Molecular Analysis of the Interaction between Human PTPN21 and the Oncoprotein E7 from Human Papillomavirus Genotype 18

Human papillomaviruses (HPVs) cause cellular hyperproliferation-associated abnormalities including cervical cancer. The HPV genome encodes two major viral oncoproteins, E6 and E7, which recruit various host proteins by direct interaction for proteasomal degradation. Recently, we reported the structure of HPV18 E7 conserved region 3 (CR3) bound to the protein tyrosine phosphatase (PTP) domain of PTPN14, a well-defined tumor suppressor, and found that this intermolecular interaction plays a key role in E7-driven transformation and tumorigenesis. In this study, we carried out a molecular analysis of the interaction between CR3 of HPV18 E7 and the PTP domain of PTPN21, a PTP protein that shares high sequence homology with PTPN14 but is putatively oncogenic rather than tumor-suppressive. Through the combined use of biochemical tools, we verified that HPV18 E7 and PTPN21 form a 2:2 complex, with a dissociation constant of 5 nM and a nearly identical binding manner with the HPV18 E7 and PTPN14 complex. Nevertheless, despite the structural similarities, the biological consequences of the E7 interaction were found to differ between the two PTP proteins. Unlike PTPN14, PTPN21 did not appear to be subjected to proteasomal degradation in HPV18-positive HeLa cervical cancer cells. Moreover, knockdown of PTPN21 led to retardation of the migration/invasion of HeLa cells and HPV18 E7-expressing HaCaT keratinocytes, which reflects its protumor activity. In conclusion, the associations of the viral oncoprotein E7 with PTPN14 and PTPN21 are similar at the molecular level but play different physiological roles.     

Influence of Human Papillomavirus E7 Oncoprotein on Maturation and Function of Plasmacytoid Dendritic Cells In Vitro

The major difficulties of human papillomavirus(HPV) treatment are its persistence and recurrence. The HPV E7 oncoprotein-loaded dendritic cells have been evaluated as cellular vaccine in previous reports. Plasmacytoid dendritic cells(pDCs) play an essential role of connecting the innate immune response and adaptive immune response in the immune system. But they function in HPV E7 loading is unclear. To investigate whether loading of the HPV type 6b, 11, and 16 E7 proteins affects the activity of pDCs, human peripheral blood-separated pDCs and mouse bone marrow-derived pDCs were pulsed with the HPV E7 proteins. The expression levels of CD40, CD80, CD86, and MHC II were significantly upregulated in pDCs upon HPV 6b/11 E7 protein pulse. The secretion and gene expression of type I IFN and IL-6 were both upregulated by HPV 6b/11 E7 proteins, more significant than HPV 16 E7 protein. The expression of essential factors of TLR signaling pathway and JNK/p38 MAP kinase signaling pathway were all increased in HPV 6b/11 E7 proteins pulsed pDCs. Our results suggest that HPV E7 proteins could promote the differentiation and maturation of pDCs and activate the TLR and MAPK pathway to induce host innate immune response. It might be conducive to explore novel immunotherapy targeting HPV infection with HPV E7 loaded pDC.     

Human Papillomavirus Type 16 E7 Oncoprotein Expressed in Peripheral Epithelium Tolerizes E7-Directed Cytotoxic T-Lymphocyte Precursors Restricted through Human (and Mouse) Major Histocompatibility Complex Class I Alleles

