Expression of DNA damage response proteins in cervical cancer patients treated with radical chemoradiotherapy

OBJECTIVE

The management of locally advanced cervical cancer has improved significantly with the advent of cisplatin-based chemoradiotherapy (CRT) as the primary treatment regimen. Nevertheless, a significant proportion of patients fail to respond or relapse on this treatment and have a very poor prognosis. Our goal was to determine the prognostic value of a panel of proteins involved in detection and repair of DNA damage.

METHODS

We performed fluorescence immunohistochemistry, and used software analysis to assess expression of DNA damage response proteins ATM, DNA-PKcs, PARP-1, Ku70 and Ku86 in 117 pre-treatment specimens from patients with locally advanced cervical cancer. We compared expression to clinicopathologic correlates to determine prognostic significance.

RESULTS

Five-year progression-free survival was significantly lower in the low expressors than in high expressors of ATM (35% vs. 58%, p=0.044) and PARP-1 (24% vs. 61%, p=0.003), and showed a trend to significance for DNA-PKcs (30% vs. 60%, p=0.050). Low expression of the same proteins also correlated significantly with lower overall survival. In multivariable analysis, adjusted for FIGO stage and tumor size, low ATM and PARP-1 expression was significantly associated with both poorer progression-free and overall survival. Pairwise analyses indicated that expression levels of these proteins were correlated.

CONCLUSIONS

Expression of DNA damage response proteins in cervical cancer is associated with outcome in patients treated with CRT. Immunohistochemical analysis of these proteins may be useful in guiding treatment decisions in such patients.

The significance of HPV status in patients with anal cancer: A comparative technical analysis

425

Background: HPV status is an important prognostic factor for outcome in head and neck cancers, with improved survival noted in HPV-positive vs -negative tumors. However, this effect has not been studied in anal cancer. Challenges exist in defining reliable and methods of detecting the presence of high-risk HPV virus or the expression of HPV-related proteins. Purpose: To determine p16 expression and HPV16 status, in association with HPV subtyping, in pre-treatment anal cancers, and to correlate marker status with clinical outcome.

Methods: This retrospective study evaluated anal cancer patients treated between 1992-2005 with definitive RT or CRT at a single institution (Tom Baker Cancer Centre, Calgary, Canada). HPV subtyping, p16 protein expression via conventional immunohistochemistry (IHC) and automated quantitative IHC (AQUA), and HPV16 in-situ hybridization (CISH) was performed in a subset of patient tumors with sufficient pretreatment tumor specimen. The correlation between results using these different techniques and association of HPV markers with clinical outcome was evaluated.

Results: 89 patients were identified; M:F 1:2; median age 57 years. Median tumor size was 3.5 cm. All patients were treated with radical external beam RT with curative intent; 81% received concurrent chemotherapy. Clinical CR was observed in 73% of patients at 3 months post-treatment. Median OS was 82 months. Of tumor specimens analyzed for HPV subtype, 78% (28/36) were HPV16. Using conventional IHC (DAB), 80% (28/35) of tumors over-expressed p16 (score 3). HPV16 CISH was positive in 81% (34/42). P16 AQUA™ score correlated with HPV16 subtyping, CISH for HPV16, and p16 DAB. With HPV16 subtyping as the gold standard for HPV16 infection, false positive and false negative rates were 8% and 4% with DAB, and 10% and 13% with CISH. In univariate analysis only p16 AQUA score > 244 (upper 15%) was associated with PFS (p= 0.006-95% CI 8.4 [1.8-38.6]) and OS (p= 0.013-95% CI 4.5 [1.4-15]).

Conclusions: p16 expression via quantitative IHC correlates with other methods, including HPV16 subtyping, and appears more reliable for defining HPV16 status than CISH. Only quantitative IHC identified a subset of patients with worse outcome.

No significant financial relationships to disclose.

The viral tropism of two distinct oncolytic viruses, reovirus and myxoma virus, is modulated by cellular tumor suppressor gene status

Replication-competent oncolytic viruses hold great potential for the clinical treatment of many cancers. Importantly, many oncolytic virus candidates, such as reovirus and myxoma virus, preferentially infect cancer cells bearing abnormal cellular signaling pathways. Reovirus and myxoma virus are highly responsive to activated Ras and Akt signaling pathways, respectively, for their specificity for viral oncolysis. However, considering the complexity of cancer cell populations, it is possible that other tumor-specific signaling pathways may also contribute to viral discrimination between normal versus cancer cells. Because carcinogenesis is a multistep process involving the accumulation of both oncogene activations and the inactivation of tumor suppressor genes, we speculated that not only oncogenes but also tumor suppressor genes may have an important role in determining the tropism of these viruses for cancer cells. It has been previously shown that many cellular tumor suppressor genes, such as p53, ATM and Rb, are important for maintaining genomic stability; dysfunction of these tumor suppressors may disrupt intact cellular antiviral activity due to the accumulation of genomic instability or due to interference with apoptotic signaling. Therefore, we speculated that cells with dysfunctional tumor suppressors may display enhanced susceptibility to challenge with these oncolytic viruses, as previously seen with adenovirus. We report here that both reovirus and myxoma virus preferentially infect cancer cells bearing dysfunctional or deleted p53, ATM and Rb tumor suppressor genes compared to cells retaining normal counterparts of these genes. Thus, oncolysis by these viruses may be influenced by both oncogenic activation and tumor suppressor status.

Examination of surface-bound Ku-DNA complexes in an aqueous environment using MAC mode atomic force microscopy

In the development of biosensors, it is essential to understand how the signal-transducing element may perturb surface-bound proteins and nucleic acids. The tip of the atomic force microscope is such an element in atomic force microscopy. In this paper, we describe the influence of tip-sample interactions on the measured height of the DNA repair protein, Ku, that has been adsorbed onto a mica surface which was submerged in aqueous solution. We find that the measured height of the Ku molecule depends critically on whether or not it is associated with DNA. Additionally, we observed that the conditions (time and concentration) under which Ku is incubated with DNA, affect the appearance (number and type) of the DNA-Ku complexes observed.

Repair of DNA double strand breaks by non-homologous end joining

DNA double strand breaks (DSB) are the most serious form of DNA damage. If not repaired they can lead to cell death. If misrepaired DSBs contribute to chromosomal aberrations and genomic instability. Non-homologous end joining (NHEJ) is one of two major pathways for the repair of DSBs in human cells. Proteins known to be required for NHEJ include the DNA-dependent protein kinase (DNA-PK), XRCC4, DNA ligase IV, and Artemis. This review discusses how these and other accessory proteins may function in the repair of DSBs produced by ionizing radiation (IR) and by V(D)J recombination.

