Mutations in the P10 region of procaspase-8 lead to chemotherapy resistance in acute myeloid leukemia by impairing procaspase-8 dimerization

Mutations in the P10 region of procaspase-8 lead to chemotherapy resistance in acute myeloid... Caspase-8 activation initiates apoptotic signaling cascades, and certain mutations in procasepase-8 have been reported to be associated with the progression and prognosis of different types of tumors. In this study, we have identified four novel mutations, which are highly correlated with chemotherapy resistance and poor prognosis of acute myeloid leukemia (AML) patients, within the P10 subunit of procaspase-8. These newly discovered mutations cause premature termination of translation, resulting in truncated procaspase-8 protein, which is incapable of forming dimer to initiate apoptosis signaling pathway. Further biochemical analysis reveals that the segment of P10 subunit of procaspase-8 consisting of three amino acid residues from L491 to F493 is crucial for the formation of procaspase-8 interdimer, and the aberration of this segment disrupts the dimerization and consequently precludes the activation of caspase-8 and downstream apoptotic signaling pathway. Therefore, the patients with AML who bear these types of P10 mutations were more likely to develop chemotherapy resistance due to impaired apoptotic signaling in cellular system, leading to significantly reduced overall survival (OS) as compared with patients carrying no such types of P10 mutations. Taken together, these newly identified P10 mutations in procaspase-8 could be used as novel biomarkers for predicting response and survival of chemotherapy-treated AML patients, as well as potential therapeutic targets for medical intervention in the future. Introduction year . To date, the 5-year survival of AML patients is only Acute myeloid leukemia (AML) is characterized by the around 27% . Chemotherapy remains the first-line treat- rapid growth of abnormal white blood cells (WBCs), ment for AML patients, and approximately 40–50% of which interferes with normal blood cell production and young patients and 10–20% of elderly patients can be differentiation in the bone marrow, and it is the most cured with conventional chemotherapeutic intervention ; common type of leukemia diagnosed in adults and chil- however, still 20–30% of young patients and 40–50% of dren, accounting for ~ 1.3% of all new cancer cases per elderly patients do not respond to chemotherapy while experience primary induction failure . In addition, a substantial number of patients who have achieved com- Correspondence: X.-L. Liu (xiaoloong.liu@gmail.com) or X.-Y. Yang plete remission (CR) with initial chemotherapy eventually (yangxiaoyu683@163.com) relapse, and the prognostic outcome for patients with The School of Basic Medical Sciences, Fujian Medical University, Fuzhou 350108, People’s Republic of China recurrence is very poor, with an estimated 5-year survival Department of Biomedicine, University of Basel, Basel 4056, Switzerland rate of approximately 11% . Several studies have Full list of author information is available at the end of the article. These authors contributed equally: Ming Li, Xiao-Mo Wu, Ju Gao. Edited by M. Diederich © The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to theCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Official journal of the Cell Death Differentiation Association 1234567890():,; 1234567890():,; Li et al. Cell Death and Disease (2018) 9:516 Page 2 of 15 suggested that AML cancerous cells developed insensi- Results tivity and resistance to chemotherapeutic agents , which Mutation identification in the P10 subunit of procaspase-8 should induce tumor cell apoptosis during treatment . from AML patients and the clinical significance of The caspase family comprising a group of aspartate- identified mutations specific cysteine proteases, is playing essential roles in To identify potential mutations within the P10 region of apoptosis signaling. Caspase-8 is an activated form of procaspase-8, the last exon of procaspase-8 was amplified procaspase-8, and produced by the proteolytic cleavage of by PCR and then subjected to Sanger sequencing. In total, procaspase-8 dimer to serve as an initiator of the 15 mutations were identified in the P10 regions out of 146 sequential activation of caspases in the apoptotic signaling AML patients, and finally 4 missense mutations out of 15 cascade . Caspase-8-deficient cells are found to be resis- at a considerable high frequency were selected for further tant to apoptosis. The procaspase-8 gene consists of 11 investigation: the p.Asn475Thrfs*22 mutation (in 39 AML exons, and encodes an inactive pro-enzyme (479 aa) patients, accounting for 26.71% of the total analyzed composed of a prodomain and a caspase domain. The population), the Y465Stop mutation (in 14 AML patients, prodomain comprises two death-effector domains (DEDs) accounting for 9.59% of the total analyzed population), which are capable of interacting with the Fas-associated the Q482Stop mutation (in 8 AML patients, accounting death domain to form the death-inducing signal complex . for 5.48% of the total analyzed population), and the The caspase domain consists of two subunits known as p.Leu491Thrfs*6 mutation (in 24 AML patients, P18 and P10 (ref ), which can be further processed to form accounting for 16.43% of the total analyzed population) the active caspase-8 heterotetramer and then be released (Fig. 1a,b). These four mutations are referred as “the P10 into the cytosol to trigger the activation of remaining mutations” in the following analysis described below. downstream signaling, leading to apoptosis . To examine whether the P10 mutations exerted any Dysfunction of apoptotic signaling due to an abnorm- effects on the chemotherapy treatment and the prognosis ality in procaspase-8 has been implicated in a variety of of AML patients, an association analysis between the P10 diseases. For example, aberrant expression of procaspase- mutations and the clinical–pathological characteristics of 8 has been linked to abnormal fetal cardiovascular patients was performed. As shown in Table 1, the patients development and hematopoietic malfunction , and sev- who carried these P10 mutations had significantly higher eral polymorphisms or mutations in the procaspase-8 WBC counts (p = 0.00432) and lower CR rates after gene have been associated with head and neck squamous chemotherapy (15.30% vs. 32.79%; p = 0.0162) than the 12,13 14 cell carcinoma patients without P10 mutations (including patients with- , vaginal squamous cell carcinoma , 15 16 lung cancer , hepatocellular carcinoma , gastric carci- out P10 mutations but with other mutations and patients 17 18 19,20 noma , colon cancer , and breast cancer . To date, with wild-type (WT) procaspase-8); however, such clinical all of these reported gene polymorphisms or mutations characteristics were not significantly different between the are located in either the promoter region or the DED WT patients and the non-P10 mutation patients (without domain of procaspase-8. However, disease-associated the identified P10 mutations but with other mutations functional mutations within the caspase domain of pro- in the P10 region), as shown in Supplementary Table 1. caspase-8, especially in cancer, have not been reported In contrast, characteristics including age, sex, previously. French–American–British subtype, and CR duration (data In this study, 15 novel mutations within the P10 subunit not shown) were not statistically correlated with the P10 of the caspase domain were identified by sequencing the mutations. Meanwhile, to further confirm above obser- procaspase-8 gene from the samples of AML patients. vations, the association between clinical–pathological Four out of these 15 mutations occurred at considerable features and well-known mutations (FLT3-ITD, CEPBA, high frequency and were discovered to be strongly NMP) was also investigated; as shown in Supplementary correlated with high WBC counts and poor prognosis. Table 2, the patients who carried FLT3-ITD mutations Further biochemical elucidation revealed that these four also had significantly lower CR rates after chemotherapy mutations undermined the caspase-8-mediated apoptotic than the patients without FLT3-ITD mutations; however, pathway by disrupting the formation of dimeric the FLT3-ITD mutations were not significantly associated procaspase-8 proteins, resulting in the mutation- with the P10 mutations in our current study (Supple- harboring cells being relatively insensitive to chemother- mentary Table 3). Furthermore, based on long-term fol- apeutic agents. This study provides a better understanding low-up data, the Kaplan–Meier analysis was performed to of the heterogeneity of AML cancerous cells and therefore investigate whether these P10 mutations had any effects the implication of procaspase-8-mediated apoptosis in on the overall survival (OS), and the patients without P10 chemotherapy treatment further provides potential ther- mutations in Table 1 were further divided into non-P10 apeutic targets for medical intervention of AML patients mutations group and the WT group (without any muta- in the future. tions). Therefore, the patients were assigned into three Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 3 of 15 Fig. 1 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 4 of 15 (see figure on previous page) Fig. 1 Identified mutations in the P10 subunit of procaspase-8 from AML patients and their clinical significances. a DNA sequencing to identify the P10 mutations. Here, 1423_del A, an adenine deletion at 1423 of procaspase-8 that causes a frameshift mutation from Asn to Thr, is denoted as p.Asn475Thrfs*22; the 1395 T > A mutation, a thymine-to-adenine mutation at 1395 of procaspase-8 that produces a stop codon mutation at Tyr, is denoted as Y465Stop; the 1444 C > T mutation, a cytosine-to-thymine mutation at 1444 of procaspase-8 that produces a stop codon mutation at Gln, is denoted as Q482Stop; the 1471 ins A mutation, an adenine insertion at 1471 of procaspase-8 that results in a frameshift mutation from Leu to Thr, is denoted as p.Leu491Thrfs*6. b Schematic illustration of the P10 mutations in the procaspase-8 protein. c The P10 mutations were closely correlated with short OS for AML patients. OS was significantly different between the patients carrying the P10 mutations and those carrying no mutations in the P10 subunit (left, Kaplan–Meier, p = 0.018) and there was no statistical difference between the non-P10 mutation group and the WT group (right, Kaplan–Meier, p > 0.05) Table 1 The clinic-pathological features of 146 AML The green fluorescent protein (GFP)/Flag-tagged patients procaspase-8 WT and P10 mutant plasmids, as well as the control vector were constructed according to the sche- Variables P10 mutations Without P10 mutations p-Value matic illustration shown in Supplementary Figure 1A. Furthermore, Annexin V–fluorescein isothiocyanate Age (FITC)/propidium iodide (PI) staining was performed to <60 Years 72 45 0.1404 detect apoptosis in the presence of the P10 mutations ≥60 Years 13 16 versus procaspase-8 WT in 293T cells. As shown in Fig- Gender ure 2a, the apoptotic rates of cells overexpressing procaspase-8 WT construct were about 40% (42.140 ± Male 49 37 0.7338 2.718%), which were considerably higher than cells Female 36 24 transfected with the control vector (13.463 ± 3.418%, p < WBC 0.05); conversely, the apoptotic rates of cells transfected 9 ** >10 × 10 /L 55 29 0.00432 with the P10 mutations were only ~ 10% (13.803 ± 2.345% for p.Asn475Thrfs*22, 10.467 ± 4.519% for Y465Stop, ≤10 × 10 /L 30 32 10.580 ± 2.760% for Q482Stop, and 12.600 ± 2.554% for p. Response to chemotherapy Leu491Thrfs*6), showing no significant differences com- Remission 13 20 0.0162 paring with the control vector (p > 0.3). The cell mor- Non-remission 72 41 phology change was further observed by confocal microscopy to study the apoptosis of each transfected χ test, * p < 0.05; ** p < 0.01 cells. As shown in Figure 2b, the cells overexpressing procaspase-8 WT construct were shrink and showed a typical apoptotic morphology, whereas the cells over- experimental groups. As shown in Fig. 1c, the patients expressing the P10 mutations or the control vector with P10 mutations in the cohort had a significantly remained intact and showed a normal morphology. Fur- shorter median survival time (86 weeks, 95% confidence thermore, the cell viability of each transfected group was interval (CI) 76.877–105.038) than the WT patients determined by the Cell Counting Kit-8 (CCK-8) assay. As without mutations in the P10 subunit (141 weeks, 95% CI shown in Figure 2c, the viability of cells overexpressing 97.037–138.358), whereas the patients carrying the non- procaspase-8 WT was 15.4 ± 1.1% after 24 h of transfec- P10 mutations (126 weeks, 95% CI 49.312–202.688) had tion and it is significantly lower than the control vector no statistically difference comparing with the WT transfected cells (100 ± 4%, p < 0.001), whereas the patients. Taken together, these data indicate that the P10 viabilities of cells transfected with the P10 mutations were mutations are highly associated with poor prognosis of 95.3 ± 1.5% (p.Asn475Thrfs*22), 96.2 ± 2.2% (Y465Stop), AML patients after chemotherapy. 95.7 ± 4% (Q482Stop), and 95.6 ± 1.2% (p.Leu491Thrfs*6), respectively, demonstrating no statistical differences P10 mutations abolish procaspase-8/caspase-8-mediated compared with the control vector (p = 0.1325, 0.2224, apoptosis 0.2563, and 0.1465, respectively). However, further pro- To investigate the molecular mechanism underlying the liferation analysis showed that these identified P10 poor clinical outcomes associated with the P10 mutations mutations had no effects on cell proliferation (Supple- and to delineate their potential functions in apoptosis, the mentary Figure 1B), indicating that the P10 mutations apoptotic signaling cascade was monitored in the pre- only abolished the procaspase-8/caspase-8-mediated 21,22 sence or absence of the P10 mutations in 293T cells . apoptosis. Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 5 of 15 Fig. 2 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 6 of 15 (see figure on previous page) Fig. 2 The P10 mutations abolish procaspase-8/caspase-8 mediated apoptosis. a Apoptotic rate was measured by flow cytometry via Annexin V–FITC/PI staining. Representative results were displayed on the left. The quantification of three independent experiments was shown on the right. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001, ns, p > 0.05). b Confocal microscopy was performed to analyze apoptosis in 293T cells. In contrast to the P10 mutations, only the cells expressing GFP-tagged procaspase-8 WT showed apoptotic morphology. Scale bar, 5 μm. c Cell viability was assessed with CCK-8 assay. The columns represented the OD values of cells from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001, ns, p > 0.05). d Δψ was analyzed by flow cytometry with JC-1 staining. Left: representative results were from three independent experiments. Right: the columns represented the average percentage of cells in P5 from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001, ns, p > 0.05). e Immunoblot analysis of Bax, Bcl-2, and Bid. Top: representative results were from three independent experiments. Bottom: the columns represented the mean level of Bax/Bcl2 (left) and Bid/Actin (right) from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001). f Immunoblot analysis of the cleavage of procaspase-8 (bloted by the anti-Flag antibody) and apoptosis-related proteins (caspase-3 and PARP). Representative results were from three independent experiments Furthermore, mitochondrial outer membrane permea- chemotherapy . To investigate the potential role of the bilization was monitored by measuring the depolarization P10 mutations in chemotherapy-associated apoptotic of the mitochondrial membrane potential (Δψ) via JC-1 dysfunction, cellular responses to apoptotic stimuli, such staining, to evaluate cellular apoptosis. As shown in Fig- as the chemotherapeutic agent etoposide and hydrogen ure 2d, the cells transfected with procaspase-8 WT dis- peroxide (H O ), were carefully analyzed. As shown in 2 2 played much higher rate of the Δψ (21.225 ± 2.708%) than Figure 3a and Supplementary Figure 1C, the etoposide the control vector-transfected cells (4.085 ± 0.262%, p < treatment significantly increased the apoptotic rates by 0.01), as well as the P10 mutations-transfected cells nearly 2.28-folds in procaspase-8 WT-transfected cells, (3.475 ± 0.290% for p.Asn475Thrfs*22, 3.665 ± 0.672% for whereas the same etoposide treatment nearly failed to Y465Stop, 3.615 ± 0.516% for Q482Stop, and 5.430 ± induce cell apoptosis in the P10 mutations-transfected 0.481% for p.Leu491Thrfs*6, with p < 0.01), whereas the cells or in the control vector-transfected cells, probably Δψ rate of the control vector-transfected cells and the P10 due to the relatively low concentration of etoposide used mutations-transfected cells were quite similar (with p > in this study. Taken together, these results suggested that 0.05). Furthermore, in-depth analysis of the related sig- the P10 mutations in procaspase-8 rendered the cells naling indicators showed the Bax/Bcl-2 ratio was much insensitive to chemotherapy treatment such as etoposide. lower, whereas the remaining Bid level was much higher It is worth noting that in an attempt to clearly observe the in the cells with P10 mutations when compared with the drug-induced apoptosis increasing and to avoid saturation procaspase-8WT-expressing cells, indicating the function of apoptosis in this study, we hereby reduced the of attenuating apoptosis by P10 mutations (Fig. 2e). expression level of each construct by using half of the Next, the apoptotic downstream signaling cascades DNA amount in previously experiments; consequently, were further investigated by analyzing the cleavage of the apoptotic rates before etoposide treatment were much procaspase-8, caspase-3, and polyadenosine diphosphate- lower than previously described in the characterization of ribose polymerase (PARP). As shown in Figure 2f, the P10 mutations (Fig. 2a). cells transfected with the P10 mutations failed to produce Furthermore, to understand whether procaspase-8 the cleaved caspase-8 proteins, in contrast with the proteins with P10 mutations are implicated in response procaspase-8 WT-transfected cells, where most of the to chemotherapy, myelogenous leukemia-derived cell line procaspase-8 proteins were proteolyzed to produce the known as K-562 and AML cell line of HEL were employed cleaved version of caspase-8. In addition, cleaved caspase- for apoptotic functional assays. We established K562, 3 and cleaved PARP could only be detected in the pre- HEL, and 293T stable cell lines expressing GFP/Flag- sence of procaspase-8 WT but not in the presence of the tagged P10 mutations and GFP/Flag control construct P10 mutations. Taken together, these results clearly through lentivirus delivering system. Due to the lethality demonstrate that these four mutations within the caused by constitutive expression of procaspase-8, the P10 subunit are capable of abolishing procaspase-8/cas- stable cell line of procaspase-8 WT could not be obtained pase-8-mediated apoptosis. and was excluded in the following experiments. Both the transcription and expression levels were subsequently Insensitive to etoposide treatment caused by the P10 examined in each established stable cell line to monitor mutations the expression efficiency of the proteins of interest, and as The aberrance of apoptotic signaling pathway in can- shown in Figure 3b, the P10 mutants were all equally cerous cells has been linked to drug resistance in expressed in K562, HEL, and 293T cells. Subsequently, Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 7 of 15 Fig. 3 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 8 of 15 (see figure on previous page) Fig. 3 Insensitive to etoposide treatment caused by the P10 mutations. a Cell apoptosis was analyzed in 293T cells transfected with the P10 mutations or procaspase-8 WT with or without the etoposide treatment for 8 h. The columns represented the average percentage of Annexin V–FITC- positive cells from three independent experiments. All data were represented as the mean ± SD (two-way ANOVA, Dunnett’s test, ****p < 0.0001). b 293T, HEL, and K562 cells were infected with lentivirus containing the protein of interest, and cells were selected with puromycin for 1 week. Left: the mRNA level was quantified by real-time PCR by targeting procaspase-8 as the template. Right: the procaspase-8 expression level was quantified by immunoblotting. The columns represented the average level from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ns, p > 0.05). c Cell apoptosis was analyzed by flow cytometry in stable 293T, HEL, and K562 cell lines treated with or without 100 nM etoposide. Representative results from three independent experiments in the stable 293T (the first panel), K562 (the second panel), and HEL (the third panel) cell lines were displayed. The columns represented the average percentage of Annexin V–FITC-positive cells for the stable 293T, K562, and HEL cell lines. All data were represented as the mean ± SD (t-test, two-tailed, **p < 0.01, ***p < 0.001, and ****p < 0.0001). The average increase of apoptotic percentage after etoposide treatment for each group was displayed at the right of the fourth panel and it was derived from three independent experiments. All data were represented as the mean ± SD (two-way ANOVA, Tukey’s test, ns, p > 0.05) the K562, HEL, and 293T stable cells were treated with preventing the dimerization of procaspase-8, co- etoposide (100 nM) for 8 h and then the cell apoptosis immunoprecipitation (Co-IP) assay was performed to was evaluated by flow cytometry. As shown in Figure 3c, examine the interaction between procaspase-8 WT and the apoptotic rates of the P10 mutants expressing K562, the P10 mutants. As shown in Figure 5a, in contrast to the HEL, and 293T cells after etoposide treatment were quite strong dimerization of the procaspase-8 WT proteins, the comparable and have no significant differences to those of interaction between the procaspase-8 P10 mutants and the control cells (p > 0.05). the procaspase-8 WT proteins can not be detected, indi- The downstream apoptotic signaling cascade in cating that the P10 mutations disrupted the dimerization response to the etoposide treatment was also investigated. of the procaspase-8 monomers. As shown in the Figure 4, the etoposide treatment led to As L491-D496 was a segment deleted in all P10 muta- increased proteolysis of endogenous procaspase-8 and tions, to delineate the role of each amino acid in the accordingly increased cleavage of caspase-3 and the PARP dimerization of procaspase-8, site-directed mutagenesis protein in all of the K562, HEL, and 293T cell lines. was performed to convert each residue within the L491- However, the etoposide treatment could not trigger the D496 segment to alanine (Ala, A) and then the constructs proteolytic cleavage of exogenous procaspase-8 proteins were transfected into 293T cells for further analysis. First, carrying P10 mutations and, consequently, there was no the apoptotic rates of these transiently transfected cells amplification of the downstream signaling cascade due to were analyzed by Annexin V–FITC/PI staining. As shown the presence of procaspase-8 P10 mutations. Consistent in Figure 5b and Supplementary Figure 4A, the apoptotic with the response to etoposide treatment, the H O rates of 293T cells expressing L491A, V492A, and F493A 2 2 treatment neither additionally enhanced apoptosis (15.853 ± 5.784%, 12.613 ± 5.684%, and 16.073 ± 7.192%, induction nor triggered the cleavage of the procaspase-8 respectively) were not significantly different from that of P10 mutants (Supplementary Figures 2 and 3). vector controlled cells (16.770 ± 6.282%, p > 0.5) but much In summary, these results demonstrated that the P10 lower than the procaspase-8 WT-expressing cells (43.327 mutations in procaspase-8 rendered the cells being ± 10.244%, p < 0.01), suggesting that the L491A, V492A, insensitive to etoposide treatment and suggested the and F493A mutations could abrogate procaspase-8/cas- possible molecular mechanism underlying the poor clin- pase-8-mediated apoptosis. However, the apoptotic rates ical outcomes to chemotherapy and drug resistance that of P494A-, S495A-, and D496A-expressing cells (37.143 ± were observed in AML patients carrying these P10 1.749%, 46.150 ± 7.272%, and 43.833 ± 1.423%, respec- mutations. tively) were quite similar to that of WT-expressing cells (p > 0.5), but significantly higher than that of the control P10 mutations in procaspase-8 inhibit dimer formation 293T cells (p < 0.05), suggesting that the P494A, S495A, Dimerization of two procaspase-8 monomers is crucial and D496A mutations retained the capability of apoptotic for exposing the hydrolysis site via conformational change induction and these substitutions had no effects on the and producing the activated caspase-8 protein through function of procaspase-8/caspase-8 signaling. Meanwhile, self-cleavage to initiate apoptosis signaling . The above the effects of these L491-D496 substitutions on Δψ were results demonstrated the P10 mutations inhibiting the also investigated. As shown in Figure 5c and Supple- activation of procaspase-8, but the underlying mechanism mentary Figure 4B, the rates of Δψ in the L491A, V492A, needs to be further explored. To investigate whether and F493A mutation-expressing cells (9.255 ± 4.094%, the P10 mutations inhibit procaspase-8 activation by 9.620 ± 3.323%, and 9.990 ± 3.790%, respectively) were not Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 9 of 15 Fig. 4 Inhibition of the proteolytic of procaspase-8 protein by the P10 mutations. The (a) 293T, (b) K562, and (c) HEL stable cell lines were treated with or without 100 nM etoposide for 8 h. After treatment, the indicated proteins were detected by immunoblot. Left: representative results were from three independent experiments (− : untreated, + : etoposide-treated). Right: the quantifications of the cleaved caspase-8, caspase-3, and PARP proteins were analyzed from three independent experiments. All data were represented as the mean ± SD (t-test, two-tailed, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001). d The increased cleavage percentage of caspase-8, caspase-3, and PARP were analyzed in each group after treatment. All data were represented as the mean ± SD (two-way ANOVA, Tukey’s test, ns, p > 0.05) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 10 of 15 Fig. 5 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 11 of 15 (see figure on previous page) Fig. 5 Inhibition of the dimerization of procaspase-8 by the P10 mutations. a 293T cells cotransfected with GFP-tagged and Flag-tagged procaspase-8 WT or the P10 mutations. After 24 h of transfection, total cell lysates were examined by IP as described in Materials and Methods. Representative results were from three independent experiments. b Cell apoptosis and c Δψ were monitored by flow cytometry. The columns represented the average percentage of Annexin V–FITC-positive cells and the average percentage of P5 from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, *p < 0.05, **p < 0.01, ***p < 0.001, ns, p > 0.05). d 293T cells expressing the indicated proteins were treated with or without 100 nM etoposide for 8 h. Apoptosis was assessed by flow cytometry. The columns represented the average percentage of Annexin V–FITC-positive cells (top) and increased apoptosis (bottom) from three independent experiments. All data were represented as the mean ± SD (Top: t-test, two-tailed, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001; Bottom: two-way ANOVA, Dunnett’s test, *p < 0.05, **p < 0.01, ****p < 0.001, ns, p > 0.05). e 293T cells were co-transfected with GFP-tagged and Flag-tagged