Mice which coexpress human papillomavirus type 16 E7 and HLA A2.1 in peripheral squamous epithelium and thymic cortical epithelium are tolerant at the cytotoxic T-lymphocyte (CTL) level to E7 epitopes restricted through HLA A*0201 and H-2^b (T. Doan, M. Chambers, M. Street, G. J. Fernando, K. Herd, P. Lambert, and R. Tindle, Virology 244:352–364, 1998). Here we used bone marrow-reconstituted radiation chimeras to distinguish whether E7-directed CTL tolerance was mediated peripherally by E7 expression in skin or centrally by E7 expression in thymus. In chimeric mice expressing E7 in skin and reconstituted with E7-naïve bone marrow and E7-naïve thymus, CTL responses to vaccine-administered E7 epitopes were not restored, i.e., the mice remained tolerant. In contrast, chimeric mice not expressing E7 in skin and reconstituted with E7-naïve bone marrow and E7-expressing thymus had full E7-directed CTL responses. These results demonstrate that E7 protein expression in peripheral squamous epithelium is sufficient to tolerize the E7-directed CTL precursor repertoire. The data have implications for E7-mediated tumorigenesis and for the development of E7-based immunotherapeutic strategies, since peripheral immunological tolerance of tumor-associated antigens may create a barrier to effective immunotherapy.The E7 oncoprotein of human papillomavirus type 16 (HPV16) is a tumor-specific antigen when expressed in HPV16-associated cervical epithelial tumors, to which immunomanipulative strategies are being directed, both experimentally (see, for example, references 7, 9, and 29) and in E7-based therapeutic vaccine clinical trials (5). We recently reported studies with mice expressing HPV16 E7 protein, driven from the keratin 14 (K14) promoter, in basal epithelium of skin and in the thymic cortex (8). We showed that immunization-induced cytotoxic T-lymphocyte (CTL) responses to each of three CTL epitopes in the E7 protein restricted through two major histocompatibility complex (MHC) class 1 haplotypes were down-regulated in these E7-transgenic mice compared with non-E7 syngeneic control mice. However, in these studies we did not determine whether the down-regulation (i.e., tolerance) was induced centrally by E7 expressed in the thymus or peripherally by E7 expressed in epithelium. In the present study, we distinguish between these two possibilities by specific immunization of bone marrow-reconstituted thymus-transplanted chimeric E7 transgenic mice. We report that chimeric mice expressing the E7 transgene in peripheral epithelium but not in the thymus showed E7-specific down-regulated CTL responses to each of two E7 CTL epitopes restricted through a human and a mouse MHC class I allele, respectively, when compared with sham chimeric but non-E7 control mice. In contrast, chimeric mice expressing the E7 transgene in thymus, but not peripheral epithelium, showed E7-directed CTL responses indistinguishable from those of non-E7 control mice. Thus, we show that the expression of E7 in peripheral squamous epithelium is sufficient to induce and maintain a state of tolerance against E7.

E7-directed bone marrow-derived precursor CTLs (pCTLs) are not tolerized in mice expressing E7 in thymus but not in skin.