Accurate in vitro end joining of a DNA double strand break with partially cohesive 3'-overhangs and 3'-phosphoglycolate termini: effect of Ku on repair fidelity

To examine determinants of fidelity in DNA end joining, a substrate containing a model of a staggered free radical-mediated double-strand break, with cohesive phosphoglycolate-terminated 3'-overhangs and a one-base gap in each strand, was constructed. In extracts of Xenopus eggs, human lymphoblastoid cells, hamster CHO-K1 cells, and a Chinese hamster ovary (CHO) derivative lacking the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs), the predominant end joining product was that corresponding to accurate restoration of the original sequence. In extracts of the Ku-deficient CHO derivative xrs6, a shorter product, consistent with 3' --> 5' resection before ligation, was formed. Similar results were seen for a substrate with 5'-overhangs and recessed 3'-phosphoglycolate ends. Supplementation of the xrs6 extracts with purified Ku restored accurate end joining. In Xenopus and human extracts, but not in hamster extracts, gap filling and ligation were blocked by wortmannin, consistent with a requirement for DNA-PKcs activity. The results suggest a Ku-dependent pathway, regulated by DNA-PKcs, that can accurately restore the original DNA sequence at sites of free radical-mediated double-strand breaks, by protecting DNA termini from degradation and maintaining the alignment of short partial complementarities during gap filling and ligation.

Inhibition of the G2 DNA damage checkpoint and of protein kinases Chk1 and Chk2 by the marine sponge alkaloid debromohymenialdisine

Cells can respond to DNA damage by activating checkpoints that delay cell cycle progression and allow time for DNA repair. Chemical inhibitors of the G(2) phase DNA damage checkpoint may be used as tools to understand better how the checkpoint is regulated and may be used to sensitize cancer cells to DNA-damaging therapies. However, few inhibitors are known. We used a cell-based assay to screen natural extracts for G(2) checkpoint inhibitors and identified debromohymenialdisine (DBH) from a marine sponge. DBH is distinct structurally from previously known G(2) checkpoint inhibitors. It inhibited the G(2) checkpoint with an IC(50) of 8 micrometer and showed moderate cytotoxicity (IC(50) = 25 micrometer) toward MCF-7 cells. DBH inhibited the checkpoint kinases Chk1 (IC(50) = 3 micrometer) and Chk2 (IC(50) = 3.5 micrometer) but not ataxia-telangiectasia mutated (ATM), ATM-Rad3-related protein, or DNA-dependent protein kinase in vitro, indicating that it blocks two major branches of the checkpoint pathway downstream of ATM. It did not cause the activation or inhibition of different signal transduction proteins, as determined by mobility shift analysis in Western blots, suggesting that it inhibits a narrow range of protein kinases in vivo.

Protein phosphatases regulate DNA-dependent protein kinase activity

DNA-dependent protein kinase (DNA-PK) is a complex of DNA-PK catalytic subunit (DNA-PKcs) and the DNA end-binding Ku70/Ku80 heterodimer. DNA-PK is required for DNA double strand break repair by the process of nonhomologous end joining. Nonhomologous end joining is a major mechanism for the repair of DNA double strand breaks in mammalian cells. As such, DNA-PK plays essential roles in the cellular response to ionizing radiation and in V(D)J recombination. In vitro, DNA-PK undergoes phosphorylation of all three protein subunits (DNA-PK catalytic subunit, Ku70 and Ku80) and phosphorylation correlates with inactivation of the serine/threonine protein kinase activity of DNA-PK. Here we show that phosphorylation-induced loss of the protein kinase activity of DNA-PK is restored by the addition of the purified catalytic subunit of either protein phosphatase 1 or protein phosphatase 2A (PP2A) and that this reactivation is blocked by the potent protein phosphatase inhibitor, microcystin. We also show that treating human lymphoblastoid cells with either okadaic acid or fostriecin, at PP2A-selective concentrations, causes a 50-60% decrease in DNA-PK protein kinase activity, although the protein phosphatase 1 activity in these cells was unaffected. In vivo phosphorylation of DNA-PKcs, Ku70, and Ku80 was observed when cells were labeled with [(32)P]inorganic phosphate in the presence of the protein phosphatase inhibitor, okadaic acid. Together, our data suggest that reversible protein phosphorylation is an important mechanism for the regulation of DNA-PK protein kinase activity and that the protein phosphatase responsible for reactivation in vivo is a PP2A-like enzyme.

The plant isoflavenoid genistein activates p53 and Chk2 in an ATM-dependent manner

Genistein is an isoflavenoid that is abundant in soy beans. Genistein has been reported to have a wide range of biological activities and to play a role in the diminished incidence of breast cancer in populations that consume a soy-rich diet. Genistein was originally identified as an inhibitor of tyrosine kinases; however, it also inhibits topoisomerase II by stabilizing the covalent DNA cleavage complex, an event predicted to cause DNA damage. The topoisomerase II inhibitor etoposide acts in a similar manner. Here we show that genistein induces the up-regulation of p53 protein, phosphorylation of p53 at serine 15, activation of the sequence-specific DNA binding properties of p53, and phosphorylation of the hCds1/Chk2 protein kinase at threonine 68. Phosphorylation and activation of p53 and phosphorylation of Chk2 were not observed in ATM-deficient cells. In contrast, the topoisomerase II inhibitor etoposide induced phosphorylation of p53 and Chk2 in ATM-positive and ATM-deficient cells. In addition, genistein-treated ATM-deficient cells were significantly more susceptible to genistein-induced killing than were ATM-positive cells. Together our data suggest that ATM is required for activation of a DNA damage-induced pathway that activates p53 and Chk2 in response to genistein.

Utilization of oriented peptide libraries to identify substrate motifs selected by ATM

The ataxia telangiectasia mutated (ATM) gene encodes a serine/threonine protein kinase that plays a critical role in genomic surveillance and development. Here, we use a peptide library approach to define the in vitro substrate specificity of ATM kinase activity. The peptide library analysis identified an optimal sequence with a central core motif of LSQE that is preferentially phosphorylated by ATM. The contributions of the amino acids surrounding serine in the LSQE motif were assessed by utilizing specific peptide libraries or individual peptide substrates. All amino acids comprising the LSQE sequence were critical for maximum peptide substrate suitability for ATM. The DNA-dependent protein kinase (DNA-PK), a Ser/Thr kinase related to ATM and important in DNA repair, was compared with ATM in terms of peptide substrate selectivity. DNA-PK was found to be unique in its preference of neighboring amino acids to the phosphorylated serine. Peptide library analyses defined a preferred amino acid motif for ATM that permits clear distinctions between ATM and DNA-PK kinase activity. Data base searches using the library-derived ATM sequence identified previously characterized substrates of ATM, as well as novel candidate substrate targets that may function downstream in ATM-directed signaling pathways.

