 and ADAM17 in U937 cells was reduced in response to PILP-1 treatment. ADAM17 has been shown to be synthesized as a zymogen, which is constitutively processed to cell membrane after removal of the prodomain of pro-ADAM17. The increase in the surface expression of ADAM17 up-regulated sheddase activity of ADAM17 (25). Flow cytometry analyses revealed that PILP-1 treatment elicited a decrease in the amount of detectable ADAM17 on the cell surface. Taken together, these results suggested that PILP-1induced decrease in TNFR2 shedding was mediated through down-regulation of ADAM17. Transcriptional level of ADAM17 mRNA of PILP-1-treated cells was notably lower than that of control untreated cells as evidenced by RT-PCR assay (Fig. 3B). Promoter assay revealed that PILP-1 attenuated the luciferase activity of the ADAM17 promoter in U937 cells (Fig. 3C). SB202190 pretreatment and overexpression of the constitutively active MEK1 abrogated PILP-1-evoked decease in ADAM17 mRNA transcription, reduction in luciferase activity of the ADAM17 promoter, and down-regulation of ADAM17 (Fig. 3, B-E). Consistently, PILP-1-elicited up-regulation of TNFR2 was restored by SB202190 and transfection of constitutively active MEK1 (Fig. 3, D and E). Moreover, SB202190 and overexpression of constitutively active MEK1 abrogated PILP-1-evoked decrease in ADAM17 activity and TNFR2 shedding (Fig. 3F). Taken together, these results suggested that PILP-1-elicited ADAM17 down-regulation was mediated through p38 MAPK activation and ERK inactivation.

Suppression of ADAM17-mediated Lyn/Akt Pathways
ADAM17 is demonstrated to be a primary sheddase for multiple EGFR pro-ligands. Activation of EGFR by its ligand, which subsequently activates downstream PI3K/Akt and Ras/MAPK/ ERK signaling pathways, is important for regulating cell proliferation and cell survival (27). Fig. 5A shows that PILP-1 inactivated Akt in U937 cells. Knockdown of ADAM17 by siRNA led to a reduction in the level of phospho-Akt (Fig. 5B). The finding that overexpression of ADAM17 abolished PILP-1-induced Akt inactivation (Fig. 5C) again supported a link between ADAM17 protein expression and Akt phosphorylation. Overexpression of constitutively active Akt (CA-AKT) restored the viability of PILP-1-treated cells (Fig. 5D). PILP-1 down-regulated ADAM17 in constitutively activated, myristoylated Akt-transfected cells, but degradation of procaspases was not observed after treatment with PILP-1 (Fig. 5E). This indicated that Akt inactivation evoked activation of caspases. Moreover, overexpression of constitutively active Akt did not alter PILP-1-evoked p38 MAPK activation and ERK inactivation (Fig. 5E), whereas SB202190 pretreatment or overexpression of constitutively active MEK1 abrogated PILP-1-induced Akt inactivation (Fig. 5F). These reflected that ADAM17 down-regulation was located at an upstream position for Akt inactivation in PILP-1-treated cells.
In primary leukemic blast cells for patients with acute myeloid leukemia, the Lyn kinase is found to be constitutively activated (28,29). Lyn is one member of the Src kinase family and has been reported to be upstream of Akt and ERK1/2 in intracellular cascades of acute myeloid leukemia (29,30). Moreover, Kang et al. (31) found that interaction between Src and ADAM12 played a role in activating Src tyrosine kinase. The commercially available antibody against the phosphorylated Src Tyr-416 residue recognizes the highly conserved tyrosine phosphorylation site (EDNEpYTAR) in other members of the Src kinase family. Thus, the corresponding tyrosine site at position 396 (the positive regulatory phosphorylation site) of Lyn is recognized by the anti-phospho-Src Tyr-416 antibody. Moreover, Western blot for phosphorylated LynY507 (the negative regulatory phosphorylation site) was also conducted. Fig. 6, A and B, shows that PILP-1 treatment or down-regulation of ADAM17 led to a decrease in the level of pY396-Lyn accompanied with an increase in the level of pY507-Lyn. Overexpression of ADAM17 suppressed PILP-1-evoked changes in the levels of pY396-Lyn and pY507-Lyn (Fig. 6B). These reflected that ADAM17 was associated with activation of Lyn in U937 cells. Inhibition of Lyn activity by PP2 caused a reduction in approximately 50% viability of U937 cells (Fig. 6C). PP2 treatment did not significantly affect PILP-1-induced p38 MAPK activation, ERK inactivation, and down-regulation of ADAM17 (Fig. 6D). Nevertheless, PP2 treatment induced degradation of procaspase-8 and a decrease in the level of phospho-Akt regardless of PILP-1 treatment. Fig. 6, C and E, shows that transfection of pcDNA3-LynY507F restored viability of PILP-1-treated cells and suppressed PILP-1-induced Akt inactivation and procaspase-8 degradation. PILP-1-induced ERK inactivation, p38 MAPK activation, and ADAM17 down-regulation were still noted with pcDNA3-LynY507F-transfected cells. Taken together, these results revealed that ADAM17 down-regulation led to inactivation of the Lyn/Akt pathway in U937 cells. Previous studies found that the EGFR inhibitor (Gefitinib)induced apoptosis of U937 cells through the Akt pathway (32). Gefitinib treatment reduced the levels of phospho-Akt and phospho-ERK but did not alter phosphorylation of Lyn and p38 MAPK and ADAM17 protein expression (Fig. 6G). It reflected that ADAM17-mediated Lyn/Akt activation was not related to the EGFR pathway. To examine if endogenous ADAM17 binds to Lyn, cell lysates of control untreated cells and PILP-1-treated cells were incubated with anti-ADAM17 or anti-Lyn antibodies, respectively. The protein complexes captured by antibodies were subjected to Western blot analyses. Fig. 6F shows that Lyn was co-immunoprecipitated with ADAM17, suggesting that ADAM17 and Lyn formed protein complexes. Meanwhile, the protein complexes immunoprecipitated by either anti-ADAM17 or anti-Lyn antibodies revealed a reduction in pY396-Lyn and an increase in pY507-Lyn after PILP-1 treatment.