(K14E7 × A2.1K^b)F₁ mice (designated KA mice) were derived by crossing male K14.HPV16E7(+/+) mice (16), which express an HPV16 E7 transgene perinatally and throughout life in skin and thymic cortical epithelium, with female HLA A2.1K^b(+/+) mice (30). (FVB × A2.1K^b)F₁ mice (designated FA) are syngeneic but do not possess the E7 transgene. KA (E7⁺) and FA (E7⁻) mice are on an H-2^b background. To inquire whether pCTLs from E7-transgenic mice were tolerized on E7-expressing thymus, we constructed thymus-transplanted radiation chimeras as described elsewhere (8) from immunologically depleted FA (E7⁻) mice reconstituted with KA (E7⁺) bone marrow cells. In half of the mice, [designated KA→FA(FA) mice], the bone marrow-derived T-cell precursors were made to mature through a thymus implant from an FA (E7⁻) donor mouse; in the other half of the mice [designated KA→FA(KA) mice], the bone marrow cells were made to mature through an E7-expressing KA thymus implant (Fig. ​(Fig.1,1, panel I). KA→FA(FA) mice, KA→FA(KA) mice, and control FA (E7⁻) and KA (E7⁺) mice were immunized for CTL response induction with a mix of peptides containing ⁸²LLMGTLGIV⁹⁰ (an HLA A*0201-restricted E7 CTL epitope [24]), ⁴⁹RAHYNIVTF⁵⁷ (an H-2D^b-restricted E7 CTL epitope [9]), and ⁵⁸GILGFVFTL⁶⁶ (an HLA A*0201-restricted influenza virus matrix CTL epitope [13]). Control mice underwent surgical procedures but without receiving cell and/or organ transplants (sham). KA (E7⁺) mice showed the previously documented (8) down-regulated CTL response to the E7 epitopes (but not to the irrelevant influenza virus matrix epitope) compared to FA (E7⁻) mice (Fig. ​(Fig.1,1, panels IIC and IID). In contrast, KA→FA(KA) mice exhibited E7 (and influenza virus matrix)-directed CTL responses of the same magnitude as those of KA→FA(FA) mice and FA (E7⁻) mice (Fig. ​(Fig.1,1, panels IIA to IIC). These data indicate that E7-directed pCTLs from E7 transgenic mice which mature through an E7-expressing thymus, and emerge into a non-E7-expressing peripheral epithelial environment, are not tolerized. Open in a separate windowFIG. 1(I) Derivation of KA→FA(FA) and KA→FA(KA) chimeric mice from immunologically ablated FA (E7⁻) mice. (II) CTL responses of splenocytes from chimeric mice and sham control FA (E7⁻) and KA (E7⁺) mice (three per group) immunized with a mix of peptides containing E7 CTL epitopes LLMGTLGIV and RAHYNIVTF and influenza virus matrix CTL epitope GILGFVFTL. Immunizations were given in Quil A adjuvant and tetanus toxoid as described elsewhere (8). Spleen cells were restimulated with individual peptides in vitro. Targets were EL4.A2 cells (8) pulsed with individual peptides as indicated. EL4.A2 cells are susceptible to specific CTL lysis through both HLA A*0201 and H-2^b restriction elements. CTL assays were conducted as described elsewhere (8). bm, bone marrow; th, thymus; sk, skin.To eliminate the possibility that bone marrow-derived precursors from KA (E7⁺) mice had somehow previously encountered E7 protein before transfer to recipient mice, thereby influencing their immunological status in the above-described experiment, we asked whether bone marrow-derived pCTLs from FA (E7⁻) mice would be tolerized during maturation in an E7-expressing thymus. We constructed chimeras from immunologically ablated FA (E7⁻) mice by reconstitution with FA (E7⁻) bone marrow cells. In half the mice [designated FA→FA(KA) mice], the bone marrow cells were made to mature through an E7-expressing KA thymus implant. In the other half of the mice [designated FA→FA(FA) mice], the bone marrow cells were made to mature through a non-E7-expressing FA thymus implant (Fig. ​(Fig.2,2, panel I). FA→FA(KA) mice, FA→FA(FA) mice, and control FA (E7⁻) and KA (E7⁺) mice were immunized for CTL induction with a mix of peptides containing the HLA A*0201-restricted and H-2^b-restricted E7 CTL epitopes and influenza virus matrix CTL epitope. FA→FA(KA) mice exhibited E7-directed CTL responses of the same magnitude as those of FA→FA(FA) mice and FA (E7⁻) mice, while KA (E7⁺) mice exhibited the expected down-regulated E7-directed (but not down-regulated influenza virus matrix-directed) CTL responses (Fig. ​(Fig.2,2, panel II). These data indicated that E7-naïve bone marrow-derived pCTLs which mature through an E7-expressing thymus and emerge into a non-E7 peripheral epithelial environment are not tolerized to E7. Open in a separate windowFIG. 2(I) Derivation of FA→FA(FA) and FA→FA(KA) chimeric mice from immunologically ablated FA (E7⁻) mice. (II) CTL responses of splenocytes from chimeric mice and sham control mice (three per group) immunized with a mix of peptides containing E7 CTL epitopes LLMGTLGIV and RAHYNIVTF and influenza virus matrix CTL epitope GILGFVFTL. Spleen cells were restimulated with individual peptides in vitro. Targets were EL4.A2 cells pulsed with individual peptides as indicated. CTL assays were conducted as described elsewhere (8). bm, bone marrow; th, thymus; sk, skin.

