If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Life on Earth has evolved in an often hostile environment, leading organisms to develop multiple protective mechanisms. Central among these are systems for minimizing DNA damage and thus maintaining genomic integrity. Complex enzyme systems for detecting and repairing DNA base damage are present in prokaryotic bacteria and have been retained throughout evolution. Nucleotide excision repair (NER), e.g., is remarkably similar between bacterial and human cells, and mutation in any of the major NER genes can produce a usually fatal human disease, xeroderma pigmentosum (XP), with principal manifestations in the skin, the target of environmental ultraviolet (UV) irradiation. The critical importance of maintaining genomic integrity is underlined by two further phenomena. First, DNA repair capacity is induced following DNA damage, allowing organisms to devote even more energy to DNA repair when stimulated by environmental insult. This so-called SOS response has been extensively studied in bacteria and a functional ortholog was more recently well documented in mammalian cells as well. Second, in higher organisms, additional pathways have evolved to assure that under circumstances of irreparable DNA damage, cells eliminate themselves from the proliferative compartment and thus do not propagate mutations.
Work in our laboratory and others has provided insights into the mechanism by which cells in higher organisms, including human cells, use these evolutionarily perfected strategies for avoiding the most devastating consequence of unrepaired DNA damage: cancer. The following sections describe a recently elucidated key role for telomeres as well as evidence that this innate cancer avoidance mechanism may be harnessed for the prevention and treatment of skin malignancies.
Telomere Structure and Function
Eukaryotic chromosomes end in telomeres, tandem repeats of TTAGGG and its complement in all mammalian cells (
). Because telomeres do not encode genes or regulatory sequences, the buffer zone of TTAGGG repeats prevents loss of genetic information as cells divide. Also, because of this progressive shortening, after approximately 60 post-natal rounds of cell division, telomeres reach a critically short length (
). Thus, a second function of telomeres is to act as a biologic clock, instructing cells either that they are young and proliferative or old and non-proliferative. Work from the laboratory of Titia de Lange has identified a third major role of telomeres, the initiation of DNA damage responses, as described below.
Telomeres are normally in a loop configuration with the double-stranded chromosome folded back on itself (
). After linearization of the chromosome, the 3′ overhang is rapidly digested and the ATM (ataxia telanglectasia mutated) kinase, the protein mutated in the disease ataxia telangiectasia, is activated (
). Both of these behaviors are believed to represent protective DNA damage responses: apoptosis because it removes from the tissue cells with extensive and presumably unrepairable DNA damage and senescence because it prevents further proliferation of cells at risk for malignant transformation.
Interpretation of replicative senescence, in addition to apoptosis, as a cancer prevention mechanism arises not only from a philosophical appreciation that non-dividing cells cannot give rise to a malignancy (
), a condition by definition in which cell growth is dysregulated and continues indefinitely, but also from recent observations that acute DNA damage or overexpression of certain oncogenes can give rise to the same senescent phenotype as does prolonged serial passage (
It is thus apparent that experimental telomere loop disruption and aging (serial cell division) share a common final pathway with acute DNA damage, such as double-strand breaks. All three responses involve activation of phosphatidyl inositol-3-like kinase(s) such as ATM, the ATM-related (ATR) kinase, or the DNA dependent-protein kinase (DNA-PK), which in turn activate p53 and other effector proteins (
). The final cellular response, apoptosis versus proliferative senescence, depends on the cell type and not on the character of the initial stimulus. These observations have led us to hypothesize that the common final pathway begins not with activation of the kinase but rather more proximally, with disruption of the telomere loop and exposure of the 3′ single-stranded overhang sequence, in mammalian cells repeats of TTAGGG.
