Induction of Apoptosis by ASK1, a Mammalian MAPKKK That Activates SAPK/JNK and p38 Signaling Pathways

See allHide authors and affiliations

Science  03 Jan 1997:
Vol. 275, Issue 5296, pp. 90-94
DOI: 10.1126/science.275.5296.90


Mitogen-activated protein (MAP) kinase cascades are activated in response to various extracellular stimuli, including growth factors and environmental stresses. A MAP kinase kinase kinase (MAPKKK), termed ASK1, was identified that activated two different subgroups of MAP kinase kinases (MAPKK), SEK1 (or MKK4) and MKK3/MAPKK6 (or MKK6), which in turn activated stress-activated protein kinase (SAPK, also known as JNK; c-Jun amino-terminal kinase) and p38 subgroups of MAP kinases, respectively. Overexpression of ASK1 induced apoptotic cell death, and ASK1 was activated in cells treated with tumor necrosis factor-α (TNF-α). Moreover, TNF-α-induced apoptosis was inhibited by a catalytically inactive form of ASK1. ASK1 may be a key element in the mechanism of stress- and cytokine-induced apoptosis.

The MAP kinase signaling cascade, a signal transduction pathway well conserved in cells from yeasts to vertebrates, consists of three distinct members of the protein kinase family, including MAP kinase (MAPK), MAPK kinase (MAPKK), and MAPKK kinase (MAPKKK) (1). MAPKKK phosphorylates and thereby activates MAPKK, and the activated form of MAPKK in turn phosphorylates and activates MAPK. Activated MAPK may translocate to the cell nucleus and regulate the activities of transcription factors and thereby control gene expression (1). At least two defined MAPK signaling modules function in mammalian cells: the Raf-MAPKK-MAPK and the MEKK-SEK1 (or MKK4)-SAPK (or JNK) pathways (2, 3, 4, 5, 6, 7). MKK3/MAPKK6 (or MKK6, a close relative of MKK3) that corresponds to MAPKK and p38 MAPK form another MAPK signaling unit (2, 8); however, the biological consequence and mechanism of activation of the p38 signaling cascade are poorly understood. We identified a mammalian MAPKKK that activates the MKK3/MAPKK6-p38 as well as the SEK1-SAPK signaling pathways.

We used a degenerate polymerase chain reaction (PCR)-based strategy to identify serine-threonine kinases (9). One PCR fragment, obtained with a set of PCR primers oriented from the conserved subdomains VI and VIII of the serine-threonine kinase family (10), was used to isolate a corresponding nearly full-length cDNA clone (11). Because of its characteristics, we refer to the putative serine-threonine kinase encoded by this cDNA as apoptosis signal- regulating kinase (ASK)1. The deduced amino acid sequence revealed that ASK1 consists of 1375 amino acids with a calculated relative molecular mass of 154,715 (Fig. 1A). Another clone that is truncated in the NH2-terminal 374 amino acids was also obtained (clone 27) (12). The serine-threonine kinase domain of ASK1 is in the middle part of the molecule and has long NH2- and COOH-terminal flanking sequences. RNA blot analysis revealed a single 5-kb transcript that was expressed in all human tissues tested (Fig. 1B).

Fig. 1.

Primary structure of ASK1 and complementation of SSK2/SSK22/SHO1 deficiency in S. cerevisiae. (A) Predicted amino acid sequence of human ASK1. The putative translation start sites for two independent cDNA clones, clone 20 and clone 27, are indicated by arrows. The protein kinase domain is shown in bold-face. An FKBP-type peptidyl-prolyl cis-trans isomerase motif present in the NH2-terminal noncatalytic portion is underlined. The cDNA sequence of ASK1 has been deposited in EMBL-GenBank data library (accession number D84476). Abbreviations for the amino acid residues are as follows: A, Ala; C, Cys; D, Asp; E, Glu; F, Phe; G, Gly; H, His; I, Ile; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, Gln; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr. (B) Tissue distribution of ASK1. Immunoblots with mRNAs from various human tissues (Clontech) were probed with ASK1 cDNA labeled by random priming. (C) Complementation of SSK2/SSK22/SHO1 triple mutant strain in S. cervisiae. Complementary DNA fragments encoding ASK1 were subcloned into the yeast expression vector pNV11 and transformed into TM257-H1 yeast strain (13), which is defective in SSK2, SSK22, and SHO1. Five independent transformants were selected and grown on YPD plates in the presence or absence of 1.5 M sorbitol. Photographs were taken after 6 days at 30°C. Transformants with pNV11 vector alone or ASK1(K709R) were also tested.