Bone marrow-derived pCTLs are specifically tolerized in mice expressing E7 in skin but not in thymus.

To inquire whether E7-directed pCTLs were tolerized in mice expressing E7 in skin but not in thymus, we constructed chimeric mice in which bone marrow-derived precursors were made to mature through a non-E7-expressing thymus and to emerge into an E7-expressing peripheral epithelial environment. In a first experiment, immunologically ablated KA (E7⁺) mice were reconstituted with KA (E7⁺) bone marrow cells which were made to mature through a thymus implanted from an FA (E7⁻) mouse. The recipient mice, designated KA→KA(FA) (Fig. ​(Fig.3,3, panel I), and control FA (E7⁻) and KA (E7⁺) mice were immunized for CTL response induction with a mix of peptides containing the HLA A*0201-restricted and H-2^b-restricted E7 CTL epitopes and the influenza virus matrix CTL epitope. In KA→KA(FA) mice, E7-directed CTL responses to both E7 epitopes were down-regulated to the level seen in control KA (E7⁺) mice (Fig. ​(Fig.3,3, panel IIB), while control FA (E7⁻) mice showed the expected high responses to both E7 CTL epitopes. Open in a separate windowFIG. 3(I) Derivation of KA→KA(FA) chimeric mice from immunologically ablated KA (E7⁺) mice. (II) CTL responses of splenocytes from chimeric mice and sham control mice (three per group) immunized with a mix of peptides containing E7 CTL epitopes LLMGTLGIV and RAHYNIVTF and influenza virus matrix CTL epitope GILGFVFTL. Spleen cells were restimulated with individual peptides in vitro. Targets were EL4.A2 cells pulsed with individual peptides as indicated. CTL assays were conducted as described elsewhere (8). bm, bone marrow; th, thymus; sk, skin.In a second experiment, immunologically ablated KA (E7⁺) mice were reconstituted with bone marrow from FA (E7⁻) mice, which was made to mature through a non-E7-expressing FA thymus. These mice, designated FA→KA(FA) mice (Fig. ​(Fig.4,4, panel I), were immunized for CTL response induction with a mix of peptides containing the HLA A*0201-restricted and the H-2^b-restricted E7 CTL epitopes and the influenza virus matrix CTL epitope. As with KA→KA(FA) mice in the previous experiment, E7-directed CTL responses to both E7 epitopes were down-regulated as in KA (E7⁺) controls and in contrast to FA (E7⁻) controls, while the influenza virus matrix response confirmed adequate reconstitution. Open in a separate windowFIG. 4(I) Derivation of FA→KA(FA) chimeric mice from immunologically ablated KA (E7⁺) mice. (II) CTL responses of splenocytes from chimeric mice and sham control mice (three per group) immunized once with a mix of peptides containing E7 CTL epitopes LLMGTLGIV and RAHYNIVTF and influenza virus matrix CTL epitope GILGFVFTL. Spleen cells were restimulated with individual peptides in vitro. Targets were EL4.A2 cells pulsed with individual peptides as indicated. CTL assays were conducted as described elsewhere (8). bm, bone marrow; th, thymus; sk, skin.The results from these two experiments indicate that bone marrow-derived E7-directed pCTLs which mature through a non-E7-expressing thymus and emerge into an E7-expressing epithelial environment are specifically tolerized to E7.We have previously reported pCTL tolerance to epitopes of the HPV16 E7 oncoprotein in KA mice expressing a K14 promoter-driven E7 transgene perinatally and throughout life in the thymus and in basal and/or suprabasal cells of peripheral epithelium (8). In the present experiments, we demonstrate that the E7-directed pCTL repertoire is tolerized in mice expressing E7 in peripheral epithelium in the absence of thymic expression. Conversely, the repertoire is not tolerized in mice expressing E7 in the thymus, in the absence of E7 expression in peripheral epithelium. These data indicate that expression of E7 in peripheral epithelium, and not the thymus, is sufficient to induce and maintain a state of pCTL tolerance to E7. In the thymus, the K14 promoter directs transgene expression to the cortical epithelial compartment (19), which, in other mouse models, has been shown to