DNA Damage Responses in Skin
In human skin, the most common medically consequential forms of DNA damage result from UV exposure. UV plays a major role in the great majority of the more than 1.3 million basal and squamous cell carcinomas diagnosed each year in the US (
). In addition to ample and/or intense intermittent sun exposure, major risk factors for skin cancer include fair complexion, poor tanning ability, and tendency to freckle, all indicative of vulnerability to UV insult (
), in skin that is genetically capable of this response (Fitzpatrick phototypes II–VI). Tanning is also well documented to protect against acute and chronic consequences of future UV exposures, including photocarcinogenesis (
). Interestingly, in intact human skin, the action spectrum for production of UV-induced DNA damage, specifically for cyclobutane pyrimidine dimers (CPD), the most common DNA photoproduct, is identical to the action spectrum for tanning, peaking broadly at 300 nm, and then falling off by four to five orders of magnitude at longer UV wavelengths (
), strongly supporting the hypothesis that DNA damage itself plays a large role in triggering the tanning response after UV exposure. Finally, work by our laboratory and others demonstrated that the rate-limiting enzyme tyrosinase is a p53-regulated gene product (
). The major form of DNA protection throughout evolution, however, has been enzymatic repair of the damaged DNA itself. Indeed, prokaryotic bacteria have not only an elaborate basal mechanism of DNA repair but also an inducible component of repair termed the SOS response (
). Specifically, UV irradiation leads to the generation of single-stranded DNA fragments that contain photoproducts, and this DNA forms a complex with the Rec A protein that then cleaves a transcription repressor, increasing the synthesis of approximately 20 bacterial genes that encode DNA repair and cell survival proteins (
). DNA damage thus induces protective responses that render bacteria more resistant to subsequent damage of the same type and hence more likely to survive in the injurious environment.
Experiments suggesting a eukaryotic SOS-like response have been reported sporadically since the 1970s. Using an experimental design similar to that used to demonstrate that prior UV irradiation of bacteria enhances their ability to repair UV-irradiated bacteriophage, human fibroblasts UV-irradiated 4 d prior to infection with UV-irradiated herpes virus were shown to support viral growth approximately twice as well as unirradiated control fibroblasts (
). As in the earlier bacterial experiments, this implied that the host cells had enhanced DNA repair capacity following UV irradiation. Others later confirmed these findings in monkey kidney cells, using SV40 virus (
). Moreover, the enhanced viral DNA repair by irradiated monkey cells was not associated with an increased mutation frequency, as had been observed in bacteria. Subsequently, human fibroblasts subjected either to ionizing radiation or UV irradiation were similarly demonstrated to have enhanced repair of virus DNA that had been damaged by the same insult prior to infection (
) were interpreted by the investigators as “consistent with the central role of single-stranded DNA as an evolutionary conserved signal for DNA damage.” Thus, there is considerable support for the concept that mammalian cells have inducible protective DNA damage responses that are functionally analogous and perhaps even loosely analogous in their molecular mechanism to the bacterial SOS response, despite obvious discrepancies between the eukaryotic and prokaryotic systems.