A database search of the ASK1 sequence outside its kinase domain showed that a short amino acid sequence in the NH2-terminal part contains a motif for an FK506-binding protein (FKBP)-type peptidyl-prolyl cis-trans isomerase, of which the functional importance is unknown (Fig. 1A). The kinase domain of ASK1 has sequence similarity with members of the MAPKKK family including MEKK1 (30.0%) in mammal and SSK2 (32.3%) and STE11 (30.4%) in Saccharomyces cerevisiae. Phylogenetic comparison suggested that ASK1 is distantly related to RAF-1, KSR1, TAK1, and TPL-2 mammalian MAPKKKs but most closely related to the SSK2 or SSK22 family of yeast MAPKKKs, which are upstream regulators of yeast HOG1 MAPK (13).

Despite differences in the overall structures of ASK1 and SSK2 or SSK22 (13), it was of interest to examine whether ASK1 might act as a functional kinase in yeast and thereby complement the loss of SSK2 or SSK22. We used yeast strain TM257-H1 (ssk2Δ ssk22Δ sho1Δ) (13, 14), which grows in a normal YPD medium but not in hyperosmotic medium. Transformants were tested for growth in the presence of 1.5 M sorbitol (Fig. 1C). Expression of ASK1, but neither vector alone nor ASK1(K709R, a catalytically inactive mutant in which Lys709 was changed to Arg), complemented TM257-H1 growth in this hyperosmotic environment (15). These observations, together with the fact that the mammalian counterpart of yeast HOG1 is p38 MAP kinase (16, 17, 18), suggested that ASK1 may be a mammalian MAPKKK that functions in a signaling cascade to activate MKK3/MAPKK6 and p38.

To investigate whether ASK1 may function as a MAPKKK in mammalian cells, we transfected ASK1 into COS7 cells together with MAPK and MAPKK expression plasmids (Fig. 2A). All the MAPK and MAPKK constructs were hemagglutinin (HA) epitope-tagged, expressed with or without ASK1, and immunoprecipitated with antibody to HA (anti-HA) (19); the immune complexes were subjected to a kinase assay with exogenous substrates (20). ASK1 expression induced 7.6- and 5.0-fold activation of SAPK and p38 MAPKs, respectively, but only weakly activated MAPK. ASK1 activated SEK1 and MKK3 up to 7.0- and 11.8-fold, respectively, whereas no detectable activation of MAPKK was observed. To investigate whether the SEK1 and MKK3 were directly phosphorylated and activated by ASK1, we used an in vitro-coupled kinase assay (21) with recombinant SEK1, MKK3, MAPKK6, and recombinant catalytically inactive p38 (MPK2) proteins (Fig. 2B). ASK1 immunoprecipitates from COS7 cells activated SEK1, MKK3, and MAPKK6 (greater than 40-fold for each) (Fig. 2B), and phosphorylation of p38 was observed only when ASK1 was present in the kinase assay. ASK1-dependent phosphorylation of p38 was further confirmed to result in the activation of p38 in a coupled kinase assay with wild-type p38 and ATF2 (21) (Fig. 2C). In contrast, ASK1 weakly activated MAPKK (2.2-fold) (Fig. 2B). When Raf was used as MAPKKK for a positive control, a 27.8-fold activation of MAPKK was observed (22). These results indicate that ASK1 selectively activates the SEK1-SAPK and MKK3/MAPKK6-p38 pathways.

Fig. 2.

Activation of MAPKKs and MAPKs by ASK1. (A) In vivo activation of MAPKKs and MAPKs by ASK1. COS7 cells were transiently transfected with pSRα-HA1-MAPK, pSRα-HA1-SAPK, pSRα-HA1-p38, pSRα-HA1-MAPKK, pSRα-HA1-SEK1, or pSRα-HA1-MKK3 together with pcDNA3-ASK1 (+) or a control vector pcDNA3 (−). After 18 hours, cells were lysed and MAPKKs and MAPKs were immunoprecipitated with anti-HA. The kinase activity was measured in an immune complex kinase assay with exogenous substrates (20). The position of each substrate is shown by a bracket or an arrowhead. The fold increase of kinase activity caused by coexpression of ASK1 is indicated above each lane and is the average of three independent experiments. Molecular sizes are indicated on the left in kilodaltons. (B) In vitro activation of MAPKKs by ASK1. COS7 cells were transiently transfected with pcDNA3-ASK1, and proteins were immunoprecipitated from cell lysates with preimmune serum (Mock ppt) or antiserum to ASK1 (ASK1 ppt). The immune complex or a buffer solution (Buffer) was first incubated with (+) or without (−) His-MAPKK, His-SEK1, His-MKK3, or His-MAPKK6, and then the kinase activity of these MAPKKs was measured with the substrate GST-catalytically inactive MAPK for MAPKK and His-catalytically inactive p38 for SEK1, MKK3, and MAPKK6. (C) The ASK1 immune complex was incubated with (+) or without (−) the indicated His-MAPKKs and His-wild-type p38, and the kinase activity of p38 was measured with the substrate ATF2.