WT or indicated procaspase-8 substitution plasmids. After 24 h of transfection, total cell lysates were examined by IP as described in Materials and Methods. Representative results were from three independent experiments significantly different from that of the control cells (5.585 and procaspase-8 containing either the L491A, V492, or ± 0.445%, p > 0.2) but much lower than that of the F493A mutation, indicating that the L491A, V492A, and procaspase-8 WT-expressing cells (p < 0.01), whereas F493A mutations could disrupt the binding ability to those of the P494A, S495A, and D496A substitutions procaspase-8 WT. In summary, the P10 mutations in (30.875 ± 2.157%, 28.450 ± 1.103%, and 28.170 ± 1.470%, procaspase-8 abolished the ability of initiating apoptotic respectively) were comparable to procaspase-8 WT- signaling by impairing the dimerization of procaspase-8 expressing cells (21.225 ± 2.708%, p > 0.05) but sig- monomers and the L491-F493 segment on the nificantly higher than the control cells (p < 0.01). P10 subunit had an essential role in self-dimerization. Furthermore, we examined whether etoposide-induced apoptosis was affected by these substitutions. As shown in Discussion Figure 5d and Supplementary Figure 4C, the etoposide Procaspase-8, as the precursor of the initiator enzyme in treatment although slightly increased the apoptotic rates the apoptotic signaling cascades, harbors various muta- of the cells transfected with L491A, V492A, and F493A tions that have been associated with a low chemother- mutations (11.0146 ± 6.160%, 24.646 ± 14.649%, and apeutic efficacy and poor prognosis in multiple types of 31.901 ± 17.392%, respectively), but showed no statisti- cancers. Here we discovered four mutations within the cally differences when compared with the increased P10 subunit of procaspase-8 from patients with AML, apoptotic rates of the cells transfected with control vector which were occurring with particularly high frequency (23.002 ± 5.902%, p > 0.05), whereas the etoposide treat- accounted for over half of the population in our screen ment significantly increased the apoptosis rates of the (85 out of 146 patients, 58.21%). The AML patients car- cells transiently transfected with the procaspase-8 WT rying these four P10 mutations had significantly higher vector or the P494A, S495A, and D496A mutations WBC counts and considerably lower CR rates after che- (114.900 ± 3.595%, 84.910 ± 2.390%, 102.966 ± 3.115%, motherapy than their counterparts carrying either the WT and 105.863 ± 13.433%, respectively) when compared with allele or non-P10 mutations. In addition, the patients with the increased apoptosis rates of the control vector- the P10 mutations had significantly shorter survival time transfected cells (p < 0.05). Interestingly, in contrast with than those without them, indicating these P10 mutations S495 and D496, the substitution at P494 residue impaired were highly correlated with poor chemotherapeutic the function of procaspase-8 to some extent, resulting in a response and prognosis in AML. statistical difference between procaspase-8 WT-expres- As procaspase-8/caspase-8 has an important role in sing cells and the P494A mutation-expressing cells (p < apoptotic regulation and signaling, we hypothesized that 0.05). Taken together, three residues of procaspase-8, the impaired apoptosis caused by these P10 mutations namely, L491, V492, and F493, were shown to be crucial was responsible for the clinical observations such as for the procaspase-8/caspase-8-mediated apoptosis sig- abnormal high levels of WBCs and, consequently, being naling pathway. resistant to chemotherapy in AML treatment. Indeed, the To verify whether these three residues affect apoptosis impairment in apoptosis was observed when the P10 by regulating the self-dimerization of procaspase-8, Co-IP mutations were overexpressed in 239T cells. The apop- assay was further performed. As shown in Figure 5e, in a totic rates of procaspase-8 proteins with the P10 muta- similar way to procaspase-8 WT, the P494A, S495A, and tions were much lower than that of procaspase-8 WT and D496A substitutions of procaspase-8 were capable of no cleavage of the procaspase-8 mutants could be detec- dimerizing with the procaspase-8 WT proteins, whereas ted. Furthermore, the treatment of apoptotic agent eto- no dimerization was evident between procaspase-8 WT poside could not increase apoptotic rates when the P10 Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 12 of 15 mutations were overexpressed, in contrast with Materials and methods procaspase-8 WT where etoposide treatment led to ~ 2.28 Clinical materials and sample preparation folds of increase in apoptosis. Taken together, these P10 Peripheral blood samples and primary bone marrow mutations identified from the AML patients are capable of specimens were collected from 146 newly diagnosed AML abolishing procaspase-8/caspase-8-mediated apoptosis patients who had only received chemotherapy at Hospital and rendering the cells being insensitive to apoptotic of Fujian Medical University between September 2015 stimulation. and December 2016. The samples were collected during Ideally, it would be better to elucidate the effects of routine diagnostic assessments in accordance with the the P10 mutations in etoposide- or H O -induced National Comprehensive Cancer Network Guidelines 2 2 apoptosis in leukemia-derived stable cells and make Version 1.2015 of Panel Members of Acute Myeloid comparison between P10 mutants and WT counterpart Leukemia Society with details provided in Table 1. The accordingly. However, due to the lethality of constitutive project was approved for all human sample collections expression of procaspase-8 in cells, the best scenario and usage by the Institution Review Board of Hospital of thereby was to compare P10 mutants with vector- Fujian Medical University. Written informed consent was controlled mock condition. In agreement with our received from each participant at the time of blood col- hypothesis, our data suggested the presence of lection in this study. procaspase-8 containing the P10 mutations could not exert any additional effects in apoptotic induction, in DNA sequencing response to apoptotic stimuli when compared with that Genomic DNA was isolated from peripheral blood of delivery the control vector. mononuclear cells using the TIANamp Genomic DNA To explore the mechanism underlying the P10 muta- Kit (TIANGEN, Beijing, China). Procaspase-8 was tions associated with apoptotic inhibition, the dimer- amplified by PCR using PrimeStar Max (TaKaRa, Dalian, ization of procaspase-8 was therefore examined. China) (forward primer 5′-GATCGGATTCCGCCAC- Compared with the strong dimerization between the CATGGACTTCAGCAGAAAT-3′ and reverse primer 5′- procaspase-8 WT proteins, the P10 mutations appeared GATCAAGCTTTTAATCAGAAGGGAAGAC-3′), and to disrupt the dimerization between the mutant and WT the final PCR products were used for Sanger sequencing monomers of procaspase-8. Residue substitution ana- (Sangon Biotech, Shanghai, China). lysis could reveal that each individual residue within the segment of L491-F493 is crucial for the formation of Cells culture procaspase-8 dimers, whereas the residues within the Human embryonic kidney cell line (HEK293T, 293T, P494-D496 segment are less important for dimerization. ATCC) was cultured in Dulbecco’s modified Eagles In line with the observations in the dimerization assay, medium (DMEM; Gibco, CA, USA). Human ery- the apoptotic rates of the P494A, S495A, and D496A troleukemia cell line (K562, ATCC) and AML cell line substitutions were comparable to that of WT procas- (HEL, ATCC) were cultured in MEM (Gibco) and RPMI- pase-8, whereas the L491A, V492, and F493A substitu- 1640 (Gibco), respectively. All the medium were supple- tions abolished the procaspase-8/caspase-8-mediated mented with 10% fetal bovine serum (FBS, Gibco) and apoptosis. 100 IU/ml penicillin–streptomycin (Sigma, Massachu- To advance our understanding of AML relapse and setts, USA), and all cells were cultured at 37 °C in 5% CO chemotherapeutic resistance, significant efforts have been incubator. made so far and numerous studies have shown that 26 27 aberrations in several genes, such as MCL-1 , TET , Plasmid generation and transfection 28–30 31–33 34–37 FLT3-ITD , NPM1 , and CEBPA , are asso- The Flag/GFP-tagged procaspase-8 expression con- ciated with drug resistance in AML treatment. Here we structs were generated using human procaspase-8 cDNA reported novel mutations within the P10 subunit of pro- ligated into the BamH1 and Xba1 sites of pcDNA3.1- caspase-8, which could inhibit the caspase-8-mediated N4Flag/NEGFP vector (recombined based on pcDNA3.1), apoptotic pathway by disrupting the dimerization of which expressed N-terminally tagged procaspase-8. PCR procaspase-8 proteins. It seems that these P10 mutations reactions were performed for all procaspase-8 cDNAs resulted in chemotherapy resistance of mutation- (forward primer: 5′-GATCGGATTCCGCCACCATG- harboring cancerous cells and therefore poor prognosis GACTTCAGCAGAAAT-3′ and reverse primer 5′-GAT- in AML patients. At last, these mutations may serve as CAAGCTTTTAATCAGAAGGGAAGAC-3′ were used, new biomarkers for predicting the chemotherapy response for WT and all P10 mutants). and prognosis of AML patients during clinical courses, or Site-directed mutations were performed with PCR-based even could provide potential therapeutic targets for mutagenesis. The Flag-procaspase-8 plasmid, which effective medical intervention in AML management. encoded procaspase-8 WT, was used as a template for Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 13 of 15 site-directed mutagenesis. The mutation site was covered further puromycin selection (2 μg/ml, Sigma, MO, USA) with the following internal primers (5′-GATCGGATTCC- for 1 week. The puromycin-resistant pools were cultured GCCACCATGGACTTCAGCAGAAAT-3′ (forward, for in DMEM or MEM containing 10% FBS and antibiotics, all site-directed mutations) and 5′-GATCTCTAGA and the expression of procaspase-8 was examined by TTAATCAGAAGGGAAGACAGCTTTTTTTCTTAG-3′ reverse transcriptase–quantitative PCR (T-qPCR). The (reverse, L491A), 5′-GATCTCTAGATTAATCAGAAGG stable cell line with only GFP/Flag was denoted as the mock control. GAAAGCAGGTTTTTTTCTTAG-3′ (reverse, V492A), 5′-GATCTCTAGATTAATCAGAAGGAGCGACAGGT TTTTTTCTTAG-3′ (reverse, F493A), 5′-GATCTCTAGA RT-qPCR analysis TTAATCAGAAGCGAAGACAAGTTTTTTTCTTAG-3′ To determine whether caspase-8 was overexpressed in (reverse, P494A), 5′-GATCTCTAGATTAATCAGCA the stable pools, the mRNA expression levels were GGGAAGAC-3′ (reverse, S495A), and 5′-GATCTCTA- examined by RT-qPCR. Briefly, total RNA was isolated GATTAAGCAGAAGGGAAGAC-3′ (reverse, D496A)). from stable pools using the TransZol Up Plus RNA Kit cDNAs containing the site-specific mutations were ligated (TRANSGEN, Beijing, China). The RNA concentration to the pcDNA3.1-N4Flag vector. was determined using the NanoDrop ND-2000 spectro- Plasmid DNA (pDNA) was amplified in Escherichia coli photometer (Thermo). Then, RNA (100 ng) was reverse- and purified according to the protocol provided in the transcribed using the Transcriptor First Strand cDNA TM E.Z.N.A. Plasmid Mini Kit (OMEGA, GA, USA). All Synthesis Kit (Roche, Basel, Switzerland) in accordance constructed plasmids were validated with restriction with the manufacturer’s protocol. Finally, the expression enzymes and sequence analysis. The pDNAs (2500 ng or levels of procaspase-8 were analyzed using RT-qPCR with 1500 ng/etoposide) were transfected into 293T cells in the Bsetar® SybrGreen qPCR Mastermix (DBI, Ludwig- six-well plates according to the protocol provided along shafen, GER) in accordance with the manufacturer’s with the Lipofectamine® 3000 transfection reagent protocol, using the gene-specific primer sets for human (Thermo, MA, USA). procaspase-8 (forward: 5′-GCTGACTTTCTGCTGGG- GAT-3′ and reverse: 5′-GACATCGCTCTCTCAGGC Generation of recombinant lentivirus particles and TC-3′) and 18S (forward: 5′-CAGCCACCCGAGATTGA procaspase-8 mutant cell lines GCA-3′ and reverse: 5′-AGTAGCGACGGGCGGTGTG- All retroviruses were produced by cotransfection of the 3′). The contents of the reactions (20 μl) were as follows: forward primer (10 nM), 0.2 μl; reverse primer (10 nM), pLenti 6-Flag/GFP or pLenti 6-Flag/GFP-procaspase-8 mutant expression constructs with packaging plasmids 0.2 μl; cDNA, 50 ng; qPCR Mastermix (2 × ), 10 μl; and PLP-1, PLP-2, and VSVG as follows: pLenti 6-Flag/GFP or water, to a final reaction volume of 20 μl. qPCR was TM pLenti 6- Flag/GFP-procaspase-8 mutants, PLP-1, PLP-2, performed on the StepOnePlus real-time PCR system and VSVG were cotransfected at a ratio of 3:1:1:1 for a (AB, MA, USA) with the following parameters: pre- total of 20 μg of DNA into the 293T cell line when cells denaturation at 95 °C for 2 min; 40 cycles at 95 °C for 15 s, were 70–80% confluent (optional 70%). The medium was 60 °C for 20 s, and 72 °C for 20 s (the signature was col- collected after 24, 48, and 96 h of transfection, then lected at this step), and a melt curve stage at 95 °C for 15 s, concentrated at 1500 g for 30 min at 4 °C, and filtered 60 °C for 1 min and 90 °C for 30 s. 18S was selected as an through a 0.45 μm filter. Afterwards, the supernatant was internal control. Raw data handling and quantification −ΔΔCt ultracentrifuged at 30,000 r.p.m. at 4 °C for 2 h (Beckman were performed with StepOne™ software. The 2 Optima XPN -100, CA, USA). The recombinant lentivirus method was used to calculated the relative level of target was resuspended with appropriate volume of phosphate- gene (procaspase-8) expression. The following equation buffered saline (PBS). The virus titer was quantified with was used: fold change = relative quantification of the the HIV-1 p24 ELISA kit (Cell Biolabs, CA, USA). The procaspase-8 P10 mutation groups/relative quantification recombinant lentivirus was stored at -80 °C for further of the mock group. usage. The 293T, K562, or HEL cells were seeding at 1 × 10 Antibodies cells (for 