Mimicking Telomere Disruption and Exposure of the 3′ Overhang
Wishing to test the hypothesis that DNA damage is a major stimulus for UV-induced tanning and unable to obtain a concentrated preparation of CPD for laboratory use, we asked whether exposure of cultured pigment cells or intact skin to thymidine dinucleotide (pTT), the obligatory substrate for the thymine dimers that account for approximately 75% of all UV-induced DNA damage (
), might increase melanogenesis in the absence of UV exposure. Both cultured human melanocytes and intact guinea-pig skin responded to pTT treatment with dramatic increases in melanin production, precisely mimicking UV-induced tanning clinically and histologically (
). The tanning was attributable to increased mRNA and protein expression of tyrosinase, the rate-limiting enzyme in melanogenesis. Subsequent studies demonstrated that the pTT-induced tan in guinea-pig skin was highly photoprotective and thus reproduced the functional as well as clinical and histologic aspects of UV-induced tanning (
), known to mediate many DNA damage responses. pTT-treated human keratinocytes and fibroblasts, transfected with a non-replicating UV-damaged choramphenicol acetyl transferase (CAT) reporter vector that had previously been damaged by UV irradiation, gave rise 24 h later to more than twice the CAT activity (a measure of vector repair) than did diluent-treated control cultures (
). pTT-treated fibroblasts showed a 2–3-fold increase in p53 protein levels and increased p53 activity as detected in an electrophoretic mobility shift assay in which p53 activation is assessed by binding to its consensus sequence in DNA. pTT treatment of human fibroblasts was subsequently shown to similarly increase the protein levels of p53, as well as several p53-regulated DNA repair enzymes over 3–5 d (
Functional consequences of p53 induction and activation could also be observed following pTT treatment, in that survival and clonogenic capacity of pre-treated UV-irradiated cells and rate of repair of CPD all increased appreciably (
). In subsequent experiments, the same CAT vector, treated with benzo[a]pyrene to induce DNA adducts rather than UV irradiated to induce CPDs, was used and vector expression in pTT-pretreated cells was again shown to be twice that in diluent-pre-treated controls (
). Cells derived from young adult patients with basal cell carcinoma versus age-matched controls demonstrated a 5%–8% decrease in ability to repair the vector, and old versus young donor cells demonstrated a 15% decrease, differences in both cases assumed to be causally related to their increased cancer risk (
). Although this can be viewed as evolutionarily beneficial to bacteria, promoting environmental adaptation, mutations in higher organisms such as humans pose unwanted risks to the individual, ranging from compromised tissue function to carcinogenesis. The low-fidelity umuC and umuD bacterial repair enzymes induced during the SOS response (
) have no known human homologs, but it was nevertheless critical to determine whether the pTT-induced enhanced repair capacity increases the mutation rate following DNA damage. Three different host cell reactivation assays using murine cells, intact murine skin, and human fibroblasts demonstrated a substantial reduction in mutation rate in pTT-pre-treated cells (
) and the intuitive notion that an evolutionarily conserved inducible DNA repair capacity would increase the probability of maintaining an intact genome in mammalian cells repeatedly exposed to environmental mutagens Figure 1.
Having established that pTT induces photoprotective tanning and increases DNA repair capacity at least in part via the p53 signaling pathway, we asked whether other oligonucleotides might also be effective in stimulating these potentially therapeutic responses. We found that several other, but certainly not all, oligonucleotides have effects similar to pTT, often at far lower molar equivalent doses, and that activity requires nuclear uptake, a phenomenon that appears to depend on the presence of a 5′ phosphate group (
) and their substantial overlap with the observed pTT effects, we asked whether homology to the telomere 3′ overhang might be the common feature of active oligonucleotides in our assay systems. This proved to be the case. All active oligonucleotides showed greater than 50% homology with the TTAGGG tandem repeat sequence, whereas inactive oligonucleotides failed to meet this criterion. In general, longer sequences (up to at least 20 nucleotides) were more effective than shorter sequences, and the presence of cytosine (C) residues specifically reduced activity (
). Relative activity among test sequences was the same in all assay systems and did not vary with the biologic endpoint used. The shortest effective sequence, pTT, was noted to represent 100% of one-third of the tandem repeat sequence (
), specifically activate DNA damage responses is consistent with the hypothesized common final pathway for experimental telomere loop disruption, cellular senescence at the time of critical telomere shortening, and acute DNA damage responses Figure 2a. Exposure of the otherwise concealed 3′ overhang was known to occur following telomere loop disruption by TRF2DN (
), consistent with its exposure and digestion. Although another technique has instead failed to show overall reduction in telomere overhang length in late-passage senescent cells, the investigators note that loss of the overhang on a minority of chromosome strands would not be detected but might still be sufficient to trigger cell senescence (
). In this context, it is of interest to note that the TTAGGG sequence contains an abundance of preferred targets for these common DNA-damaging agents: one-third of the sequence is TT, a preferred UV target (
). It is thus plausible that preferential damage to telomeres at the time of overall genomic damage might result in telomere disruption, as a consequence either of introducing photoproducts, chemical adducts, or the like, or as a consequence of attempted repair of these lesions. Thus, all three conditions might be expected to expose the TTAGGG repeat sequence, which interestingly does not otherwise appear to occur twice consecutively elsewhere in the human genome (NCBI Human Genome BLAST search [http:www.ncbi.nlm.nih.gov/genome], accessed in 2004), allowing it to interact with a sensor protein or protein complex, which in turn activates of the ATM, ATR, and/or DNA PK kinases with subsequent DNA damage signaling Figure 2b.