We determined the biological activity of ASK1 in mink lung epithelial (Mv1Lu) cell lines that were stably transfected with metallothionein promoter-based expression plasmids (23). Immunoprecipitation of ASK1 with antiserum to ASK1 (24) after metabolic labeling of the cells revealed that ASK1 was highly expressed only when induced by ZnCl2 (Fig. 3A). ASK1(K709R)-transfected cells expressed the recombinant protein in similar amounts. SAPK or p38 MAPK signaling cascade or both may participate in signaling pathways leading to apoptosis (25, 26, 27, 28). To investigate the effects of ASK1 on cellular growth or viability, we measured [3H]thymidine incorporation in the stable transfectants. Inhibition of [3H]thymidine incorporation was observed in the cells transfected with ASK1 but not in cells transfected with the vector alone or ASK1(K709R) (Fig. 3B). The dose-dependent inhibition of [3H]thymidine incorporation by ZnCl2 correlated well with the dose-dependent expression and activation of ASK1 (Fig. 3C). Furthermore, endogenous SAPK and p38 were activated (29) in parallel with the ASK1 activities (Fig. 3D).

Fig. 3.

Induction of apoptosis by ASK1 in stably transfected Mv1Lu cells. (A) ZnCl2-dependent expression of ASK1 and ASK1(K709R) in stably transfected Mv1Lu cells. Cells were metabolically labeled with a mixture of [35S]methionine and [35S]cysteine in the presence or absence of 100 μM ZnCl2 for 5 hours, and the cellular lysates were subjected to immunoprecipitation as described (33) with antiserum to ASK1, SDS-PAGE, and fluorography. (B) ASK1-dependent inhibition of [3H]thymidine incorporation. Mv1Lu cells stably transfected with vector alone (▪), ASK1 (•), and ASK1(K709R) (ˆ) were incubated in minimal essential medium (MEM) containing fetal bovine serum (FBS) (1%) with the indicated concentration of ZnCl2 for 16 hours. The cells were then incubated with [3H]thymidine for 1 hour, and 3H incorporation into the DNA was determined. The average of the counts from duplicate wells is shown as the percentage of control. Data shown are the representatives of three independent experiments. Error bars represent the SD. (C) Dose-dependent expression (top) and activation (bottom) of ASK1 by ZnCl2 induction. (Top) ASK1 was immunoprecipitated as described in (A) after the ASK1-transfected Mv1Lu cells were induced by the indicated amount of ZnCl2 for 5 hours. (Bottom) Immunoprecipitated ASK1 was subjected to the MKK3-p38 coupled kinase assay. (D) ASK1-dependent activation of endogenous SAPK (top) and p38 (bottom). Mv1Lu cells stably transfected with ASK1 were incubated with the indicated concentration of ZnCl2 for 5 hours. The cell extracts were assayed for SAPK activity (29) with gel-immobilized c-Jun as a substrate (top), or for p38 activity in an immune complex kinase assay (29) with a polyclonal antibody to p38 with His-ATF2 as a substrate (bottom). The positions of p54 and p46 SAPKs (top) and ATF2 (bottom) are indicated by arrowheads and an arrow, respectively. (E) ASK1-dependent cell death. (Top) Cells were incubated with MEM containing FBS (1%) in the presence (+) or absence (−) of 100 μM ZnCl2 for 26 hours. Representative cell morphology was determined by phase-contrast microscopy (original magnification, ×33). (Bottom) Cells were incubated in MEM without FBS in the presence or absence of 100 μM ZnCl2 for 25 hours and stained by the TUNEL method with an in situ cell death detection kit (Boehringer Mannheim). Apoptotic cells are shown by dark brown staining. Photographs were taken at a higher magnification (×82.5) than in the top panel.