293T cells) or 2 × 10 cells (for K562 or HEL All antibodies were purchased from Cell Signaling cells) per well in six-well plates in the presence of poly- Technologies (MA, USA) unless indicated. Antibodies brene (5 μg/ml). The amount of recombinant lentivirus used were as following: anti-caspase8 (1:2000, #9746), was calculated ([multiplicity of infection = 30] × cells/ anti-caspase3 (1:1000, #9665), anti-cleaved caspase3 titer), and then certain amount of virus was added to the (1:1000, #9664 P), anti-cleaved PARP (1:1000, #5625 P), cells. After 6 h of infection, the medium was changed to anti-PARP (1:1000, #9532 P), anti-GFP (1:2000, #2956), fresh medium and the cells were cultured normally. A anti-DYKDDDK (1:1000, Flag, TRANSGEN, HT201-02), confluence of 70–80% of infected cells was chosen for anti-Bax (1:1000, #5023), anti-Bcl2 (1:1000, Abcam, Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 14 of 15 ab32124, Cambridge, UK), anti-β-Actin (1:1000, stable transfected cells were cultured to 70–80% con- TRANSGEN, HC201-02), anti-mouse (1:5000, #7076P2), fluence and then used for further analysis. For the apop- and anti-rabbit (1:5000, 7074P2) for immunoblotting and tosis induction analysis, the cells were treated with Co-IPs. etoposide (Sigma-Aldrich, MO, USA) or H O (Sigma- 2 2 Aldrich) for 8 h. Then, all cells were collected for apop- Immunoblotting and CO-IP tosis analysis by flow cytometry (BD FACSCalibur™, CA, For immunoblotting, collected cells transfected with the USA) using the Annexin V–FITC/PI Apoptosis Detection procaspase-8 WT or P10 mutants, or stable pools were Kit (Dojindo, Kumamoto, Japan) according to the man- washed in cold PBS at 4 °C and lysed in RIPA buffer (50 ufacturer’s protocol. For the membrane potential analysis, mM Tris-HCl pH 7.4, 150 mM NaCl, 1% Triton X-100, the transfected cells (cells were transfected with the Flag- and 0.1% SDS) supplemented with 0.1 mM phe- procaspase-8 WT, Flag-procaspase-8 P10 mutants, or nylmethylsulfonyl fluoride and a protease inhibitor cock- Flag-procaspase-8 L491A/V492A/F493A/P494A/S495A/ tail (Roche) on ice for 20–30 min. Lysates were D496A single site mutation plasmids) were cultured for centrifuged at 12,000 × g for 30 min at 4 °C. Supernatants 24 h; then, the rate of Δψ was measured by flow cytometry were collected and the protein concentrations were using the Mitochondrial Membrane Potential Detection quantified according to the protocol provided in the Easy JC-1 Kit (BD, CA, USA) according to the manufacturer’s II Protein Quantitative Kit (BCA, TRANSGEN) by the protocol. SpectraMax® M5 Multi-Mode microplate reader (Mole- cular Devices, CA, USA). Supernatants were mixed with CCK-8 assay the same volumes of 2 × loading buffer and boiled for 10 Cell viability was measured with the CCK-8 (Dojindo) min. Approximately 50 µg of protein was separated by according to the manufacturer’s protocol. Briefly, 10% SDS-polyacrylamide gel electrophoresis (PAGE), 293T cells were seeded 5000 cells/well into 96-well plates. transferred onto a nitrocellulose membrane, blocked with At 60–70% confluence, the cells were transfected with the 5% skim milk (BD, MD, USA) in TBST (20 mM Tris-HCl, Flag-procaspase-8 WT or Flag-procaspase-8 P10 muta- 500 mM NaCl pH 7.5, 0.1% Tween-20) at room tem- tions, or Flag-procaspase-8 L491A/V492A/F493A/P494A/ perature for 2 h and then incubated with relevant primary S495A/D496A single site mutation plasmids (100 ng/well) antibodies at 4 °C overnight. Subsequently, the mem- according to the protocol provided along with the Lipo- branes were incubated with the corresponding secondary fectamine® 3000 transfection reagent. After culture for 24 antibodies at room temperature for 2 h. Finally, the results h, 10 μl of CCK-8 solution was added to each well, and the were visualized by chemiluminescence. plate was incubated in incubator (37 °C, 5% CO ) for For Co-IPs, 293T cells were co-transfected with the another 1 h. Then, the optical density value was measured GFP-procaspase-8 WT and Flag/Flag-procaspase-8 P10 at the wavelength of 450 nm on the SpectraMax® M5 mutant constructs in 60 mm dishes. Cells were cultured Multi-Mode microplate reader to analyze cell viability. for 24 h and lysed in an IP buffer (50 mM Tris-HCl pH 8.0, 1% NP-40, 150 mM NaCl, 1 mM EDTA) containing a Statistical analysis protease inhibitor cocktail on ice for 30 min. Following Data were presented as the mean ± SD and analyzed centrifugation at 13,300 × g for 30 min at 4 °C, super- with SPSS 19.0 and GraphPad Prism 6.0. Three inde- natants were collected. As the input samples, several of pendent experiments were performed for all measure- the cell lysates were mixed with the same volume of 2 × ments. The differences between two groups were analyzed loading buffer and boiled for 10 min. The remaining cell with Student’s t-test. Statistical comparisons of the CR lysates were precleared by incubation with 1 μg of anti- duration and OS were performed using the Kaplan–Meier GFP or anti-Flag antibody at 4 °C overnight. Meanwhile, analysis. p < 0.05 was considered as statistical significance. Dynabeads protein G (Thermo) were washed three times with IP buffer using a magnetic particle concentrator Acknowledgements (Thermo). Then, the rinsed beads (40 μl) were added to This work is supported by the National Natural Science Foundation of China (81602102, 31071311, 31600616), the Natural Science Foundation of Fujian each sample and incubated for 2–4 h at 4 °C. Finally, the Province (2017J05142), and the Project of Fuzhou Science and Technology bead–antibody complexes were washed in IP buffer for 3 Program of China (2014-S-146, 2016-S-124-10, 2016-S-124-4). times, resuspended in 50 μl of 1 × loading buffer, boiled for 10 min, and separated by 10% SDS-PAGE for Author details immunoblotting. The School of Basic Medical Sciences, Fujian Medical University, Fuzhou 350108, People’s Republic of China. Department of Biomedicine, University of Basel, Basel 4056, Switzerland. The United Innovation of Mengchao Apoptosis and mitochondrial membrane potential analysis Hepatobiliary Technology Key Laboratory of Fujian Province, Mengchao The transiently transfected cells were cultured for 24 h Hepatobiliary Hospital of Fujian Medical University, Fuzhou 350002, People’s after the transfection for further analysis, whereas the Republic of China. The First People’s Hospital of Huaihua, Huaihua 418000, Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 15 of 15 People’s Republic of China. Fuzhou Maternity and Child Healthcare Hospital, 16. Cho, S. et al. Epigenetic methylation and expression of caspase 8 and survivin Fuzhou 350005, People’s Republic of China in hepatocellular carcinoma. Pathol. Int. 60,203–211 (2010). 17. Soung,Y. H.et al. CASPASE-8 gene is inactivated by somatic mutations in Conflict of interest gastric carcinomas. Cancer Res. 65, 815–821 (2005). The authors declare that they have no conflict of interest. 18. Kim, H. S. et al. Inactivating mutations of caspase-8 gene in colorectal carci- nomas. Gastroenterology 125,708–715 (2003). 19. Brynychová, V. et al. Clinical and functional importance of selected CASP8 andCASP9 polymorphismsinbreastcarcinoma. Klin. Onkol. 29,445–453 Publisher's note (2016). Springer Nature remains neutral with regard to jurisdictional claims in 20. Hashemi,M.et al. Bi-directional PCR allele-specificamplification (bi-PASA) for published maps and institutional affiliations. detection of caspase-8 -652 6N ins_del promoter polymorphism (rs3834129) in breast cancer. Gene 505,176–179 (2012). Supplementary Information accompanies this paper at https://doi.org/ 21. Tamm, I. et al. IAP-family protein Survivin inhibits caspase activity and apop- 10.1038/s41419-018-0511-3. tosis induced by Fas (CD95), Bax, caspases, and anticancer drugs. Cancer Res. 58,5315–5320 (1998). Received: 2 November 2017 Revised: 16 March 2018 Accepted: 19 March 22. Patel, V., Balakrishnan, K., Keating, M. J., Wierda, W. G. & Gandhi, V. Expression of executioner procaspases and their activation by a procaspase-activating compound in chronic lymphocytic leukemia cells. Blood 125,1126–1136 (2015). 23. Rathore, R., McCallum, J. E., Varghese, E., Florea, A. M. & Busselberg, D. Over- coming chemotherapy drug resistance by targeting inhibitors of apoptosis References proteins (IAPs). Apoptosis 22,898–919 (2017). 1. Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics. CA Cancer J. Clin. 67,7–30 24. Valter, K. et al. Contrasting effects of glutamine deprivation on apoptosis (2017). induced by conventionally used anticancer drugs. Biochim. Biophys. Acta 1864, 2. Bose,P., Vachhani, P.&Cortes,J.E.Treatment of relapsed/refractoryacute 498–506 (2017). myeloid leukemia. Curr. Treat. Options Oncol. 18, 17 (2017). 25. Muzio,M., Stockwell, B. R.,Stennicke,H.R., Salvesen,G.S.&Dixit, V. M. An 3. Ofran, Y. & Rowe, J. M. Treatment for relapsed acute myeloid leukemia: what is induced proximity model for caspase-8 activation. J. Biol. Chem. 273, new? Curr. Opin. Hematol. 19,89–94 (2012). 2926–2930 (1998). 4. Rowe,J.M., Li,X.& Cassileth,P.A.VerypoorsurvivalofpatientswithAML who 26. Glaser, S. P. et al. Anti-apoptotic Mcl-1 is essential for the development and relapse after achieving a first complete remission: the Eastern Cooperative sustained growth of acute myeloid leukemia. Genes Dev. 26,120–125 (2012). Oncology Group Experience. Blood 106, 546a (2005). 27. Nakajima, H. & Kunimoto, H. TET2 as an epigenetic master regulator for normal 5. Konopleva, M. et al. Mechanisms of apoptosis sensitivity and resistance to the and malignant hematopoiesis. Cancer Sci. 105,1093–1099 (2014). BH3 mimetic ABT-737 in acute myeloid leukemia. Cancer Cell 10,375–388 28. Annesley, C. E. & Brown, P. The biology and targeting of FLT3 in pediatric (2006). leukemia. Front. Oncol. 4, 263 (2014). 6. Carter,B.Z.et al. Synergistic targeting of AML stem/progenitor cells with IAP 29. Kantarjian, H. Acute myeloid leukemia–majorprogressoverfourdecades and antagonist birinapant and demethylating agents. J. Natl. Cancer Inst. 106, glimpses into the future. Am.J.Hematol. 91,131–145 (2016). djt440 (2014). 30. Levis, M. FLT3 mutations in acute myeloid leukemia: what is the best 7. Kruidering, M. & Evan, G. I. Caspase-8 in apoptosis: the beginning of the end. approach in 2013? Hematol. Am. Soc. Hematol. Educ. Program 2013,220–226 Iubmb. Life 50,85–90 (2000). 8. Carrington, P. E. et al. The structure of FADD and its mode of interaction with (2013). 31. Falini, B. et al. Cytoplasmic nucleophosmin in acute myelogenous leukemia procaspase-8. Mol. Cell 22,599–610 (2006). with a normal karyotype. N. Engl. J. Med. 352,254–266 (2005). 9. Schug, Z. T., Gonzalvez, F., Houtkooper, R. H., Vaz, F. M. & Gottlieb, E. BID is 32. Suzuki, T. et al. Clinical characteristics and prognostic implications of NPM1 cleaved by caspase-8 within a native complex on the mitochondrial mem- mutations in acute myeloid leukemia. Blood 106,2854–2861 (2005). brane. Cell Death Differ. 18,538–548 (2011). 33. Kottaridis, P. D. et al. The presence of a FLT3 internal tandem duplication in 10. Hoffmann, J. C.,Pappa,A., Krammer, P. H. & Lavrik, I. N. A new C-terminal patients with acute myeloid leukemia (AML) adds important prognostic cleavage product of procaspase-8, p30, defines an alternative pathway of information to cytogenetic risk group and response to the first cycle of procaspase-8 activation. Mol. Cell. Biol. 29, 4431–4440 (2009). chemotherapy: analysis of 854 patients from the United Kingdom Medical 11. Kang, T. B. et al. Mutation of a self-processing site in caspase-8 compromises its Research Council AML 10 and 12 trials. Blood 98, 1752–1759 (2001). apoptotic but not its nonapoptotic functions in bacterial artificial chromo- 34. Walter, R. B. et al. Significance of FAB subclassification of “acute myeloid some transgenic mice. J. Immunol. 181,2522–2532 (2008). leukemia, NOS” in the2008WHO classification: analysis of 5848 newly diag- 12. Ando, M. et al. Cancer-associated missense mutations of caspase-8 activate nosed patients. Blood 121,2424–2431 (2013). nuclear factor-κB signaling. Cancer Sci. 104,1002–1008 (2013). 35. Fröhling, S. et al. CEBPA mutations in younger adults with acute myeloid 13. Li, C., Egloff, A. M., Sen, M., Grandis, J. R. & Johnson, D. E. Caspase-8 mutations in leukemia and normal cytogenetics: prognostic relevance and analysis of head and neck cancer confer resistance to death receptor-mediated apoptosis cooperating mutations. J. Clin. Oncol. 22,624–633 (2004). and enhance migration, invasion, and tumor growth. Mol. Oncol. 8,1220–1230 36. Bienz, M. et al. Risk assessment in patients with acute myeloid leukemia and a (2014). normal karyotype. Clin. Cancer Res. 11,1416–1424 (2005). 14. Liu,B., Peng,D., Lu,Y., Jin, W. &Fan,Z.Anovelsingleamino acid deletion 37. Boissel, N. et al. Prevalence, clinical profile, and prognosis of NPM mutations in caspase-8 mutant in cancer cells that lost proapoptotic activity. J. Biol. Chem. AML with normal karyotype. Blood 106,3618–3620 (2005). 277, 30159–30164 (2002). 38. Livak, K.J.&Schmittgen,T.D.Analysisofrelativegeneexpressiondata using 15. Qian, J. et al. Association between CASP8 and CASP10 polymorphisms and real-time quantitative PCR and the 2−ΔΔCT method. Methods 25,402–408 toxicity outcomes with platinum-based chemotherapy in Chinese patients (2001). with non-small cell lung cancer. Oncologist 17,1551–1561 (2012). Official journal of the Cell Death Differentiation Association http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Cell Death & Disease Springer Journals

Mutations in the P10 region of procaspase-8 lead to chemotherapy resistance in acute myeloid leukemia by impairing procaspase-8 dimerization

Free
15 pages

Loading next page...
 
/lp/springer_journal/mutations-in-the-p10-region-of-procaspase-8-lead-to-chemotherapy-VCAErhARvh
Publisher
Springer Journals
Copyright
Copyright © 2018 by The Author(s)
Subject
Life Sciences; Life Sciences, general; Biochemistry, general; Cell Biology; Immunology; Cell Culture; Antibodies
eISSN
2041-4889
D.O.I.