Three yeast proteins related to the human candidate tumor suppressor p33(ING1) are associated with histone acetyltransferase activities

Three Saccharomyces cerevisiae proteins (Yng1/YOR064c, Yng2/YHR090c, and Pho23) and two Schizosaccharomyces pombe proteins (Png1/CAA15917 and Png2/CAA21250) share significant sequence identity with the human candidate tumor suppressor p33(ING1) in their C-terminal regions. The homologous regions contain PHD finger domains which have been implicated in chromatin-mediated transcriptional regulation. We show that GFP-Yng2, like human Ing1, is localized in the nucleus. Deletion of YNG2 results in several phenotypes, including an abnormal multibudded morphology, an inability to utilize nonfermentable carbon sources, heat shock sensitivity, slow growth, temperature sensitivity, and sensitivity to caffeine. These phenotypes are suppressed by expression of either human Ing1 or S. pombe Png1, suggesting that the yeast and human proteins are functionally conserved. Yng1- and Pho23-deficient cells also share some of these phenotypes. We demonstrated by yeast two-hybrid and coimmunoprecipitation tests that Yng2 interacts with Tra1, a component of histone acetyltransferase (HAT) complexes. We further demonstrated by coimmunoprecipitation that HA-Yng1, HA-Yng2, HA-Pho23, and HA-Ing1 are associated with HAT activities in yeast. Genetic and biochemical evidence indicate that the Yng2-associated HAT is Esa1, suggesting that Yng2 is a component of the NuA4 HAT complex. These studies suggest that the yeast Ing1-related proteins are involved in chromatin remodeling. They further suggest that these functions may be conserved in mammals and provide a possible mechanism for the human Ing1 candidate tumor suppressor.

Purification and characterization of ATM from human placenta. A manganese-dependent, wortmannin-sensitive serine/threonine protein kinase

ATM is mutated in the human genetic disorder ataxia telangiectasia, which is characterized by ataxia, immune defects, and cancer predisposition. Cells that lack ATM exhibit delayed up-regulation of p53 in response to ionizing radiation. Serine 15 of p53 is phosphorylated in vivo in response to ionizing radiation, and antibodies to ATM immunoprecipitate a protein kinase activity that, in the presence of manganese, phosphorylates p53 at serine 15. Immunoprecipitates of ATM also phosphorylate PHAS-I in a manganese-dependent manner. Here we have purified ATM from human cells using nine chromatographic steps. Highly purified ATM phosphorylated PHAS-I, the 32-kDa subunit of RPA, serine 15 of p53, and Chk2 in vitro. The majority of the ATM phosphorylation sites in Chk2 were located in the amino-terminal 57 amino acids. In each case, phosphorylation was strictly dependent on manganese. ATM protein kinase activity was inhibited by wortmannin with an IC(50) of approximately 100 nM. Phosphorylation of RPA, but not p53, Chk2, or PHAS-I, was stimulated by DNA. The related protein, DNA-dependent protein kinase catalytic subunit, also phosphorylated PHAS-I, RPA, and Chk2 in the presence of manganese, suggesting that the requirement for manganese is a characteristic of this class of enzyme.

Protein-DNA complexes containing DNA-dependent protein kinase in crude extracts from human and rodent cells

The DNA-dependent protein kinase (DNA-PK) is composed of a large catalytic subunit (DNA-PKcs) and a DNA-binding protein, Ku. Cells lacking DNA-PK activity are radiosensitive and are defective in DNA double-strand break repair and V(D)J recombination. Although much information regarding the interactions of Ku with DNA ends is available, relatively little is known about the interaction of DNA-PKcs with DNA-bound Ku. Here we show, using electrophoretic mobility shift assays, that chemical crosslinkers enhance the formation of protein-DNA complexes containing DNA-PKcs, Ku and other proteins in extracts from cells of normal human cell lines. Extracts from cells of the radiosensitive human cell line M059J, which lacks DNA-PKcs, are not competent to form these protein-DNA complexes, while addition of purified DNA-PKcs protein restores complex formation. This assay may be useful for screening for DNA-PK function in cells of human cell lines and for identifying proteins that interact with the DNA-PK-DNA complex. We also show that Ku protein in rodent cells can interact with human DNA-PKcs; however, this assay may be less useful for studying Ku/DNA-PKcs interactions in cells of rodent cell lines due to the low abundance of DNA-PKcs in these cells.

Relative affinities of poly(ADP-ribose) polymerase and DNA-dependent protein kinase for DNA strand interruptions

Poly(ADP-ribose) polymerase (PARP) and DNA-dependent protein kinase (DNA-PK) are important nuclear enzymes that cooperate to minimize genomic damage caused by DNA strand interruptions. DNA strand interruptions trigger the ADP-ribosylation activity and phosphorylation activity of PARP and DNA-PK respectively. In order to understand the relationship of PARP and DNA-PK with respect to DNA binding required for their activation, we analyzed the kinetics of the reactions and determined the apparent dissociation constants (Kd app) of the enzymes for DNA strand interruptions. PARP has a high binding affinity for blunt ends of DNA (Kd app=116 pM) and 3' single-base overhangs (Kd app=332 pM) in comparison to long overhangs (Kd app=2.6-5.0 nM). Nicks are good activators of PARP although the affinity of PARP for nicks (Kd app=467 pM) is 4-fold less than that for blunt ends. The Kd app of DNA-PK for 3' single-base overhangs, blunt ends and long overhangs is 704 pM, 1.3 nM and 1.4-2.2 nM respectively. These results demonstrate that (1) PARP, when compared to DNA-PK, has a greater preference for blunt ends and 3' single-base overhangs but a weaker preference for long overhangs, and (2) nicks are effective in attracting and activating PARP. The possible implications of the preferences of PARP and DNA-PK for DNA strand interruptions in vivo are discussed.

DNA-dependent protein kinase phosphorylation sites in Ku 70/80 heterodimer

Ku antigen is composed of 70 and 82 kDa subunits (Ku70 and Ku80, respectively) that together bind with high affinity to ends of double-stranded DNA and other DNA structures in vitro. When bound to DNA, the Ku 70/80 heterodimer enhances the kinase activity of the catalytic subunit of the DNA-dependent protein kinase, DNA-PKcs. Ku and DNA-PKcs are required for V(D)J recombination and DNA double-strand break repair in vivo and may also play a role in regulation of transcription. Ku is phosphorylated by DNA-PKcs in vitro, and cells that lack DNA-PKcs are deficient in Ku phosphorylation in vitro, suggesting that Ku may be a physiological target for DNA-PK. Here we have identified the sites of DNA-PK phosphorylation in human Ku protein. We find that Ku70 is phosphorylated at a single serine residue, serine 6, located in the putative transcriptional activation domain, and Ku80 is phosphorylated at serines 577 and 580 and at threonine 715. Interestingly, none of the phosphorylation sites identified in Ku correspond to the serine-glutamine consensus for DNA-PK phosphorylation, consistent with previous reports that DNA-PK can recognize additional phosphorylation motifs.