contribute to the shaping of the T-cell repertoire by positive rather than negative selection (for example, see reference 20). Melero et al. (21) observed no functional down-regulation of the CTL responses induced by immunization with E7 peptide epitope RAHYNIVTF in H-2^b mice expressing HPV16 E7 from a K14 promoter and concluded that the mice remain immunologically ignorant of this epitope. This result contrasts with ours. Together, they provide further examples of T-cell tolerance to peripheral antigens in some systems (see, for example, references 1, 2, and 23) and T-cell ignorance in others (see, for example, references 14, 17, and 25). While determinants of immunological outcome of peripheral antigen expression are clearly complex (22), the level of expression (as well as timing and site of expression) can determine whether an antigen induces tolerance or is ignored by naïve T cells. This consideration may explain the difference between the results of Melero et al. and ours. The effect of the level of E7 expression on peripheral tolerance induction is under investigation in our laboratory.Specific CTL tolerance has implications for E7-mediated tumorigenesis. Nascent E7-expressing tumor cells will escape surveillance where little or no positive priming of cognate pCTLs by endogenous E7 occurs. Additionally, specific CTL tolerance which inhibits the generation of an immunization-induced CTL response will detract from effective immunotherapy (26). We have previously reported that (K14.E7 × C57)F₁ mice fail to control a challenge with an E7-expressing tumor following immunization with E7 CTL epitope RAHYNIVTF, whereas in immunized non-E7-transgenic control mice the tumors did not become established (12). Failure to control the tumor was correlated with a lack of an inducible RAHYNIVTF-directed CTL response in E7-transgenic mice, in contrast to non-E7-transgenic control mice, where a powerful CTL response was observed. In further experiments, multiple immunization of KA mice with E7 CTL epitopes or whole E7 protein failed to arrest the development of E7-associated endogenous tumors (8), again being correlated with a lack of E7-directed CTL responses.The current therapeutic vaccine strategy for HPV16-associated cervical carcinoma targets the E7 tumor-specific antigen by CTL induction (5, 28). The possibility arises that chronic expression of E7 in transformed cervical epithelial cells during the life of the tumor functionally tolerizes E7-directed pCTLs.Ongoing experiments in our laboratory will distinguish between presentation of E7 to pCTLs directly by keratinocytes and cross presentation of exogenously acquired E7 by bone marrow-derived professional antigen-presenting cells. Presentation of antigen by either of these routes can be tolerogenic (3, 6, 15, 27). Additionally, we will determine whether loss of functional E7-directed CTLs results from pCTL deletion (4, 11) or anergy (10, 25). Peripheral tolerance of tissue-specific antigen depends, at least in some cases, on the generation of regulatory CD4⁺ cells (see, for example, reference 18). That E7-directed CTL tolerance in KA (E7⁺) mice reflects an impairment of cognate CD4⁺ help is unlikely in view of our finding that (K14.E7 × C57)F₁ mice immunized with full-length E7 and displaying E7-specific pCTL tolerance showed concomitant enhanced E7-directed CD4⁺ T-helper responses (12).The data reported in the present study demonstrate the induction of peripheral tolerance in E7-directed pCTLs by HPV16 E7 expressed in squamous epithelial cells, in the context of human (and mouse) MHC class 1 haplotypes. There are direct implications for the development and progression of cervical cancers which express E7 in transformed squamous epithelium and for the design of E7-based immunotherapeutic strategies for cervical cancer. In the broader context, there are implications for CTL response induction to any foreign or aberrant protein expressed constitutively in squamous epithelial cells as a result of infection, tumorigenesis, or appearance of autoantigen.     