10.1038/s41419-018-0511-3
Publisher site
See Article on Publisher Site

Abstract

Caspase-8 activation initiates apoptotic signaling cascades, and certain mutations in procasepase-8 have been reported to be associated with the progression and prognosis of different types of tumors. In this study, we have identified four novel mutations, which are highly correlated with chemotherapy resistance and poor prognosis of acute myeloid leukemia (AML) patients, within the P10 subunit of procaspase-8. These newly discovered mutations cause premature termination of translation, resulting in truncated procaspase-8 protein, which is incapable of forming dimer to initiate apoptosis signaling pathway. Further biochemical analysis reveals that the segment of P10 subunit of procaspase-8 consisting of three amino acid residues from L491 to F493 is crucial for the formation of procaspase-8 interdimer, and the aberration of this segment disrupts the dimerization and consequently precludes the activation of caspase-8 and downstream apoptotic signaling pathway. Therefore, the patients with AML who bear these types of P10 mutations were more likely to develop chemotherapy resistance due to impaired apoptotic signaling in cellular system, leading to significantly reduced overall survival (OS) as compared with patients carrying no such types of P10 mutations. Taken together, these newly identified P10 mutations in procaspase-8 could be used as novel biomarkers for predicting response and survival of chemotherapy-treated AML patients, as well as potential therapeutic targets for medical intervention in the future. Introduction year . To date, the 5-year survival of AML patients is only Acute myeloid leukemia (AML) is characterized by the around 27% . Chemotherapy remains the first-line treat- rapid growth of abnormal white blood cells (WBCs), ment for AML patients, and approximately 40–50% of which interferes with normal blood cell production and young patients and 10–20% of elderly patients can be differentiation in the bone marrow, and it is the most cured with conventional chemotherapeutic intervention ; common type of leukemia diagnosed in adults and chil- however, still 20–30% of young patients and 40–50% of dren, accounting for ~ 1.3% of all new cancer cases per elderly patients do not respond to chemotherapy while experience primary induction failure . In addition, a substantial number of patients who have achieved com- Correspondence: X.-L. Liu (xiaoloong.liu@gmail.com) or X.-Y. Yang plete remission (CR) with initial chemotherapy eventually (yangxiaoyu683@163.com) relapse, and the prognostic outcome for patients with The School of Basic Medical Sciences, Fujian Medical University, Fuzhou 350108, People’s Republic of China recurrence is very poor, with an estimated 5-year survival Department of Biomedicine, University of Basel, Basel 4056, Switzerland rate of approximately 11% . Several studies have Full list of author information is available at the end of the article. These authors contributed equally: Ming Li, Xiao-Mo Wu, Ju Gao. Edited by M. Diederich © The Author(s) 2018 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to theCreativeCommons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/. Official journal of the Cell Death Differentiation Association 1234567890():,; 1234567890():,; Li et al. Cell Death and Disease (2018) 9:516 Page 2 of 15 suggested that AML cancerous cells developed insensi- Results tivity and resistance to chemotherapeutic agents , which Mutation identification in the P10 subunit of procaspase-8 should induce tumor cell apoptosis during treatment . from AML patients and the clinical significance of The caspase family comprising a group of aspartate- identified mutations specific cysteine proteases, is playing essential roles in To identify potential mutations within the P10 region of apoptosis signaling. Caspase-8 is an activated form of procaspase-8, the last exon of procaspase-8 was amplified procaspase-8, and produced by the proteolytic cleavage of by PCR and then subjected to Sanger sequencing. In total, procaspase-8 dimer to serve as an initiator of the 15 mutations were identified in the P10 regions out of 146 sequential activation of caspases in the apoptotic signaling AML patients, and finally 4 missense mutations out of 15 cascade . Caspase-8-deficient cells are found to be resis- at a considerable high frequency were selected for further tant to apoptosis. The procaspase-8 gene consists of 11 investigation: the p.Asn475Thrfs*22 mutation (in 39 AML exons, and encodes an inactive pro-enzyme (479 aa) patients, accounting for 26.71% of the total analyzed composed of a prodomain and a caspase domain. The population), the Y465Stop mutation (in 14 AML patients, prodomain comprises two death-effector domains (DEDs) accounting for 9.59% of the total analyzed population), which are capable of interacting with the Fas-associated the Q482Stop mutation (in 8 AML patients, accounting death domain to form the death-inducing signal complex . for 5.48% of the total analyzed population), and the The caspase domain consists of two subunits known as p.Leu491Thrfs*6 mutation (in 24 AML patients, P18 and P10 (ref ), which can be further processed to form accounting for 16.43% of the total analyzed population) the active caspase-8 heterotetramer and then be released (Fig. 1a,b). These four mutations are referred as “the P10 into the cytosol to trigger the activation of remaining mutations” in the following analysis described below. downstream signaling, leading to apoptosis . To examine whether the P10 mutations exerted any Dysfunction of apoptotic signaling due to an abnorm- effects on the chemotherapy treatment and the prognosis ality in procaspase-8 has been implicated in a variety of of AML patients, an association analysis between the P10 diseases. For example, aberrant expression of procaspase- mutations and the clinical–pathological characteristics of 8 has been linked to abnormal fetal cardiovascular patients was performed. As shown in Table 1, the patients development and hematopoietic malfunction , and sev- who carried these P10 mutations had significantly higher eral polymorphisms or mutations in the procaspase-8 WBC counts (p = 0.00432) and lower CR rates after gene have been associated with head and neck squamous chemotherapy (15.30% vs. 32.79%; p = 0.0162) than the 12,13 14 cell carcinoma patients without P10 mutations (including patients with- , vaginal squamous cell carcinoma , 15 16 lung cancer , hepatocellular carcinoma , gastric carci- out P10 mutations but with other mutations and patients 17 18 19,20 noma , colon cancer , and breast cancer . To date, with wild-type (WT) procaspase-8); however, such clinical all of these reported gene polymorphisms or mutations characteristics were not significantly different between the are located in either the promoter region or the DED WT patients and the non-P10 mutation patients (without domain of procaspase-8. However, disease-associated the identified P10 mutations but with other mutations functional mutations within the caspase domain of pro- in the P10 region), as shown in Supplementary Table 1. caspase-8, especially in cancer, have not been reported In contrast, characteristics including age, sex, previously. French–American–British subtype, and CR duration (data In this study, 15 novel mutations within the P10 subunit not shown) were not statistically correlated with the P10 of the caspase domain were identified by sequencing the mutations. Meanwhile, to further confirm above obser- procaspase-8 gene from the samples of AML patients. vations, the association between clinical–pathological Four out of these 15 mutations occurred at considerable features and well-known mutations (FLT3-ITD, CEPBA, high frequency and were discovered to be strongly NMP) was also investigated; as shown in Supplementary correlated with high WBC counts and poor prognosis. Table 2, the patients who carried FLT3-ITD mutations Further biochemical elucidation revealed that these four also had significantly lower CR rates after chemotherapy mutations undermined the caspase-8-mediated apoptotic than the patients without FLT3-ITD mutations; however, pathway by disrupting the formation of dimeric the FLT3-ITD mutations were not significantly associated procaspase-8 proteins, resulting in the mutation- with the P10 mutations in our current study (Supple- harboring cells being relatively insensitive to chemother- mentary Table 3). Furthermore, based on long-term fol- apeutic agents. This study provides a better understanding low-up data, the Kaplan–Meier analysis was performed to of the heterogeneity of AML cancerous cells and therefore investigate whether these P10 mutations had any effects the implication of procaspase-8-mediated apoptosis in on the overall survival (OS), and the patients without P10 chemotherapy treatment further provides potential ther- mutations in Table 1 were further divided into non-P10 apeutic targets for medical intervention of AML patients mutations group and the WT group (without any muta- in the future. tions). Therefore, the patients were assigned into three Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 3 of 15 Fig. 1 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 4 of 15 (see figure on previous page) Fig. 1 Identified mutations in the P10 subunit of procaspase-8 from AML patients and their clinical significances. a DNA sequencing to identify the P10 mutations. Here, 1423_del A, an adenine deletion at 1423 of procaspase-8 that causes a frameshift mutation from Asn to Thr, is denoted as p.Asn475Thrfs*22; the 1395 T > A mutation, a thymine-to-adenine mutation at 1395 of procaspase-8 that produces a stop codon mutation at Tyr, is denoted as Y465Stop; the 1444 C > T mutation, a cytosine-to-thymine mutation at 1444 of procaspase-8 that produces a stop codon mutation at Gln, is denoted as Q482Stop; the 1471 ins A mutation, an adenine insertion at 1471 of procaspase-8 that results in a frameshift mutation from Leu to Thr, is denoted as p.Leu491Thrfs*6. b Schematic illustration of the P10 mutations in the procaspase-8 protein. c The P10 mutations were closely correlated with short OS for AML patients. OS was significantly different between the patients carrying the P10 mutations and those carrying no mutations in the P10 subunit (left, Kaplan–Meier, p = 0.018) and there was no statistical difference between the non-P10 mutation group and the WT group (right, Kaplan–Meier, p > 0.05) Table 1 The clinic-pathological features of 146 AML The green fluorescent protein (GFP)/Flag-tagged patients procaspase-8 WT and P10 mutant plasmids, as well as the control vector were constructed according to the sche- Variables P10 mutations Without P10 mutations p-Value matic illustration shown in Supplementary Figure 1A. Furthermore, Annexin V–fluorescein isothiocyanate Age (FITC)/propidium iodide (PI) staining was performed to <60 Years 72 45 0.1404 detect apoptosis in the presence of the P10 mutations ≥60 Years 13 16 versus procaspase-8 WT in 293T cells. As shown in Fig- Gender ure 2a, the apoptotic rates of cells overexpressing procaspase-8 WT construct were about 40% (42.140 ± Male 49 37 0.7338 2.718%), which were considerably higher than cells Female 36 24 transfected with the control vector (13.463 ± 3.418%, p < WBC 0.05); conversely, the apoptotic rates of cells transfected 9 ** >10 × 10 /L 55 29 0.00432 with the P10 mutations were only ~ 10% (13.803 ± 2.345% for p.Asn475Thrfs*22, 10.467 ± 4.519% for Y465Stop, ≤10 × 10 /L 30 32 10.580 ± 2.760% for Q482Stop, and 12.600 ± 2.554% for p. Response to chemotherapy Leu491Thrfs*6), showing no significant differences com- Remission 13 20 0.0162 paring with the control vector (p > 0.3). The cell mor- Non-remission 72 41 phology change was further observed by confocal microscopy to study the apoptosis of each transfected χ test, * p < 0.05; ** p < 0.01 cells. As shown in Figure 2b, the cells overexpressing procaspase-8 WT construct were shrink and showed a typical apoptotic morphology, whereas the cells over- experimental groups. As shown in Fig. 1c, the patients expressing the P10 mutations or the control vector with P10 mutations in the cohort had a significantly remained intact and showed a normal morphology. Fur- shorter median survival time (86 weeks, 95% confidence thermore, the cell viability of each transfected group was interval (CI) 76.877–105.038) than the WT patients determined by the Cell Counting Kit-8 (CCK-8) assay. As without mutations in the P10 subunit (141 weeks, 95% CI shown in Figure 2c, the viability of cells overexpressing 97.037–138.358), whereas the patients carrying the non- procaspase-8 WT was 15.4 ± 1.1% after 24 h of transfec- P10 mutations (126 weeks, 95% CI 49.312–202.688) had tion and it is significantly lower than the control vector no statistically difference comparing with the WT transfected cells (100 ± 4%, p < 0.001), whereas the patients. Taken together, these data indicate that the P10 viabilities of cells transfected with the P10 mutations were mutations are highly associated with poor prognosis of 95.3 ± 1.5% (p.Asn475Thrfs*22), 96.2 ± 2.2% (Y465Stop), AML patients after chemotherapy. 95.7 ± 4% (Q482Stop), and 95.6 ± 1.2% (p.Leu491Thrfs*6), respectively, demonstrating no statistical differences P10 mutations abolish procaspase-8/caspase-8-mediated compared with the control vector (p = 0.1325, 0.2224, apoptosis 0.2563, and 0.1465, respectively). However, further pro- To investigate the molecular mechanism underlying the liferation analysis showed that these identified P10 poor clinical outcomes associated with the P10 mutations mutations had no effects on cell proliferation (Supple- and to delineate their potential functions in apoptosis, the mentary Figure 1B), indicating that the P10 mutations apoptotic signaling cascade was monitored in the pre- only abolished the procaspase-8/caspase-8-mediated 21,22 sence or absence of the P10 mutations in 293T cells . apoptosis. Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 5 of 15 Fig. 2 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 6 of 15 (see figure on previous page) Fig. 2 The P10 mutations abolish procaspase-8/caspase-8 mediated apoptosis. a Apoptotic rate was measured by flow cytometry via Annexin V–FITC/PI staining. Representative results were displayed on the left. The quantification of three independent experiments was shown on the right. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001, ns, p > 0.05). b Confocal microscopy was performed to analyze apoptosis in 293T cells. In contrast to the P10 mutations, only the cells expressing GFP-tagged procaspase-8 WT showed apoptotic morphology. Scale bar, 5 μm. c Cell viability was assessed with CCK-8 assay. The columns represented the OD values of cells from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001, ns, p > 0.05). d Δψ was analyzed by flow cytometry with JC-1 staining. Left: representative results were from three independent experiments. Right: the columns represented the average percentage of cells in P5 from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001, ns, p > 0.05). e Immunoblot analysis of Bax, Bcl-2, and Bid. Top: representative results were from three independent experiments. Bottom: the columns represented the mean level of Bax/Bcl2 (left) and Bid/Actin (right) from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ****p < 0.0001). f Immunoblot analysis of the cleavage of procaspase-8 (bloted by the anti-Flag antibody) and apoptosis-related proteins (caspase-3 and PARP). Representative results were from three independent experiments Furthermore, mitochondrial outer membrane permea- chemotherapy . To investigate the potential role of the bilization was monitored by measuring the depolarization P10 mutations in chemotherapy-associated apoptotic of the mitochondrial membrane potential (Δψ) via JC-1 dysfunction, cellular responses to apoptotic stimuli, such staining, to evaluate cellular apoptosis. As shown in Fig- as the chemotherapeutic agent etoposide and hydrogen ure 2d, the cells transfected with procaspase-8 WT dis- peroxide (H O ), were carefully analyzed. As shown in 2 2 played much higher rate of the Δψ (21.225 ± 2.708%) than Figure 3a and Supplementary Figure 1C, the etoposide the control vector-transfected cells (4.085 ± 0.262%, p < treatment significantly increased the apoptotic rates by 0.01), as well as the P10 mutations-transfected cells nearly 2.28-folds in procaspase-8 WT-transfected cells, (3.475 ± 0.290% for p.Asn475Thrfs*22, 3.665 ± 0.672% for whereas the same etoposide treatment nearly failed to Y465Stop, 3.615 ± 0.516% for Q482Stop, and 5.430 ± induce cell apoptosis in the P10 mutations-transfected 0.481% for p.Leu491Thrfs*6, with p < 0.01), whereas the cells or in the control vector-transfected cells, probably Δψ rate of the control vector-transfected cells and the P10 due to the relatively low concentration of etoposide used mutations-transfected cells were quite similar (with p > in this study. Taken together, these results suggested that 0.05). Furthermore, in-depth analysis of the related sig- the P10 mutations in procaspase-8 rendered the cells naling indicators showed the Bax/Bcl-2 ratio was much insensitive to chemotherapy treatment such as etoposide. lower, whereas the remaining Bid level was much higher It is worth noting that in an attempt to clearly observe the in the cells with P10 mutations when compared with the drug-induced apoptosis increasing and to avoid saturation procaspase-8WT-expressing cells, indicating the function of apoptosis in this study, we hereby reduced the of attenuating apoptosis by P10 mutations (Fig. 2e). expression level of each construct by using half of the Next, the apoptotic downstream signaling cascades DNA amount in previously experiments; consequently, were further investigated by analyzing the cleavage of the apoptotic rates before etoposide treatment were much procaspase-8, caspase-3, and polyadenosine diphosphate- lower than previously described in the characterization of ribose polymerase (PARP). As shown in Figure 2f, the P10 mutations (Fig. 2a). cells transfected with the P10 mutations failed to produce Furthermore, to understand whether procaspase-8 the cleaved caspase-8 proteins, in contrast with the proteins with P10 mutations are implicated in response procaspase-8 WT-transfected cells, where most of the to chemotherapy, myelogenous leukemia-derived cell line procaspase-8 proteins were proteolyzed to produce the known as K-562 and AML cell line of HEL were employed cleaved version of caspase-8. In addition, cleaved caspase- for apoptotic functional assays. We established K562, 3 and cleaved PARP could only be detected in the pre- HEL, and 293T stable cell lines expressing GFP/Flag- sence of procaspase-8 WT but not in the presence of the tagged P10 mutations and GFP/Flag control construct P10 mutations. Taken together, these results clearly through lentivirus delivering system. Due to the lethality demonstrate that these four mutations within the caused by constitutive expression of procaspase-8, the P10 subunit are capable of abolishing procaspase-8/cas- stable cell line of procaspase-8 WT could not be obtained pase-8-mediated apoptosis. and was excluded in the following experiments. Both the transcription and expression levels were subsequently Insensitive to etoposide treatment caused by the P10 examined in each established stable cell line to monitor mutations the expression efficiency of the proteins of interest, and as The aberrance of apoptotic signaling pathway in can- shown in Figure 3b, the P10 mutants were all equally cerous cells has been linked to drug resistance in expressed in K562, HEL, and 293T cells. Subsequently, Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 7 of 15 Fig. 3 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 8 of 15 (see figure on previous page) Fig. 3 Insensitive to etoposide treatment caused by the P10 mutations. a Cell apoptosis was analyzed in 293T cells transfected with the P10 mutations or procaspase-8 WT with or without the etoposide treatment for 8 h. The columns represented the average percentage of Annexin V–FITC- positive cells from three independent experiments. All data were represented as the mean ± SD (two-way ANOVA, Dunnett’s test, ****p < 0.0001). b 293T, HEL, and K562 cells were infected with lentivirus containing the protein of interest, and cells were selected with puromycin for 1 week. Left: the mRNA level was quantified by real-time PCR by targeting procaspase-8 as the template. Right: the procaspase-8 expression level was quantified by immunoblotting. The columns represented the average level from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, ns, p > 0.05). c Cell apoptosis was analyzed by flow cytometry in stable 293T, HEL, and K562 cell lines treated with or without 100 nM etoposide. Representative results from three independent experiments in the stable 293T (the first panel), K562 (the second panel), and HEL (the third panel) cell lines were displayed. The columns represented the average percentage of Annexin V–FITC-positive cells for the stable 293T, K562, and HEL cell lines. All data were represented as the mean ± SD (t-test, two-tailed, **p < 0.01, ***p < 0.001, and ****p < 0.0001). The average increase of apoptotic percentage after etoposide treatment for each group was displayed at the right of the fourth panel and it was derived from three independent experiments. All data were represented as the mean ± SD (two-way ANOVA, Tukey’s test, ns, p > 0.05) the K562, HEL, and 293T stable cells were treated with preventing the dimerization of procaspase-8, co- etoposide (100 nM) for 8 h and then the cell apoptosis immunoprecipitation (Co-IP) assay was performed to was evaluated by flow cytometry. As shown in Figure 3c, examine the interaction between procaspase-8 WT and the apoptotic rates of the P10 mutants expressing K562, the P10 mutants. As shown in Figure 5a, in contrast to the HEL, and 293T cells after etoposide treatment were quite strong dimerization of the procaspase-8 WT proteins, the comparable and have no significant differences to those of interaction between the procaspase-8 P10 mutants and the control cells (p > 0.05). the procaspase-8 WT proteins can not be detected, indi- The downstream apoptotic signaling cascade in cating that the P10 mutations disrupted the dimerization response to the etoposide treatment was also investigated. of the procaspase-8 monomers. As shown in the Figure 4, the etoposide treatment led to As L491-D496 was a segment deleted in all P10 muta- increased proteolysis of endogenous procaspase-8 and tions, to delineate the role of each amino acid in the accordingly increased cleavage of caspase-3 and the PARP dimerization of procaspase-8, site-directed mutagenesis protein in all of the K562, HEL, and 293T cell lines. was performed to convert each residue within the L491- However, the etoposide treatment could not trigger the D496 segment to alanine (Ala, A) and then the constructs proteolytic cleavage of exogenous procaspase-8 proteins were transfected into 293T cells for further analysis. First, carrying P10 mutations and, consequently, there was no the apoptotic rates of these transiently transfected cells amplification of the downstream signaling cascade due to were analyzed by Annexin V–FITC/PI staining. As shown the presence of procaspase-8 P10 mutations. Consistent in Figure 5b and Supplementary Figure 4A, the apoptotic with the response to etoposide treatment, the H O rates of 293T cells expressing L491A, V492A, and F493A 2 2 treatment neither additionally enhanced apoptosis (15.853 ± 5.784%, 12.613 ± 5.684%, and 16.073 ± 7.192%, induction nor triggered the cleavage of the procaspase-8 respectively) were not significantly different from that of P10 mutants (Supplementary Figures 2 and 3). vector controlled cells (16.770 ± 6.282%, p > 0.5) but much In summary, these results demonstrated that the P10 lower than the procaspase-8 WT-expressing cells (43.327 mutations in procaspase-8 rendered the cells being ± 10.244%, p < 0.01), suggesting that the L491A, V492A, insensitive to etoposide treatment and suggested the and F493A mutations could abrogate procaspase-8/cas- possible molecular mechanism underlying the poor clin- pase-8-mediated apoptosis. However, the apoptotic rates ical outcomes to chemotherapy and drug resistance that of P494A-, S495A-, and D496A-expressing cells (37.143 ± were observed in AML patients carrying these P10 1.749%, 46.150 ± 7.272%, and 43.833 ± 1.423%, respec- mutations. tively) were quite similar to that of WT-expressing cells (p > 0.5), but significantly higher than that of the control P10 mutations in procaspase-8 inhibit dimer formation 293T cells (p < 0.05), suggesting that the P494A, S495A, Dimerization of two procaspase-8 monomers is crucial and D496A mutations retained the capability of apoptotic for exposing the hydrolysis site via conformational change induction and these substitutions had no effects on the and producing the activated caspase-8 protein through function of procaspase-8/caspase-8 signaling. Meanwhile, self-cleavage to initiate apoptosis signaling . The above the effects of these L491-D496 substitutions on Δψ were results demonstrated the P10 mutations inhibiting the also investigated. As shown in Figure 5c and Supple- activation of procaspase-8, but the underlying mechanism mentary Figure 4B, the rates of Δψ in the L491A, V492A, needs to be further explored. To investigate whether and F493A mutation-expressing cells (9.255 ± 4.094%, the P10 mutations inhibit procaspase-8 activation by 9.620 ± 3.323%, and 9.990 ± 3.790%, respectively) were not Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 9 of 15 Fig. 4 Inhibition of the proteolytic of procaspase-8 protein by the P10 mutations. The (a) 293T, (b) K562, and (c) HEL stable cell lines were treated with or without 100 nM etoposide for 8 h. After treatment, the indicated proteins were detected by immunoblot. Left: representative results were from three independent experiments (− : untreated, + : etoposide-treated). Right: the quantifications of the cleaved caspase-8, caspase-3, and PARP proteins were analyzed from three independent experiments. All data were represented as the mean ± SD (t-test, two-tailed, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001). d The increased cleavage percentage of caspase-8, caspase-3, and PARP were analyzed in each group after treatment. All data were represented as the mean ± SD (two-way ANOVA, Tukey’s test, ns, p > 0.05) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 10 of 15 Fig. 5 (See legend on next page.) Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 11 of 15 (see figure on previous page) Fig. 5 Inhibition of the dimerization of procaspase-8 by the P10 mutations. a 293T cells cotransfected with GFP-tagged and Flag-tagged procaspase-8 WT or the P10 mutations. After 24 h of transfection, total cell lysates were examined by IP as described in Materials and Methods. Representative results were from three independent experiments. b Cell apoptosis and c Δψ were monitored by flow cytometry. The columns represented the average percentage of Annexin V–FITC-positive cells and the average percentage of P5 from three independent experiments. All data were represented as the mean ± SD (one-way ANOVA, *p < 0.05, **p < 0.01, ***p < 0.001, ns, p > 0.05). d 293T cells expressing the indicated proteins were treated with or without 100 nM etoposide for 8 h. Apoptosis was assessed by flow cytometry. The columns represented the average percentage of Annexin V–FITC-positive cells (top) and increased apoptosis (bottom) from three independent experiments. All data were represented as the mean ± SD (Top: t-test, two-tailed, *p < 0.05, **p < 0.01, ***p < 0.001 and ****p < 0.0001; Bottom: two-way ANOVA, Dunnett’s test, *p < 0.05, **p < 0.01, ****p < 0.001, ns, p > 0.05). e 293T cells were co-transfected with GFP-tagged and Flag-tagged WT or indicated procaspase-8 substitution plasmids. After 24 h of transfection, total cell lysates were examined by IP as described in Materials and Methods. Representative results were from three independent experiments significantly different from that of the control cells (5.585 and procaspase-8 containing either the L491A, V492, or ± 0.445%, p > 0.2) but much lower than that of the F493A mutation, indicating that the L491A, V492A, and procaspase-8 WT-expressing cells (p < 0.01), whereas F493A mutations could disrupt the binding ability to those of the P494A, S495A, and D496A substitutions procaspase-8 WT. In summary, the P10 mutations in (30.875 ± 2.157%, 28.450 ± 1.103%, and 28.170 ± 1.470%, procaspase-8 abolished the ability of initiating apoptotic respectively) were comparable to procaspase-8 WT- signaling by impairing the dimerization of procaspase-8 expressing cells (21.225 ± 2.708%, p > 0.05) but sig- monomers and the L491-F493 segment on the nificantly higher than the control cells (p < 0.01). P10 subunit had an essential role in self-dimerization. Furthermore, we examined whether etoposide-induced apoptosis was affected by these substitutions. As shown in Discussion Figure 5d and Supplementary Figure 4C, the etoposide Procaspase-8, as the precursor of the initiator enzyme in treatment although slightly increased the apoptotic rates the apoptotic signaling cascades, harbors various muta- of the cells transfected with L491A, V492A, and F493A tions that have been associated with a low chemother- mutations (11.0146 ± 6.160%, 24.646 ± 14.649%, and apeutic efficacy and poor prognosis in multiple types of 31.901 ± 17.392%, respectively), but showed no statisti- cancers. Here we discovered four mutations within the cally differences when compared with the increased P10 subunit of procaspase-8 from patients with AML, apoptotic rates of the cells transfected with control vector which were occurring with particularly high frequency (23.002 ± 5.902%, p > 0.05), whereas the etoposide treat- accounted for over half of the population in our screen ment significantly increased the apoptosis rates of the (85 out of 146 patients, 58.21%). The AML patients car- cells transiently transfected with the procaspase-8 WT rying these four P10 mutations had significantly higher vector or the P494A, S495A, and D496A mutations WBC counts and considerably lower CR rates after che- (114.900 ± 3.595%, 84.910 ± 2.390%, 102.966 ± 3.115%, motherapy than their counterparts carrying either the WT and 105.863 ± 13.433%, respectively) when compared with allele or non-P10 mutations. In addition, the patients with the increased apoptosis rates of the control vector- the P10 mutations had significantly shorter survival time transfected cells (p < 0.05). Interestingly, in contrast with than those without them, indicating these P10 mutations S495 and D496, the substitution at P494 residue impaired were highly correlated with poor chemotherapeutic the function of procaspase-8 to some extent, resulting in a response and prognosis in AML. statistical difference between procaspase-8 WT-expres- As procaspase-8/caspase-8 has an important role in sing cells and the P494A mutation-expressing cells (p < apoptotic regulation and signaling, we hypothesized that 0.05). Taken together, three residues of procaspase-8, the impaired apoptosis caused by these P10 mutations namely, L491, V492, and F493, were shown to be crucial was responsible for the clinical observations such as for the procaspase-8/caspase-8-mediated apoptosis sig- abnormal high levels of WBCs and, consequently, being naling pathway. resistant to chemotherapy in AML treatment. Indeed, the To verify whether these three residues affect apoptosis impairment in apoptosis was observed when the P10 by regulating the self-dimerization of procaspase-8, Co-IP mutations were overexpressed in 239T cells. The apop- assay was further performed. As shown in Figure 5e, in a totic rates of procaspase-8 proteins with the P10 muta- similar way to procaspase-8 WT, the P494A, S495A, and tions were much lower than that of procaspase-8 WT and D496A substitutions of procaspase-8 were capable of no cleavage of the procaspase-8 mutants could be detec- dimerizing with the procaspase-8 WT proteins, whereas ted. Furthermore, the treatment of apoptotic agent eto- no dimerization was evident between procaspase-8 WT poside could not increase apoptotic rates when the P10 Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 12 of 15 mutations were overexpressed, in contrast with Materials and methods procaspase-8 WT where etoposide treatment led to ~ 2.28 Clinical materials and sample preparation folds of increase in apoptosis. Taken together, these P10 Peripheral blood samples and primary bone marrow mutations identified from the AML patients are capable of specimens were collected from 146 newly diagnosed AML abolishing procaspase-8/caspase-8-mediated apoptosis patients who had only received chemotherapy at Hospital and rendering the cells being insensitive to apoptotic of Fujian Medical University between September 2015 stimulation. and December 2016. The samples were collected during Ideally, it would be better to elucidate the effects of routine diagnostic assessments in accordance with the the P10 mutations in etoposide- or H O -induced National Comprehensive Cancer Network Guidelines 2 2 apoptosis in leukemia-derived stable cells and make Version 1.2015 of Panel Members of Acute Myeloid comparison between P10 mutants and WT counterpart Leukemia Society with details provided in Table 1. The accordingly. However, due to the lethality of constitutive project was approved for all human sample collections expression of procaspase-8 in cells, the best scenario and usage by the Institution Review Board of Hospital of thereby was to compare P10 mutants with vector- Fujian Medical University. Written informed consent was controlled mock condition. In agreement with our received from each participant at the time of blood col- hypothesis, our data suggested the presence of lection in this study. procaspase-8 containing the P10 mutations could not exert any additional effects in apoptotic induction, in DNA sequencing response to apoptotic stimuli when compared with that Genomic DNA was isolated from peripheral blood of delivery the control vector. mononuclear cells using the TIANamp Genomic DNA To explore the mechanism underlying the P10 muta- Kit (TIANGEN, Beijing, China). Procaspase-8 was tions associated with apoptotic inhibition, the dimer- amplified by PCR using PrimeStar Max (TaKaRa, Dalian, ization of procaspase-8 was therefore examined. China) (forward primer 5′-GATCGGATTCCGCCAC- Compared with the strong dimerization between the CATGGACTTCAGCAGAAAT-3′ and reverse primer 5′- procaspase-8 WT proteins, the P10 mutations appeared GATCAAGCTTTTAATCAGAAGGGAAGAC-3′), and to disrupt the dimerization between the mutant and WT the final PCR products were used for Sanger sequencing monomers of procaspase-8. Residue substitution ana- (Sangon Biotech, Shanghai, China). lysis could reveal that each individual residue within the segment of L491-F493 is crucial for the formation of Cells culture procaspase-8 dimers, whereas the residues within the Human embryonic kidney cell line (HEK293T, 293T, P494-D496 segment are less important for dimerization. ATCC) was cultured in Dulbecco’s modified Eagles In line with the observations in the dimerization assay, medium (DMEM; Gibco, CA, USA). Human ery- the apoptotic rates of the P494A, S495A, and D496A troleukemia cell line (K562, ATCC) and AML cell line substitutions were comparable to that of WT procas- (HEL, ATCC) were cultured in MEM (Gibco) and RPMI- pase-8, whereas the L491A, V492, and F493A substitu- 1640 (Gibco), respectively. All the medium were supple- tions abolished the procaspase-8/caspase-8-mediated mented with 10% fetal bovine serum (FBS, Gibco) and apoptosis. 100 IU/ml penicillin–streptomycin (Sigma, Massachu- To advance our understanding of AML relapse and setts, USA), and all cells were cultured at 37 °C in 5% CO chemotherapeutic resistance, significant efforts have been incubator. made so far and numerous studies have shown that 26 27 aberrations in several genes, such as MCL-1 , TET , Plasmid generation and transfection 28–30 31–33 34–37 FLT3-ITD , NPM1 , and CEBPA , are asso- The Flag/GFP-tagged procaspase-8 expression con- ciated with drug resistance in AML treatment. Here we structs were generated using human procaspase-8 cDNA reported novel mutations within the P10 subunit of pro- ligated into the BamH1 and Xba1 sites of pcDNA3.1- caspase-8, which could inhibit the caspase-8-mediated N4Flag/NEGFP vector (recombined based on pcDNA3.1), apoptotic pathway by disrupting the dimerization of which expressed N-terminally tagged procaspase-8. PCR procaspase-8 proteins. It seems that these P10 mutations reactions were performed for all procaspase-8 cDNAs resulted in chemotherapy resistance of mutation- (forward primer: 5′-GATCGGATTCCGCCACCATG- harboring cancerous cells and therefore poor prognosis GACTTCAGCAGAAAT-3′ and reverse primer 5′-GAT- in AML patients. At last, these mutations may serve as CAAGCTTTTAATCAGAAGGGAAGAC-3′ were used, new biomarkers for predicting the chemotherapy response for WT and all P10 mutants). and prognosis of AML patients during clinical courses, or Site-directed mutations were performed with PCR-based even could provide potential therapeutic targets for mutagenesis. The Flag-procaspase-8 plasmid, which effective medical intervention in AML management. encoded procaspase-8 WT, was used as a template for Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 13 of 15 site-directed mutagenesis. The mutation site was covered further puromycin selection (2 μg/ml, Sigma, MO, USA) with the following internal primers (5′-GATCGGATTCC- for 1 week. The puromycin-resistant pools were cultured GCCACCATGGACTTCAGCAGAAAT-3′ (forward, for in DMEM or MEM containing 10% FBS and antibiotics, all site-directed mutations) and 5′-GATCTCTAGA and the expression of procaspase-8 was examined by TTAATCAGAAGGGAAGACAGCTTTTTTTCTTAG-3′ reverse transcriptase–quantitative PCR (T-qPCR). The (reverse, L491A), 5′-GATCTCTAGATTAATCAGAAGG stable cell line with only GFP/Flag was denoted as the mock control. GAAAGCAGGTTTTTTTCTTAG-3′ (reverse, V492A), 5′-GATCTCTAGATTAATCAGAAGGAGCGACAGGT TTTTTTCTTAG-3′ (reverse, F493A), 5′-GATCTCTAGA RT-qPCR analysis TTAATCAGAAGCGAAGACAAGTTTTTTTCTTAG-3′ To determine whether caspase-8 was overexpressed in (reverse, P494A), 5′-GATCTCTAGATTAATCAGCA the stable pools, the mRNA expression levels were GGGAAGAC-3′ (reverse, S495A), and 5′-GATCTCTA- examined by RT-qPCR. Briefly, total RNA was isolated GATTAAGCAGAAGGGAAGAC-3′ (reverse, D496A)). from stable pools using the TransZol Up Plus RNA Kit cDNAs containing the site-specific mutations were ligated (TRANSGEN, Beijing, China). The RNA concentration to the pcDNA3.1-N4Flag vector. was determined using the NanoDrop ND-2000 spectro- Plasmid DNA (pDNA) was amplified in Escherichia coli photometer (Thermo). Then, RNA (100 ng) was reverse- and purified according to the protocol provided in the transcribed using the Transcriptor First Strand cDNA TM E.Z.N.A. Plasmid Mini Kit (OMEGA, GA, USA). All Synthesis Kit (Roche, Basel, Switzerland) in accordance constructed plasmids were validated with restriction with the manufacturer’s protocol. Finally, the expression enzymes and sequence analysis. The pDNAs (2500 ng or levels of procaspase-8 were analyzed using RT-qPCR with 1500 ng/etoposide) were transfected into 293T cells in the Bsetar® SybrGreen qPCR Mastermix (DBI, Ludwig- six-well plates according to the protocol provided along shafen, GER) in accordance with the manufacturer’s with the Lipofectamine® 3000 transfection reagent protocol, using the gene-specific primer sets for human (Thermo, MA, USA). procaspase-8 (forward: 5′-GCTGACTTTCTGCTGGG- GAT-3′ and reverse: 5′-GACATCGCTCTCTCAGGC Generation of recombinant lentivirus particles and TC-3′) and 18S (forward: 5′-CAGCCACCCGAGATTGA procaspase-8 mutant cell lines GCA-3′ and reverse: 5′-AGTAGCGACGGGCGGTGTG- All retroviruses were produced by cotransfection of the 3′). The contents of the reactions (20 μl) were as follows: forward primer (10 nM), 0.2 μl; reverse primer (10 nM), pLenti 6-Flag/GFP or pLenti 6-Flag/GFP-procaspase-8 mutant expression constructs with packaging plasmids 0.2 μl; cDNA, 50 ng; qPCR Mastermix (2 × ), 10 μl; and PLP-1, PLP-2, and VSVG as follows: pLenti 6-Flag/GFP or water, to a final reaction volume of 20 μl. qPCR was TM pLenti 6- Flag/GFP-procaspase-8 mutants, PLP-1, PLP-2, performed on the StepOnePlus real-time PCR system and VSVG were cotransfected at a ratio of 3:1:1:1 for a (AB, MA, USA) with the following parameters: pre- total of 20 μg of DNA into the 293T cell line when cells denaturation at 95 °C for 2 min; 40 cycles at 95 °C for 15 s, were 70–80% confluent (optional 70%). The medium was 60 °C for 20 s, and 72 °C for 20 s (the signature was col- collected after 24, 48, and 96 h of transfection, then lected at this step), and a melt curve stage at 95 °C for 15 s, concentrated at 1500 g for 30 min at 4 °C, and filtered 60 °C for 1 min and 90 °C for 30 s. 18S was selected as an through a 0.45 μm filter. Afterwards, the supernatant was internal control. Raw data handling and quantification −ΔΔCt ultracentrifuged at 30,000 r.p.m. at 4 °C for 2 h (Beckman were performed with StepOne™ software. The 2 Optima XPN -100, CA, USA). The recombinant lentivirus method was used to calculated the relative level of target was resuspended with appropriate volume of phosphate- gene (procaspase-8) expression. The following equation buffered saline (PBS). The virus titer was quantified with was used: fold change = relative