Herpes simplex virus type 1 immediate-early protein vmw110 induces the proteasome-dependent degradation of the catalytic subunit of DNA-dependent protein kinase

Herpes simplex virus type 1 (HSV-1) infection causes the active degradation of the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs), and this process is reliant on the expression of the HSV-1 immediate-early protein Vmw110. In this study we investigated in more detail the mechanism by which the degradation occurs, the domains of Vmw110 which are required, and whether Vmw110 is by itself sufficient for the effect. We found that proteasome inhibitors prevented the degradation of DNA-PKcs, indicating the involvement of a proteasome pathway. Furthermore, the continued activity of DNA-PK during infection in the presence of these inhibitors indicated that Vmw110 does not directly alter the enzyme activity of DNA-PKcs prior to its degradation in a normal infection. Indeed, Vmw110 was found to bind to neither the catalytic nor Ku subunits of DNA-PK. Using mutant Vmw110 viruses we show that the RING finger domain of Vmw110 is essential for the induced degradation of DNA-PKcs but that the ability of Vmw110 to bind to a cellular ubiquitin-specific protease (HAUSP) is not required. When expressed in the absence of other viral proteins, Vmw110 was sufficient to cause the degradation of DNA-PKcs, indicating that the effect on the stability of DNA-PKcs was a direct consequence of Vmw110 activity and not an indirect Vmw110-dependent effect of virus infection. Finally, the Vmw110-induced degradation of DNA-PKcs and loss in DNA-PK activity appears to be beneficial to HSV-1 infection, as virus replication was more efficient in cells lacking DNA-PKcs, especially at low multiplicities of infection.

ATM associates with and phosphorylates p53: mapping the region of interaction

The human genetic disorder ataxia-telangiectasia (AT) is characterized by immunodeficiency, progressive cerebellar ataxia, radiosensitivity, cell cycle checkpoint defects and cancer predisposition. The gene mutated in this syndrome, ATM (for AT mutated), encodes a protein containing a phosphatidyl-inositol 3-kinase (PI-3 kinase)-like domain. ATM also contains a proline-rich region and a leucine zipper, both of which implicate this protein in signal transduction. The proline-rich region has been shown to bind to the SH3 domain of c-Abl, which facilitates its phosphorylation and activation by ATM. Previous results have demonstrated that AT cells are defective in the G1/S checkpoint activated after radiation damage and that this defect is attributable to a defective p53 signal transduction pathway. We report here direct interaction between ATM and p53 involving two regions in ATM, one at the amino terminus and the other at the carboxy terminus, corresponding to the PI-3 kinase domain. Recombinant ATM protein phosphorylates p53 on serine 15 near the N terminus. Furthermore, ectopic expression of ATM in AT cells restores normal ionizing radiation (IR)-induced phosphorylation of p53, whereas expression of ATM antisense RNA in control cells abrogates the rapid IR-induced phosphorylation of p53 on serine 15. These results demonstrate that ATM can bind p53 directly and is responsible for its serine 15 phosphorylation, thereby contributing to the activation and stabilization of p53 during the IR-induced DNA damage response.

Inactivation of DNA-dependent protein kinase by protein kinase Cdelta: implications for apoptosis

Protein kinase Cdelta (PKCdelta) is proteolytically cleaved and activated at the onset of apoptosis induced by DNA-damaging agents, tumor necrosis factor, and anti-Fas antibody. A role for PKCdelta in apoptosis is supported by the finding that overexpression of the catalytic fragment of PKCdelta (PKCdelta CF) in cells is associated with the appearance of certain characteristics of apoptosis. However, the functional relationship between PKCdelta cleavage and induction of apoptosis is unknown. The present studies demonstrate that PKCdelta associates constitutively with the DNA-dependent protein kinase catalytic subunit (DNA-PKcs). The results show that PKCdelta CF phosphorylates DNA-PKcs in vitro. Interaction of DNA-PKcs with PKCdelta CF inhibits the function of DNA-PKcs to form complexes with DNA and to phosphorylate its downstream target, p53. The results also demonstrate that cells deficient in DNA-PK are resistant to apoptosis induced by overexpressing PKCdelta CF. These findings support the hypothesis that functional interactions between PKCdelta and DNA-PK contribute to DNA damage-induced apoptosis.

Tra1p is a component of the yeast Ada.Spt transcriptional regulatory complexes

The yeast Ada and TBP class of Spt proteins interact in multiple complexes that are required for transcriptional regulation. We have identified Tra1p as a component of these complexes through tandem mass spectrometry analysis of proteins that associate with Ngg1p/Ada3p. TRA1 is an essential gene and encodes a 3744-amino acid protein that is a member of a group of proteins including the catalytic subunit of DNA-dependent protein kinase, ATM and TRRAP, with carboxyl-terminal regions related to phosphatidylinositol 3-kinases. The interaction between Tra1p and Ada/Spt components was verified by the reciprocal coimmunoprecipitation of Ada2p and Tra1p from whole cell extracts in one or more complexes containing Spt7p. Tra1p cofractionated with Ngg1p and Spt7p through consecutive chromatography on Mono Q, DNA-cellulose, and Superose 6 columns. Binding of Tra1p to DNA-cellulose required Ada components. The association of Tra1p with two Ada.Spt complexes was suggested by its cofractionation with Ngg1p and Spt7p in two peaks on the Mono Q column. In the absence of Ada2p, the elution profile of Tra1p shifted to a distinct peak. Despite the similarity of Tra1p to a group of putative protein kinases, we have not detected protein kinase activity within immunoprecipitates of Tra1p or the Ada.Spt complexes.