Degradation of the Retinoblastoma Tumor Suppressor by the Human Papillomavirus Type 16 E7 Oncoprotein Is Important for Functional Inactivation and Is Separable from Proteasomal Degradation of E7

The steady-state level and metabolic half-life of retinoblastoma tumor suppressor protein pRB are decreased in cells that express high-risk human papillomavirus (HPV) E7 proteins. Here we show that pRB degradation is a direct activity of E7 and does not reflect a property of cell lines acquired during the selection process for E7 expression. An amino-terminal domain of E7 that does not directly contribute to pRB binding but is required for transformation is also necessary for E7-mediated pRB degradation. Treatment with inhibitors of the 26S proteasome not only blocks E7-mediated pRB degradation but also causes the stabilization of E7. Mutagenic analyses, however, reveal that the processes of proteasomal degradation of E7 and pRB are not linked processes. HPV type 16 E7 also targets the pRB-related proteins p107 and p130 for destabilization by a proteasome-dependent mechanism. Using the SAOS2 flat-cell assay as a biological indicator for pRB function, we demonstrate that pRB degradation, not solely binding, is important for the E7-induced inactivation of pRB.     

Papillomavirus E6 PDZ Interactions Can Be Replaced by Repression of p53 To Promote Episomal Human Papillomavirus Genome Maintenance

Cancer-associated human papillomaviruses (HPVs) express E6 oncoproteins that target the degradation of p53 and have a carboxy-terminal PDZ ligand that is required for stable episomal maintenance of the HPV genome. We find that the E6 PDZ ligand can be deleted and the HPV genome stably maintained if cellular p53 is inactivated. This indicates that the E6-PDZ interaction promotes HPV genome maintenance at least in part by neutralization of an activity that can arise from residual undegraded p53.     

Codon Optimization of the Human Papillomavirus E7 Oncogene Induces a CD8+ T Cell Response to a Cryptic Epitope Not Harbored by Wild-Type E7

Codon optimization of nucleotide sequences is a widely used method to achieve high levels of transgene expression for basic and clinical research. Until now, immunological side effects have not been described. To trigger T cell responses against human papillomavirus, we incubated T cells with dendritic cells that were pulsed with RNA encoding the codon-optimized E7 oncogene. All T cell receptors isolated from responding T cell clones recognized target cells expressing the codon-optimized E7 gene but not the wild type E7 sequence. Epitope mapping revealed recognition of a cryptic epitope from the +3 alternative reading frame of codon-optimized E7, which is not encoded by the wild type E7 sequence. The introduction of a stop codon into the +3 alternative reading frame protected the transgene product from recognition by T cell receptor gene-modified T cells. This is the first experimental study demonstrating that codon optimization can render a transgene artificially immunogenic through generation of a dominant cryptic epitope. This finding may be of great importance for the clinical field of gene therapy to avoid rejection of gene-corrected cells and for the design of DNA- and RNA-based vaccines, where codon optimization may artificially add a strong immunogenic component to the vaccine.     

Triptolide Induces Growth Inhibition and Apoptosis of Human Laryngocarcinoma Cells by Enhancing p53 Activities and Suppressing E6-Mediated p53 Degradation