Induction of expression of ASK1 but not ASK1(K709R) induced cytoplasmic shrinkage and nuclear condensation, which became apparent within 6 hours after addition of ZnCl2 (30). The typical morphological properties of apoptotic cells became most evident after more than 24 hours (Fig. 3E). The apoptotic cell death induced by ASK1 was confirmed by an in situ staining of cells with the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) method (Fig. 3E) as well as by genomic DNA fragmentation (30). Taken together, these observations indicate that overexpression of ASK1 results in apoptosis.

Tumor necrosis factor-α (TNF-α) causes apoptosis and activates both SAPK and p38 signaling systems (4, 31). We therefore examined whether treatment of cells with TNF-α resulted in the activation of ASK1. ASK1 immunoprecipitates from TNF-α-treated cells were subjected to a coupled kinase assay with MKK3 and catalytically inactive p38 (Fig. 4, A and B). Treatment of cells with TNF-α activated the ASK1 in ASK1-transfected Mv1Lu cells within 5 min (Fig. 4A) in a dose-dependent manner (Fig. 4B). In addition, endogenous ASK1 was also activated by TNF-α in other cell types in which apoptosis is induced by TNF-α, including human 293 embryonal kidney cells, A673 rhabdomyosarcoma cells (Fig. 4A), Jurkat T cells, and KB epidermal carcinoma cells (22). Furthermore, when ASK1(K709R) was transiently transfected into 293 cells (transfection efficiency was greater than 60% as determined by β-galactosidase staining) (Fig. 4C) or Jurkat cells (32) (Fig. 4D), DNA fragmentation induced by TNF-α was reduced, suggesting that ASK1(K709R) acts as a dominant-negative mutant, and that ASK1 is necessary for the TNF-α-induced apoptotic response.

Fig. 4.

TNF-α-induced activation of ASK1 and dominant-negative effect of ASK1(K709R) on TNF-α-induced apoptosis. Mv1Lu cells stably transfected with ASK1 were first treated with 50 μM ZnCl2 for 5 hours and then stimulated with TNF-α (100 ng/ml) for the indicated time (A) (top) or with the indicated concentration of TNF-α for 30 min (B). The kinase activity of ASK1 was measured (21) in arbitrary units. The results are means from at least five independent experiments. Error bars represent the SD. (A) (bottom) ASK1 activity from ASK1-transfected Mv1Lu cells and nontransfected 293 and A673 cells treated with TNF-α (100 ng/ml). ASK1(K709R)-dependent inhibition of DNA fragmentation in TNF-α-treated 293 (C) and Jurkat (D) cells. 293 cells (2 × 106) were transiently transfected with 2 μg of pcDNA3 control vector or pcDNA3-ASK1(K709R) by the use of Tfx-50 (Promega) according to the manufacturer's protocol. Eight hours after transfection, cells were treated with TNF-α (100 ng/ml) with or without 300 nM actinomycin D (ActD, which enhanced TNF-α-induced apoptosis) for 16 hours. Apoptotic cells detached from culture plate were collected, and total DNA was isolated and analyzed by 2% agarose gel electrophoresis. This experiment was repeated more than five times with similar results. In (D), 3 × 106 nontransfected Jurkat cells or selectively isolated populations of ASK1(K709R)-expressing Jurkat cells (32) were treated with the indicated concentrations of TNF-α for 5.5 hours. Cytoplasmic DNA was extracted (35) and analyzed by 2% agarose gel electrophoresis.

Our results suggest that proinflammatory cytokines may activate ASK1, which contributes to cellular apoptosis, possibly by activating SEK1-SAPK and MKK3/MAPKK6-p38 signaling cascades. SAPK and p38 signaling pathways are both indispensable for apoptosis triggered by nerve growth factor withdrawal in differentiated neuronal cells (25). However, a possibility that ASK1-induced apoptosis in our system is mediated by effector pathways other than SAPK and p38 is not excluded. ASK1 might be a constitutively active kinase because ASK1 that was transiently expressed in COS7 cells or induced by ZnCl2 in Mv1Lu cells was active in vitro and in vivo. However, overexpression of ASK1 might result in activation, or the transfection procedure or the treatment with a heavy metal may subject the cells to stress, and thereby activate ASK1. Thus, identification of regulatory mechanisms of ASK1 should provide insights into signaling pathways leading to apoptosis.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
View Abstract

Navigate This Article