Resistance to etoposide-induced apoptosis in a Burkitt's lymphoma cell line

Burkitt's lymphoma cells that vary in their phenotypic characteristics show significantly different degrees of susceptibility to radiation-induced apoptosis. Propensity to undergo apoptosis is reflected in the degradation of substrates such as DNA-dependent protein kinase but the status of bcl-2, c-myc and p53 has been uninformative. In this study, we have focused on 2 Epstein-Barr virus (EBV)-associated Burkitt's cell lines, one (WW2) susceptible and the other (BL29) resistant to etoposide-induced apoptosis. Differences in expression of BHRF1, an EBV gene that is homologous to the Bcl-2 proto-oncogene and known to inhibit apoptosis, or changes in apoptosis inhibitory proteins (IAPs), did not appear to account for the difference in susceptibility in the 2 cell lines. Cytoplasmic extracts from etoposide-treated WW2 cells caused apoptotic changes in nuclei isolated from either BL29 or WW2 cells, whereas extracts from BL29 cells failed to do so. In addition, extracts from etoposide-treated WW2 cells degraded the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs), an important indicator of apoptosis, but this protein was resistant to degradation by BL29 extracts. It appears likely that caspase 3 (CPP32) is involved in this degradation since it was activated only in the apoptosis susceptible cells and the pattern of cleavage of DNA-PKcs was similar to that reported previously with recombinant caspase 3. As observed previously, addition of caspase 3 to nuclei failed to induce morphological changes indicative of apoptosis, but addition of caspase 3 to nuclei in the presence of extract from the resistant cells led to apoptotic changes. We conclude that resistance to apoptosis in BL29 cells is due to a failure of etoposide to activate upstream effectors of caspase activity.

DNA-dependent protein kinase acts upstream of p53 in response to DNA damage

The tumour suppressor p53 becomes activated as a transcription factor in response to DNA damage, but the mechanism for this activation is unclear. A good candidate for an upstream activator of p53 is the DNA-dependent protein kinase (DNA-PK) that depends on the presence of DNA breaks for its activity. Here we investigate the link between DNA damage and the activation of DNA-PK and of p53. To determine whether DNA-PK is an upstream mediator of the p53 DNA-damage response, we analysed a severe combined-immunodeficiency (SCID) mouse cell line, SCGR11, and the human glioma cell line M059J . Both cell lines lack any detectable DNA-PK activity. We find that p53 is incapable of binding to DNA in the absence of DNA-PK, that DNA-PK is necessary but not sufficient for activation of p53 sequence-specific DNA binding, and that this activation occurs in response to DNA damage. Our results establish DNA-PK as a link between DNA damage and p53 activation, and reveal the existence of a mammalian DNA-damage-response pathway.

Lack of correlation between ATM protein expression and tumour cell radiosensitivity

PURPOSE

Cells derived from individuals in which the ataxia telangiectasia (ATM) gene is mutated are hypersensitive to ionizing radiation. Whether differences in ATM protein levels exist among human malignant glioma cell lines and whether such differences are correlated with cellular radiosensitivity were determined.

MATERIALS AND METHODS

Polyclonal antibodies were raised to separate regions of the ATM protein. ATM protein expression in human malignant glioma cell lines, SV40 transformed normal human fibroblasts and SV40 transformed AT fibroblasts was analysed by Western blotting. Reverse transcriptase polymerase chain reaction (RT-PCR) was used to assess the presence of ATM transcript.

RESULTS

While ATM protein was detected in all cell extracts, significant differences in the level of expression were observed. There was no apparent correlation between cellular radiosensitivity and differences in ATM protein levels in these human glioma cells. Extremely low levels of ATM protein were observed in M059J cells, which provide the only example of DNA-dependent protein kinase (DNA-PKcs) deficiency in a cell line of human origin.

CONCLUSIONS

Variations in the levels of ATM protein are insufficient to explain the differences in cellular radiosensitivity observed in a panel of human malignant glioma cell lines.

DNA-dependent protein kinase interacts with antigen receptor response element binding proteins NF90 and NF45

The DNA-dependent protein kinase (DNA-PK) is composed of a large catalytic subunit of approximately 470 kDa (DNA-PKcs) and the DNA-binding protein, Ku. Absence of DNA-PK activity confers sensitivity to x-rays and defects in both DNA double-strand break repair and V(D)J recombination. However the precise function of DNA-PK in DNA double-strand break repair is not known. Here we show, using electrophoretic mobility shift assays, that polypeptides in a fraction purified from human cells interact with DNA-PK and stabilize the formation of a complex containing DNA-PKcs-Ku and DNA. Five polypeptides in this fraction have been identified by amino-terminal sequence analysis and/or immunoblotting. These proteins are NF90 and NF45, which are the 90- and 45-kDa subunits of a protein known to bind specifically to the antigen receptor response element of the interleukin 2 promoter, and the alpha, beta, and gamma subunits of eukaryotic translation initiation factor eIF-2. We also show that NF90, NF45, and eIF-2 beta are substrates for DNA-PK in vitro. In addition, recombinant NF90 promotes formation of a complex between DNA-PKcs, Ku, and DNA, and antibodies to recombinant NF90 or recombinant NF45 immunoprecipitate DNA-PKcs in vitro. Together, our data suggest that NF90, in complex with NF45, interacts with DNA-PKcs and Ku on DNA and that NF90 and NF45 may be important for the function of DNA-PK.

Interaction of DNA-dependent protein kinase and poly(ADP-ribose) polymerase with radiation-induced DNA strand breaks

Two of the enzymes involved in the response of mammalian cells to ionizing radiation are the DNA-dependent protein kinase and poly(ADP-ribose) polymerase. These enzymes are known to be activated by binding to DNA strand breaks, but previous studies designed to look at strand break specificity have employed enzymatically generated strand breaks and not irradiated DNA. Using highly purified DNA-dependent protein kinase, we compared enzyme activation by a series of DNA substrates. Irradiated plasmid DNA activated DNA-dependent protein kinase in a dose-dependent manner. When calculated in terms of the molar concentration of double-strand breaks, the enzyme activation by irradiated DNA was comparable to that by restriction enzyme-cleaved DNA. Linear DNA purified after plasmid irradiation also activated DNA-dependent protein kinase to a comparable extent, but nicked DNA, either isolated from irradiated plasmid or generated by DNase I, failed to activate the enzyme. A comparison of the enzyme activation by plasmid molecules with different 3'- and 5'-terminal groups indicated that the chemical nature of the DNA termini at the double-strand break does not significantly influence the response of the DNA-dependent protein kinase. Similar experiments with poly(ADP-ribose) polymerase demonstrated that single- and double-strand breaks activate this enzyme with almost equal efficiency, but because of their greater number, single-strand breaks dominate the response of poly(ADP-ribose) polymerase to irradiated DNA.

Cell cycle checkpoints and DNA repair in Nijmegen breakage syndrome

Nijmegen breakage syndrome is characterized by a variable T cell and B cell immunodeficiency, growth failure, and an increased risk of malignancy. It is inherited in an autosomal recessive manner and is biochemically related to ataxia-telangiectasia. Cells from a patient with Nijmegen breakage syndrome were unable to arrest cell cycle progression after exposure to ionizing radiation, and BrdU incorporation into newly synthesized DNA was uninhibited, demonstrating that these cells have an aberrant response to radiation exposure. Although gross chromosomal breakage was observed, dinucleotide repeat segments were stable over time, suggesting that other types of DNA stability were not affected. DNA-PK activity, which is mediated by a protein related to the ataxia-telangiectasia gene product and is intimately involved in DNA repair and VDJ recombination, was normal in cells from an NBS patient. Therefore, cells from patients with Nijmegen breakage syndrome have an abnormal response to radiation exposure similar to that seen in ataxia-telangiectasia.