Triptolide, an active compound extracted from Chinese herb Leigongteng (Tripterygium wilfordii Hook F.), shows a broad-spectrum of anticancer activity through its cytotoxicity. However, the efficacy of triptolide on laryngocarcinoma rarely been evaluated, and the mechanism by which triptolide-induced cellular apoptosis is still not well understood. In this study, we found that triptolide significantly inhibited the laryngocarcinoma HEp-2 cells proliferation, migration and survivability. Triptolide induces HEp-2 cell cycle arrest at the G1 phase and apoptosis through intrinsic and extrinsic pathways since both caspase-8 and -9 are activated. Moreover, triptolide enhances p53 expression by increasing its stability via down-regulation of E6 and E6AP. Increased p53 transactivates down-stream target genes to initiate apoptosis. In addition, we found that short time treatment with triptolide induced DNA damage, which was consistent with the increase in p53. Furthermore, the cytotoxicity of triptolide is decreased by p53 knockdown or use of caspases inhibitor. In conclusion, our results demonstrated that triptolide inhibits cell proliferation and induces apoptosis in laryngocarcinoma cells by enhancing p53 expression and activating p53 functions through induction of DNA damage and suppression of E6 mediated p53 degradation. These studies indicate that triptolide is a potential anti-laryngocarcinoma drug.     

Sonoporation Delivery of Monoclonal Antibodies against Human Papillomavirus 16 E6 Restores p53 Expression in Transformed Cervical Keratinocytes

High-risk types of human papillomavirus (HPV), such as HPV16, have been found in nearly all cases of cervical cancer. Therapies targeted at blocking the HPV16 E6 protein and its deleterious effects on the tumour suppressor pathways of the cell can reverse the malignant phenotype of affected keratinocytes while sparing uninfected cells. Through a strong interdisciplinary collaboration between engineering and biology, a novel, non-invasive intracellular delivery method for the HPV16 E6 antibody, F127-6G6, was developed. The method employs high intensity focused ultrasound (HIFU) in combination with microbubbles, in a process known as sonoporation. In this proof of principle study, it was first demonstrated that sonoporation antibody delivery into the HPV16 positive cervical carcinoma derived cell lines CaSki and SiHa was possible, using chemical transfection as a baseline for comparison. Delivery of the E6 antibody using sonoporation significantly restored p53 expression in these cells, indicating the antibody is able to enter the cells and remains active. This delivery method is targeted, non-cytotoxic, and non-invasive, making it more easily translatable for in vivo experiments than other transfection methods.     

7-Hydroxystaurosporine (UCN-01) Induces Apoptosis in Human Colon Carcinoma and Leukemia Cells Independently of p53

7-hydroxystaurosporine (UCN-01) is a more selective protein kinase C inhibitor than staurosporine. UCN-01 exhibits antitumor activity in experimental tumor models and is presently in clinical trials. Our study reveals that human myeloblastic leukemia HL60 and K562 and colon carcinoma HT29 cells undergo internucleosomal DNA fragmentation and morphological changes characteristic of apoptosis after UCN-01 treatment. These three cell lines lack functional p53, and K562 and HT29 cells are usually resistant to apoptosis. DNA fragmentation in HT29 and K562 cells occurred after 1 day of treatment while it took less than 4 h in HL60 cells. Cycloheximide prevented UCN-01-induced DNA fragmentation in HT-29 cells, but not in HL60 and K562 cells, suggesting that macromolecular synthesis is selectively required for apoptotic DNA fragmentation in HT29 cells. UCN-01-induced DNA fragmentation was preceded by activation of cyclin B1/cdc2 kinase. Further studies in HL60 cells showed that UCN-01-induced apoptosis was associated with degradation of CPP32, PARP, and lamin B and that the inhibitor of caspases (ICE/CED-3 cysteine proteases), Z-VAD-FMK, and the serine protease inhibitor, DCI, protected HL60 cells from UCN-01-induced DNA fragmentation. However, only DCI and TPCK, but not Z-VAD-FMK, inhibited DNA fragmentation in the HL60 cell-free system, suggesting that serine protease(s) may play a role in the execution phase of apoptosis in HL60 cells treated with UCN-01. Z-VAD-FMK and DCI also inhibited apoptosis in HT29 cells. These data demonstrate that the protein kinase C inhibitor and antitumor agent, UCN-01 is a potent apoptosis inducer in cell lines that are usually resistant to apoptosis and lack p53 and that caspases and probably serine proteases are activated during UCN-01-induced apoptosis.