quantification of the the HIV-1 p24 ELISA kit (Cell Biolabs, CA, USA). The procaspase-8 P10 mutation groups/relative quantification recombinant lentivirus was stored at -80 °C for further of the mock group. usage. The 293T, K562, or HEL cells were seeding at 1 × 10 Antibodies cells (for 293T cells) or 2 × 10 cells (for K562 or HEL All antibodies were purchased from Cell Signaling cells) per well in six-well plates in the presence of poly- Technologies (MA, USA) unless indicated. Antibodies brene (5 μg/ml). The amount of recombinant lentivirus used were as following: anti-caspase8 (1:2000, #9746), was calculated ([multiplicity of infection = 30] × cells/ anti-caspase3 (1:1000, #9665), anti-cleaved caspase3 titer), and then certain amount of virus was added to the (1:1000, #9664 P), anti-cleaved PARP (1:1000, #5625 P), cells. After 6 h of infection, the medium was changed to anti-PARP (1:1000, #9532 P), anti-GFP (1:2000, #2956), fresh medium and the cells were cultured normally. A anti-DYKDDDK (1:1000, Flag, TRANSGEN, HT201-02), confluence of 70–80% of infected cells was chosen for anti-Bax (1:1000, #5023), anti-Bcl2 (1:1000, Abcam, Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 14 of 15 ab32124, Cambridge, UK), anti-β-Actin (1:1000, stable transfected cells were cultured to 70–80% con- TRANSGEN, HC201-02), anti-mouse (1:5000, #7076P2), fluence and then used for further analysis. For the apop- and anti-rabbit (1:5000, 7074P2) for immunoblotting and tosis induction analysis, the cells were treated with Co-IPs. etoposide (Sigma-Aldrich, MO, USA) or H O (Sigma- 2 2 Aldrich) for 8 h. Then, all cells were collected for apop- Immunoblotting and CO-IP tosis analysis by flow cytometry (BD FACSCalibur™, CA, For immunoblotting, collected cells transfected with the USA) using the Annexin V–FITC/PI Apoptosis Detection procaspase-8 WT or P10 mutants, or stable pools were Kit (Dojindo, Kumamoto, Japan) according to the man- washed in cold PBS at 4 °C and lysed in RIPA buffer (50 ufacturer’s protocol. For the membrane potential analysis, mM Tris-HCl pH 7.4, 150 mM NaCl, 1% Triton X-100, the transfected cells (cells were transfected with the Flag- and 0.1% SDS) supplemented with 0.1 mM phe- procaspase-8 WT, Flag-procaspase-8 P10 mutants, or nylmethylsulfonyl fluoride and a protease inhibitor cock- Flag-procaspase-8 L491A/V492A/F493A/P494A/S495A/ tail (Roche) on ice for 20–30 min. Lysates were D496A single site mutation plasmids) were cultured for centrifuged at 12,000 × g for 30 min at 4 °C. Supernatants 24 h; then, the rate of Δψ was measured by flow cytometry were collected and the protein concentrations were using the Mitochondrial Membrane Potential Detection quantified according to the protocol provided in the Easy JC-1 Kit (BD, CA, USA) according to the manufacturer’s II Protein Quantitative Kit (BCA, TRANSGEN) by the protocol. SpectraMax® M5 Multi-Mode microplate reader (Mole- cular Devices, CA, USA). Supernatants were mixed with CCK-8 assay the same volumes of 2 × loading buffer and boiled for 10 Cell viability was measured with the CCK-8 (Dojindo) min. Approximately 50 µg of protein was separated by according to the manufacturer’s protocol. Briefly, 10% SDS-polyacrylamide gel electrophoresis (PAGE), 293T cells were seeded 5000 cells/well into 96-well plates. transferred onto a nitrocellulose membrane, blocked with At 60–70% confluence, the cells were transfected with the 5% skim milk (BD, MD, USA) in TBST (20 mM Tris-HCl, Flag-procaspase-8 WT or Flag-procaspase-8 P10 muta- 500 mM NaCl pH 7.5, 0.1% Tween-20) at room tem- tions, or Flag-procaspase-8 L491A/V492A/F493A/P494A/ perature for 2 h and then incubated with relevant primary S495A/D496A single site mutation plasmids (100 ng/well) antibodies at 4 °C overnight. Subsequently, the mem- according to the protocol provided along with the Lipo- branes were incubated with the corresponding secondary fectamine® 3000 transfection reagent. After culture for 24 antibodies at room temperature for 2 h. Finally, the results h, 10 μl of CCK-8 solution was added to each well, and the were visualized by chemiluminescence. plate was incubated in incubator (37 °C, 5% CO ) for For Co-IPs, 293T cells were co-transfected with the another 1 h. Then, the optical density value was measured GFP-procaspase-8 WT and Flag/Flag-procaspase-8 P10 at the wavelength of 450 nm on the SpectraMax® M5 mutant constructs in 60 mm dishes. Cells were cultured Multi-Mode microplate reader to analyze cell viability. for 24 h and lysed in an IP buffer (50 mM Tris-HCl pH 8.0, 1% NP-40, 150 mM NaCl, 1 mM EDTA) containing a Statistical analysis protease inhibitor cocktail on ice for 30 min. Following Data were presented as the mean ± SD and analyzed centrifugation at 13,300 × g for 30 min at 4 °C, super- with SPSS 19.0 and GraphPad Prism 6.0. Three inde- natants were collected. As the input samples, several of pendent experiments were performed for all measure- the cell lysates were mixed with the same volume of 2 × ments. The differences between two groups were analyzed loading buffer and boiled for 10 min. The remaining cell with Student’s t-test. Statistical comparisons of the CR lysates were precleared by incubation with 1 μg of anti- duration and OS were performed using the Kaplan–Meier GFP or anti-Flag antibody at 4 °C overnight. Meanwhile, analysis. p < 0.05 was considered as statistical significance. Dynabeads protein G (Thermo) were washed three times with IP buffer using a magnetic particle concentrator Acknowledgements (Thermo). Then, the rinsed beads (40 μl) were added to This work is supported by the National Natural Science Foundation of China (81602102, 31071311, 31600616), the Natural Science Foundation of Fujian each sample and incubated for 2–4 h at 4 °C. Finally, the Province (2017J05142), and the Project of Fuzhou Science and Technology bead–antibody complexes were washed in IP buffer for 3 Program of China (2014-S-146, 2016-S-124-10, 2016-S-124-4). times, resuspended in 50 μl of 1 × loading buffer, boiled for 10 min, and separated by 10% SDS-PAGE for Author details immunoblotting. The School of Basic Medical Sciences, Fujian Medical University, Fuzhou 350108, People’s Republic of China. Department of Biomedicine, University of Basel, Basel 4056, Switzerland. The United Innovation of Mengchao Apoptosis and mitochondrial membrane potential analysis Hepatobiliary Technology Key Laboratory of Fujian Province, Mengchao The transiently transfected cells were cultured for 24 h Hepatobiliary Hospital of Fujian Medical University, Fuzhou 350002, People’s after the transfection for further analysis, whereas the Republic of China. The First People’s Hospital of Huaihua, Huaihua 418000, Official journal of the Cell Death Differentiation Association Li et al. Cell Death and Disease (2018) 9:516 Page 15 of 15 People’s Republic of China. Fuzhou Maternity and Child Healthcare Hospital, 16. Cho, S. et al. Epigenetic methylation and expression of caspase 8 and survivin Fuzhou 350005, People’s Republic of China in hepatocellular carcinoma. Pathol. Int. 60,203–211 (2010). 17. Soung,Y. H.et al. CASPASE-8 gene is inactivated by somatic mutations in Conflict of interest gastric carcinomas. Cancer Res. 65, 815–821 (2005). The authors declare that they have no conflict of interest. 18. Kim, H. S. et al. Inactivating mutations of caspase-8 gene in colorectal carci- nomas. Gastroenterology 125,708–715 (2003). 19. Brynychová, V. et al. Clinical and functional importance of selected CASP8 andCASP9 polymorphismsinbreastcarcinoma. Klin. Onkol. 29,445–453 Publisher's note (2016). Springer Nature remains neutral with regard to jurisdictional claims in 20. Hashemi,M.et al. Bi-directional PCR allele-specificamplification (bi-PASA) for published maps and institutional affiliations. detection of caspase-8 -652 6N ins_del promoter polymorphism (rs3834129) in breast cancer. Gene 505,176–179 (2012). Supplementary Information accompanies this paper at https://doi.org/ 21. Tamm, I. et al. IAP-family protein Survivin inhibits caspase activity and apop- 10.1038/s41419-018-0511-3. tosis induced by Fas (CD95), Bax, caspases, and anticancer drugs. Cancer Res. 58,5315–5320 (1998). Received: 2 November 2017 Revised: 16 March 2018 Accepted: 19 March 22. Patel, V., Balakrishnan, K., Keating, M. J., Wierda, W. G. & Gandhi, V. Expression of executioner procaspases and their activation by a procaspase-activating compound in chronic lymphocytic leukemia cells. Blood 125,1126–1136 (2015). 23. Rathore, R., McCallum, J. E., Varghese, E., Florea, A. M. & Busselberg, D. Over- coming chemotherapy drug resistance by targeting inhibitors of apoptosis References proteins (IAPs). Apoptosis 22,898–919 (2017). 1. Siegel, R. L., Miller, K. D. & Jemal, A. Cancer statistics. CA Cancer J. Clin. 67,7–30 24. Valter, K. et al. Contrasting effects of glutamine deprivation on apoptosis (2017). induced by conventionally used anticancer drugs. Biochim. Biophys. Acta 1864, 2. Bose,P., Vachhani, P.&Cortes,J.E.Treatment of relapsed/refractoryacute 498–506 (2017). myeloid leukemia. Curr. Treat. Options Oncol. 18, 17 (2017). 25. Muzio,M., Stockwell, B. R.,Stennicke,H.R., Salvesen,G.S.&Dixit, V. M. An 3. Ofran, Y. & Rowe, J. M. Treatment for relapsed acute myeloid leukemia: what is induced proximity model for caspase-8 activation. J. Biol. Chem. 273, new? Curr. Opin. Hematol. 19,89–94 (2012). 2926–2930 (1998). 4. Rowe,J.M., Li,X.& Cassileth,P.A.VerypoorsurvivalofpatientswithAML who 26. Glaser, S. P. et al. Anti-apoptotic Mcl-1 is essential for the development and relapse after achieving a first complete remission: the Eastern Cooperative sustained growth of acute myeloid leukemia. Genes Dev. 26,120–125 (2012). Oncology Group Experience. Blood 106, 546a (2005). 27. Nakajima, H. & Kunimoto, H. TET2 as an epigenetic master regulator for normal 5. Konopleva, M. et al. Mechanisms of apoptosis sensitivity and resistance to the and malignant hematopoiesis. Cancer Sci. 105,1093–1099 (2014). BH3 mimetic ABT-737 in acute myeloid leukemia. Cancer Cell 10,375–388 28. Annesley, C. E. & Brown, P. The biology and targeting of FLT3 in pediatric (2006). leukemia. Front. Oncol. 4, 263 (2014). 6. Carter,B.Z.et al. Synergistic targeting of AML stem/progenitor cells with IAP 29. Kantarjian, H. Acute myeloid leukemia–majorprogressoverfourdecades and antagonist birinapant and demethylating agents. J. Natl. Cancer Inst. 106, glimpses into the future. Am.J.Hematol. 91,131–145 (2016). djt440 (2014). 30. Levis, M. FLT3 mutations in acute myeloid leukemia: what is the best 7. Kruidering, M. & Evan, G. I. Caspase-8 in apoptosis: the beginning of the end. approach in 2013? Hematol. Am. Soc. Hematol. Educ. Program 2013,220–226 Iubmb. Life 50,85–90 (2000). 8. Carrington, P. E. et al. The structure of FADD and its mode of interaction with (2013). 31. Falini, B. et al. Cytoplasmic nucleophosmin in acute myelogenous leukemia procaspase-8. Mol. Cell 22,599–610 (2006). with a normal karyotype. N. Engl. J. Med. 352,254–266 (2005). 9. Schug, Z. T., Gonzalvez, F., Houtkooper, R. H., Vaz, F. M. & Gottlieb, E. BID is 32. Suzuki, T. et al. Clinical characteristics and prognostic implications of NPM1 cleaved by caspase-8 within a native complex on the mitochondrial mem- mutations in acute myeloid leukemia. Blood 106,2854–2861 (2005). brane. Cell Death Differ. 18,538–548 (2011). 33. Kottaridis, P. D. et al. The presence of a FLT3 internal tandem duplication in 10. Hoffmann, J. C.,Pappa,A., Krammer, P. H. & Lavrik, I. N. A new C-terminal patients with acute myeloid leukemia (AML) adds important prognostic cleavage product of procaspase-8, p30, defines an alternative pathway of information to cytogenetic risk group and response to the first cycle of procaspase-8 activation. Mol. Cell. Biol. 29, 4431–4440 (2009). chemotherapy: analysis of 854 patients from the United Kingdom Medical 11. Kang, T. B. et al. Mutation of a self-processing site in caspase-8 compromises its Research Council AML 10 and 12 trials. Blood 98, 1752–1759 (2001). apoptotic but not its nonapoptotic functions in bacterial artificial chromo- 34. Walter, R. B. et al. Significance of FAB subclassification of “acute myeloid some transgenic mice. J. Immunol. 181,2522–2532 (2008). leukemia, NOS” in the2008WHO classification: analysis of 5848 newly diag- 12. Ando, M. et al. Cancer-associated missense mutations of caspase-8 activate nosed patients. Blood 121,2424–2431 (2013). nuclear factor-κB signaling. Cancer Sci. 104,1002–1008 (2013). 35. Fröhling, S. et al. CEBPA mutations in younger adults with acute myeloid 13. Li, C., Egloff, A. M., Sen, M., Grandis, J. R. & Johnson, D. E. Caspase-8 mutations in leukemia and normal cytogenetics: prognostic relevance and analysis of head and neck cancer confer resistance to death receptor-mediated apoptosis cooperating mutations. J. Clin. Oncol. 22,624–633 (2004). and enhance migration, invasion, and tumor growth. Mol. Oncol. 8,1220–1230 36. Bienz, M. et al. Risk assessment in patients with acute myeloid leukemia and a (2014). normal karyotype. Clin. Cancer Res. 11,1416–1424 (2005). 14. Liu,B., Peng,D., Lu,Y., Jin, W. &Fan,Z.Anovelsingleamino acid deletion 37. Boissel, N. et al. Prevalence, clinical profile, and prognosis of NPM mutations in caspase-8 mutant in cancer cells that lost proapoptotic activity. J. Biol. Chem. AML with normal karyotype. Blood 106,3618–3620 (2005). 277, 30159–30164 (2002). 38. Livak, K.J.&Schmittgen,T.D.Analysisofrelativegeneexpressiondata using 15. Qian, J. et al. Association between CASP8 and CASP10 polymorphisms and real-time quantitative PCR and the 2−ΔΔCT method. Methods 25,402–408 toxicity outcomes with platinum-based chemotherapy in Chinese patients (2001). with non-small cell lung cancer. Oncologist 17,1551–1561 (2012). Official journal of the Cell Death Differentiation Association

Journal

Cell Death & DiseaseSpringer Journals

Published: May 3, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off