The supplemental Fig. 3A shows that PILP-1 treatment induced ADAM17 down-regulation and procaspase-8 degradation in chronic myeloid leukemia K562 cells. Knockdown of ADAM17 by siRNA also elicited a decrease in viability of K562 cells (supplemental Fig. 3B). Moreover, p38 MAPK activation, ERK inactivation, Akt inactivation, and Lyn inactivation were also noted with PILP-1-treated K562 cells (supplemental Fig. 3A). Obviously, PILP-1-induced death of K562 cells and U937 cells was likely mediated through the same pathways. In sharp contrast, PILP-1 did not induce significant changes in protein expression of ADAM17 and the levels of phospho-ERK, phospho-p38 MAPK, pY396-Lyn, and phospho-Akt in PBMC cells (supplemental Fig. 4A). Moreover, marginal reduction in viability of PBMCs was noted after PILP-1 treatment (supplemental Fig. 4B). These data indicated that PILP-1 showed selective cytotoxicity toward U937 and K562 cells.
Given that ADAM17, ADAM9, and ADAM10 share common membrane substrates (33), the effect of PILP-1 on protein expression of ADAM9 and ADAM10 was examined. As shown in supplemental Fig. 5A, protein expression of mature ADAM9 and ADAM10 did not significantly change in PILP-1-treated cells. Flow cytometry analyses also revealed that PILP-1 treatment did not significantly alter the amount of detectable ADAM9 and ADAM10 on the cell surface (supplemental Fig. 5B). To examine if endogenous ADAM9 or ADAM10 binds to Lyn, cell lysates of control untreated cells and PILP-1treated cells were incubated with anti-Lyn antibodies. The protein complexes captured by antibodies were subjected to Western blot analysis. It was found that Lyn could not form protein complexes with ADAM9 or ADAM10 (supplemental Fig. 5C).

DISCUSSION
ADAM is a gene family of multidomain membrane-anchored proteins, including more than 30 members in various animal species, and has been implicated in pathophysiological conditions (33,34). ADAM17 is a member of the ADAM family and originally described as being responsible for the proteolytic cleavage of the membrane-anchoring precursor form of TNF-␣ to generate a soluble form of TNF-␣ (35). Subsequent studies have shown that ADAM17 is also involved in the shedding of other biologically active proteins, including heparin-binding epidermal growth factor, transforming growth factor ␤, TNFR1, TNFR2, EGFR, vascular cell adhesion molecule-1, L-selectin, interleukin receptors, Notch, and prion proteins (36,37). ADAM17 has been shown to be synthesized as a zymogen, which is constitutively processed in the secretory pathway. An increase in the surface expression of ADAM17 up-regulated sheddase activity of ADAM17 (25). Thus, in addition to protein expression, generation of mature ADAM17 from its proform is proved to be related to its sheddase activity (37,38).
Several reports have focused on the importance of ADAM17 up-regulation in tumor malignancy. In colon carcinoma, the up-regulated expression of ADAM17 induced the activation of EGFR through the shedding of EGFR ligands (39). Tanaka et al. (40) demonstrated an increase in the expression of heparin-binding epidermal growth factor in advanced ovarian cancer and also found that it correlated significantly with the ADAM17 expression in ovarian cancer. Aberrant expression of ADAM17 has also been reported in breast cancer (41), prostate cancer (42), pancreatic ductal adenocarcinoma (43), and oral squamous cell carcinoma (44). Thus, blocking of ADAM17 expression or development of specific ADAM17 inhibitors might have potential for cancer therapy. Although specific ADAM17 inhibitors have been reported previously (45,46), ADAM17-targeted drugs also lead to the inhibition of several nontarget ADAMs or metalloproteinases (47). Thus, suppres- sion of ADAM17 protein expression may become specific modalities for cancer therapy. Our data show that PILP-1-induced death of leukemia cells is mediated through down-regulation of ADAM17 and subsequent inactivation of Lyn and Akt. Lyn has been reported to be the major Src kinase in acute myeloid leukemia, and its constitutive activation is associated with proliferation of leukemia cells (28,29). Src family contains a unique N-terminal region, an Src homology 2 (SH2) domain, an SH3 domain, a catalytic (tyrosine kinase) domain, and a short C-terminal tail. The SH3 domain is important for inter-as well as intramolecular interactions that regulate Src catalytic activity, cellular location, and recruitment of protein substrates (31). Noticeably, the cytoplasmic tail of ADAM17 contains an SH3binding site, which is suggested to potentially activate SH3 domain-containing signaling molecules such as Src and Grb (36). Consistent with the finding that the interaction between ADAM12 and Src elicits activation of Src, our data reveal that ADAM17 is probably involved in Lyn phosphorylation through a similar mechanism. Noticeably, as compared with PBMC cells, leukemia cells, including U937 cells and K562 cells, are susceptible to being death induced by PILP-1. Thus, down-regulation of ADAM17 by PILP-1 may be an adaptable strategy in improving leukemia therapy. Conclusively, PILP-1 treatment suppresses ADAM17 expression through p38 MAPK activation and ERK inactivation-mediated pathways. Down-regulation of ADAM17 leads to inactivation of Lyn/Akt pathways and consequently evokes the caspase-8/mitochondria-mediated death pathway in U937 cells.