Attenuation of DNA-dependent protein kinase activity and its catalytic subunit by the herpes simplex virus type 1 transactivator ICP0

The DNA-dependent protein kinase (DNA-PK) is involved in several fundamental nuclear processes, including DNA double-strand break repair, V(D)J recombination, and transcription by RNA polymerases I and II. In this study, we show that infection of mammalian cells with herpes simplex virus type 1 attenuates DNA-PK activity by specifically depleting the p350/DNA-PKcs catalytic subunit. The half-life of the p350/DNA-PKcs protein decreases from greater than 24 h to less than 4 h following infection. The depletion of DNA-PK activity and p350/DNA-PKcs abundance is dependent on expression of the viral immediate-early protein ICP0. As ICP0 acts as a promoter-independent transactivator of gene expression, these data suggest that ICP0 may function by directly or indirectly targeting the p350/DNA-PKcs subunit of DNA-PK, thereby altering the inhibitory effects of DNA-PK on RNA polymerase II transcription.

Interleukin-1 beta-converting enzyme-like protease cleaves DNA-dependent protein kinase in cytotoxic T cell killing

Cytotoxic T cells (CTL) represent the major defense mechanism against the spread of virus infection. It is believed that the pore-forming protein, perforin, facilitates the entry of a series of serine proteases (particularly granzyme B) into the target cell which ultimately leads to DNA fragmentation and apoptosis. We demonstrate here that during CTL-mediated cytolysis the catalytic subunit of DNA-dependent protein kinase (DNA-PKcs), an enzyme implicated in the repair of double strand breaks in DNA, is specifically cleaved by an interleukin (IL)-1 beta-converting enzyme (ICE)-like protease. A serine protease inhibitor, 3,4-dichloroisocoumarin (DCl), which is known to block granzyme B activity, inhibited CTL-induced apoptosis and prevented the degradation of DNA-PKcs in cells but failed to prevent the degradation of purified DNA-PKcs by CTL extracts. However, Tyr-Val-Ala-Asp-CH2Cl (YVAD-CMK) and other cysteine protease inhibitors prevented the degradation of purified DNA-PKcs by CTL extracts. Furthermore, incubation of DNA-PKcs with granzyme B did not produce the same cleavage pattern observed in cells undergoing apoptosis and when this substrate was incubated with either CTL extracts or the ICE-like protease, CPP32. Sequence analysis revealed that the cleavage site in DNA-PKcs during CTL killing was the same as that when this substrate was exposed to CPP32. This study demonstrates for the first time that the cleavage of DNA-PKcs in this intact cell system is exclusively due to an ICE-like protease.

DNA-dependent protein kinase catalytic subunit: a target for an ICE-like protease in apoptosis

Radiosensitive cell lines derived from X-ray cross complementing group 5 (XRCC5), SCID mice and a human glioma cell line lack components of the DNA-dependent protein kinase, DNA-PK, suggesting that DNA-PK plays an important role in DNA double-strand break repair. Another enzyme implicated in DNA repair, poly(ADP-ribose) polymerase, is cleaved and inactivated during apoptosis, suggesting that some DNA repair proteins may be selectively targeted for destruction during apoptosis. Here we demonstrate that DNA-PKcs, the catalytic subunit of DNA-PK, is preferentially degraded after the exposure of different cell types to a variety of agents known to cause apoptosis. However, Ku, the DNA-binding component of the enzyme, remains intact. Degradation of DNA-PKcs was accompanied by loss of DNA-PK activity. One cell line resistant to etoposide-induced apoptosis failed to show degradation of DNA-PKcs. Protease inhibitor data implicated an ICE-like protease in the cleavage of DNA-PKcs, and it was subsequently shown that the cysteine protease CPP32, but not Mch2alpha, ICE or TX, cleaved purified DNA-PKcs into three fragments of comparable size with those observed in cells undergoing apoptosis. Cleavage sites in DNA-PKcs, determined by antibody mapping and microsequencing, were shown to be the same for CPP32 cleavage and for cleavage catalyzed by extracts from cells undergoing apoptosis. These observations suggest that DNA-PKcs is a critical target for proteolysis by an ICE-like protease during apoptosis.

The DNA-dependent protein kinase, DNA-PK: 10 years and no ends in sight

The phenomenon of DNA-dependent protein phosphorylation was first described little over 10 years ago. Since then a DNA-dependent protein kinase, DNA-PK, has been purified from human cells and many of its biochemical properties have been characterized. DNA-PK is composed of a large catalytic subunit, DNA-PKcs, and a DNA-targeting protein, Ku. When assembled on a suitable DNA molecule, the DNA-PK holoenzyme acts as a serine/threonine protein kinase that in vitro phosphorylates many DNA binding and non-binding proteins and transcription factors. Recent genetic studies point strongly to functions in DNA double-strand break repair and V(D)J recombination. In addition, biochemical studies suggest a role in the regulation of transcription. Here we discuss, from a historical perspective, the events leading up to our current understanding of the function of DNA-PK, including recent results from our own studies suggesting the involvement of DNA-PK in apoptosis and in viral infection of human cells.

The DNA-dependent protein kinase is inactivated by autophosphorylation of the catalytic subunit

The DNA-dependent protein kinase (DNA-PK) requires for activity free ends or other discontinuities in the structure of double strand DNA. In vitro, DNA-PK phosphorylates several transcription factors and other DNA-binding proteins and is thought to function in DNA damage recognition or repair and/or transcription. Here we show that in vitro DNA-PK undergoes autophosphorylation of all three protein subunits (DNA-PKcs, Ku p70 and Ku p80) and that phosphorylation correlates with inactivation of the serine/threonine kinase activity of DNA-PK. Significantly, activity is restored by the addition of purified native DNA-PKcs but not Ku, suggesting that inactivation is due to autophosphorylation of DNA-PKcs. Our data also suggest that autophosphorylation results in dissociation of DNA-PKcs from the Ku-DNA complex. We suggest that autophosphorylation is an important mechanism for the regulation of DNA-PK activity.

Purification and characterization of the double-stranded DNA-activated protein kinase, DNA-PK, from human placenta

The double-stranded DNA-activated protein kinase (DNA-PK) is a serine-threonine protein kinase that is composed of a large catalytic subunit (p350) and a DNA-binding protein of 70 and 80 kDa subunits known as the Ku autoantigen. When targeted to DNA by free DNA ends, DNA-PK phosphorylates many DNA-binding proteins and transcription factors. Previously, DNA-PK had only been purified and characterized from transformed human tissue culture cells. Here we report that DNA-PK is an abundant protein in human placenta and lymphocytes. We have purified the placental DNA-PK to homogeneity and show that its biochemical properties are similar to those of the HeLa cell enzyme.

Lack of correlation between DNA-dependent protein kinase activity and tumor cell radiosensitivity

Lack of DNA-dependent protein kinase (DNA-PK) activity confers radiosensitivity and defective DNA double-strand break repair. Nine human malignant glioma cell lines were studied to determine whether differences in DNA-PK activity reflect differences in inherent radiosensitivity or are predictive of tumor treatment response. DNA-PK activity was present in all cell extracts, as were the DNA-PK proteins, DNA-PK catalytic subunit, Ku p70, and Ku p80. No correlation was found between the levels of DNA-PK activity and inherent radiosensitivity or in the tumor treatment response. These preliminary results suggest that variation in DNA-PK activity may not be a determinant of clinical response in malignant glioma.

DNA-dependent protein kinase catalytic subunit: a relative of phosphatidylinositol 3-kinase and the ataxia telangiectasia gene product

DNA-dependent protein kinase (DNA-PK), which is involved in DNA double-stranded break repair and V(D)J recombination, comprises a DNA-targeting component called Ku and an approximately 460 kDa catalytic subunit, DNA-PKcs. Here, we describe the cloning of the DNA-PKcs cDNA and show that DNA-PKcs falls into the phosphatidylinositol (PI) 3-kinase family. Biochemical assays, however, indicate that DNA-PK phosphorylates proteins but has no detectable activity toward lipids. Strikingly, DNA-PKcs is most similar to PI kinase family members involved in cell cycle control, DNA repair, and DNA damage responses. These include the FKBP12-rapamycin-binding proteins Tor1p, Tor2p, and FRAP, S. pombe rad3, and the product of the ataxia telangiectasia gene, mutations in which lead to genomic instability and predisposition to cancer. The relationship of these proteins to DNA-PKcs provides important clues to their mechanisms of action.

Absence of p350 subunit of DNA-activated protein kinase from a radiosensitive human cell line

The radiosensitive rodent mutant cell line xrs-5 is defective in DNA double-strand break repair and lacks the Ku component of the DNA-activated protein kinase, DNA-PK. Here radiosensitive human cell lines were analyzed for DNA-PK activity and for the presence of related proteins. The radiosensitive human malignant glioma M059J cell line was found to be defective in DNA double-strand break repair, but fails to express the p350 subunit of DNA-PK. These results suggest that DNA-PK kinase activity is involved in DNA double-strand break repair.

The carboxyl-terminal transactivation domain of human serum response factor contains DNA-activated protein kinase phosphorylation sites

The serum response factor (SRF) is a 67-kDa phosphoprotein that, together with auxiliary factors, modulates transcription of immediate early genes containing serum response elements in their promoters. Here we show that the carboxyl-terminal domain of human SRF is phosphorylated in vivo and is recognized in vitro by the double-stranded DNA-activated serine/threonine-specific protein kinase, DNA-PK. SRF phosphorylation by DNA-PK was stimulated by its cognate binding site. Protein microsequence analysis of a 22-amino acid synthetic SRF peptide and phosphopeptide analysis of genetically altered glutathione S-transferase-SRF fusion proteins identified Ser-435 and Ser-446 of human SRF as sites phosphorylated by DNA-PK. Both serines are followed by glutamine. Changing Gln-436 and Gln-447 to other residues reduced or eliminated phosphorylation by DNA-PK, confirming that these glutamines are important determinants for kinase recognition. The carboxyl-terminal transcription activation domain was mapped within a 71-amino acid region that contains both DNA-PK phosphorylation sites. Amino acid substitutions that interfered with phosphorylation by DNA-PK at Ser-435/446 in GAL4-SRF fusion proteins were reduced in transactivation potency. From these data we suggest that DNA-PK phosphorylation may modulate SRF activity in vivo.

Dimerization and DNA binding alter phosphorylation of Fos and Jun

Fos and Jun form dimeric complexes that bind to activator protein 1 (AP-1) DNA sequences and regulate gene expression. The levels of expression and activities of these proteins are regulated by a variety of extracellular stimuli. They are thought to function in nuclear signal transduction processes in many different cell types. The role of Fos and Jun in gene transcription is complex and may be regulated in several ways including association with different dimerization partners, interactions with other transcriptional factors, effects on DNA topology, and reduction/oxidation of a conserved cysteine residue in the DNA-binding domain. In addition, phosphorylation has been suggested to control the activity of Fos and Jun. Here we show that phosphorylation of Fos and Jun by several protein kinases is affected by dimerization and binding to DNA. Jun homodimers are phosphorylated efficiently by casein kinase II, whereas Fos-Jun heterodimers are not. DNA binding also reduces phosphorylation of Jun by casein kinase II, p34cdc2 (cdc2) kinase, and protein kinase C. Phosphorylation of Fos by cAMP-dependent protein kinase and cdc2 is relatively insensitive to dimerization and DNA binding, whereas phosphorylation of Fos and Jun by DNA-dependent protein kinase is dramatically stimulated by binding to the AP-1 site. These results imply that different protein kinases can distinguish among Fos and Jun proteins in the form of monomers, homodimers, and heterodimers and between DNA-bound and non-DNA-bound proteins. Thus, potentially, these different states of Fos and Jun can be recognized and regulated independently by phosphorylation.

Phosphorylation at Ser-15 and Ser-392 in mutant p53 molecules from human tumors is altered compared to wild-type p53

The product of the p53 gene suppresses cell growth and plays a critical role in suppressing development of human tumors. p53 protein binds DNA, activates transcription, and can be phosphorylated at N- and C-terminal sites. Previously, wild-type p53 was shown to be hyperphosphorylated compared to mutant p53 during p53-mediated growth arrest in vivo. Here we show that Ser-15 and Ser-9 in the N-terminal transactivation domain of wild-type human p53 are phosphorylated in vivo in cells derived from the human glioblastoma line T98G. In [Ile237]p53 and [Ala143]p53, two natural p53 mutants from human tumors that are defective for activation of transcription, phosphorylation at Ser-15 was reduced and phosphorylation at Ser-392 was increased compared to wild-type p53. No change was observed at Ser-9. [His273]p53, a third mutant, had a phosphorylation state similar to that of wild-type p53. We suggest that phosphorylation of Ser-15 may depend on the ability of p53 to adopt a wild-type conformation and may contribute to p53's ability to block cell growth.

Mutation of the serine 15 phosphorylation site of human p53 reduces the ability of p53 to inhibit cell cycle progression

Overexpression of wild-type p53 prevents cells from entering the S phase of the cell cycle. The amino-terminal transactivation region of p53 is phosphorylated by several protein kinases, including DNA-PK, a nuclear serine/threonine protein kinase that in vitro requires DNA for activity. DNA-PK was recently shown to phosphorylate serines 15 and 37 of human p53 (Lees-Miller et al., 1992. Mol. Cell. Biol., 12, 5041-5049). To prevent phosphorylation at these sites, mutants were constructed that changed the codons for serine 15 or serine 37 to alanine codons. Expression of p53-Ala-37 in stably transformed T98G cells blocked progression of the cells into S phase as well as did the expression of wild-type p53. In contrast, p53-Ala-15 was partially defective in blocking cell cycle progression. Several cell clones transformed with the mutant p53-Ala-15 gene expressed normal levels of p53 mRNA but accumulated little or no detectable p53 protein. However, by using a transient expression system driven by a strong cytomegalovirus promoter, we showed that the inability of p53-Ala-15 to fully block cell cycle progression was not due to inadequate levels of expression or to a failure of the mutant protein to accumulate in the nucleus. These results suggest that phosphorylation of Ser-15 may affect p53 function.

Human DNA-activated protein kinase phosphorylates serines 15 and 37 in the amino-terminal transactivation domain of human p53

Human DNA-PK is a nuclear, serine/threonine protein kinase that, when activated by DNA, phosphorylates several DNA-binding substrates, including the tumor suppressor protein p53. To identify which p53 residues are phosphorylated, we examined DNA-PK's ability to phosphorylate synthetic peptides corresponding to human p53 sequences. Serines 15 and 37 in the amino-terminal transactivation domain of human p53, and serines 7 and 18 of mouse p53, were phosphorylated by DNA-PK in the context of synthetic peptides. Other serines in these p53 peptides, and serines in other p53 peptides, including peptides containing the serine 315 p34cdc2 site and the serine 392 casein kinase II site, were not recognized by DNA-PK or were phosphorylated less efficiently. Phosphorylation of the conserved serine 15 in human p53 peptides depended on the presence of an adjacent glutamine, and phosphorylation was inhibited by the presence of a nearby lysine. Phosphorylation of recombinant wild-type mouse p53 was inhibited at high DNA concentrations, suggesting that DNA-PK may phosphorylate p53 only when both are bound to DNA at nearby sites. Our study suggests that DNA-PK may have a role in regulating cell growth and indicates how phosphorylation of serine 15 in DNA-bound p53 could alter p53 function.

The human DNA-activated protein kinase phosphorylates simian virus 40 T antigen at amino- and carboxy-terminal sites

Protein phosphorylation modulates the functions of simian virus 40 large T antigen (TAg) in productive and transforming infections. We recently described a DNA-activated protein kinase (DNA-PK) that efficiently phosphorylates TAg and several other nuclear, DNA-binding proteins in vitro (S.P. Lees-Miller, Y.-R. Chen, and C. W. Anderson, Mol. Cell. Biol. 10:6472-6481, 1990). In this report, we show by direct amino acid sequence analysis that DNA-PK phosphorylates TAg strongly at Ser-677, a residue known to be important for TAg interaction with origin site I and for transformation. We propose that DNA-PK may modulate the role of TAg in repressing early viral transcription and cell transformation, but a role for DNA-PK in regulating simian virus 40 DNA synthesis is not excluded. DNA-PK also phosphorylates Ser-665, and Ser-667, and one or more serines between amino acids 110 and 131. At least six serines, Ser-111, Ser-112, Ser-120, Ser-665, Ser-667, and Ser-677, are phosphorylated in TAg purified from baculovirus vector-infected insect cells.

The DNA-activated protein kinase, DNA-PK: a potential coordinator of nuclear events

DNA-PK is a moderately abundant serine/threonine protein kinase found in the nucleus of a wide range of eukaryotic cells. It is one of the few known cellular enzymes whose activity is regulated directly by DNA. Many DNA binding proteins, including a number of transcription factors, are substrates for DNA-PK in vitro. We suggest that this kinase may coordinate signal transduction pathways and nuclear events, including transcription, in response to changes in DNA or chromatin state.

Human cells contain a DNA-activated protein kinase that phosphorylates simian virus 40 T antigen, mouse p53, and the human Ku autoantigen

HeLa cells contain a serine/threonine protein kinase (DNA-PK) that is strongly activated in vitro by low concentrations of double-stranded DNA (dsDNA). Activation was specific for dsDNA; both natural DNAs and synthetic oligonucleotides functioned as kinase activators. The fact that DNA-PK activity was rapidly inhibited by incubation with dsDNA and ATP suggests that DNA-PK activity also may be regulated by autophosphorylation. During gel filtration, DNA-PK activity behaved as a 350-kDa protein, and highly purified DNA-PK contained a dsDNA-binding, 350-kDa polypeptide that was phosphorylated in a dsDNA-dependent manner. We conclude that this 350-kDa polypeptide is likely to be DNA-PK. Previously we showed that the dsDNA-activated kinase phosphorylates two threonines at the N terminus of hsp90 alpha (S. P. Lees-Miller and C. W. Anderson, J. Biol. Chem. 264:17275-17280, 1989). Here we show that DNA-PK also phosphorylates the simian virus 40 large tumor antigen, the mouse tumor-suppressor protein p53, the human Ku autoantigen, and two unidentified HeLa DNA-associated polypeptides of 52 and 110 kDa. Identification of these and other newly identified DNA-binding substrates suggest that the dsDNA-activated kinase may regulate transcription, DNA replication, or cell growth.

The human double-stranded DNA-activated protein kinase phosphorylates the 90-kDa heat-shock protein, hsp90 alpha at two NH2-terminal threonine residues

The 90-kDa heat-shock protein, hsp90, is an abundant cytoplasmic protein that can be phosphorylated in vitro by a double-stranded (ds) DNA-activated protein kinase found in cells from several species. Here we show that the dsDNA-activated protein kinase from human HeLa cells phosphorylates 2 threonine residues in the sequence PEETQTQDQPME at the amino terminus of human hsp90 alpha. Hsp90 beta, which is 97% identical to hsp90 alpha but lacks both amino-terminal threonines, is not phosphorylated by the dsDNA-activated protein kinase. Mouse hsp86 and rabbit hsp90 alpha are homologous to human hsp90 alpha; both heterologous proteins are phosphorylated at the same amino-terminal threonines by the human dsDNA-activated protein kinase.