Whether or not the hypothesized mechanism is responsible, we have documented extensive parallels between T-oligo treatment and experimental telomere loop disruption, as well as between T-oligo treatment and DNA damage responses following UV irradiation. These include not only tanning and cell cycle arrest but also senescence (
Enhancing DNA repair after UV irradiation while reducing the mutation rate would predict that T-oligo therapy might reduce photocarcinogenesis. To examine this possibility, we used the widely studied SKH1 hairless mouse engineered to express a lacZ mutation reporter transgene and a chronic UV irradiation protocol known to produce squamous cell carcinomas within 6 mo (
). Mice with both wild-type DNA repair capacity and mice heterozygous for deletion of the XPC repair enzyme were used to model the broad range of the DNA repair capacity observed in the human population. pTT or diluent alone was applied daily to the back for the first 5 d of each month, and then the mice were UV irradiated 5 d per wk for the remaining 3 wk of each month, for a total of 6 mo. The UV-irradiated mice progressively developed skin tumors as expected, and at the end of 6 mo only 12% of the control vehicle-treated mice were tumor free. Seventy-eight percent of the intermittently pTT-treated mice, however, were tumor free. Histologic analysis revealed both actinic keratosis-like lesions and invasive squamous cell carcinomas, with each tumor type being more prevalent in the vehicle-treated than in the pTT-treated mice. There was no evidence of pTT toxicity in either the irradiated mice or in sham-irradiated controls. Assessment of the lac-Z/pUR288 reporter plasmid showed a decreased mutation rate in pTT-treated versus control mice after either a single UV exposure or after 6 mo of intermittent irradiation, confirming the previous in vitro data (see above). These data suggest that topical T-oligo therapy might be a valuable adjunct to sunscreen use and sun avoidance in patients at high risk of photocarcinogenesis.
Our initial work with T-oligos fully homologous to the 3′ overhang demonstrated apoptosis in multiple established cell lines (
), including melanoma cells. Because apoptosis is a major mechanism by which chemotherapeutic agents reduce tumor burden, and resistance to apoptosis is a well-documented mechanism of treatment resistance (
), we asked whether T-oligos might also have a role in treating established malignancies. As a first experimental model in which to determine whether T-oligos can serve as an effective cancer therapy, we selected melanoma, the most fatal of skin cancers and a malignant cell type characterized by resistance to conventional cancer therapy once metastasis has occurred. In preliminary experiments, several human melanoma cell lines were examined and found to readily undergo apoptosis within 72–96 h of exposure to an 11-base T-oligo (pGTTAGGGTTAG). For further study, we selected the aggressive MM-AN line (
). T-oligos were added once to culture medium at time 0, and the cells were examined daily for 4 d. By 96 h, two-thirds of the MM-AN cells were undergoing apoptosis, mediated at least in part by the p53 homolog p73 (
), suggesting that T-oligo-induced apoptosis preferentially affects malignant cells. Apoptosis of the MM-AN cells was explained in part by marked downregulation of livin/ML-IAP, a member of the inhibitor of apoptosis family of proteins known to be expressed normally during embryogenesis and inappropriately re-expressed in many melanomas, thereby contributing to their resistance to chemotherapy-induced apoptosis (
). T-oligo-treated MM-AN cells also showed upregulation of the melanogenic proteins tyrosinase and TRP1, as well as the other differentiation markers gp100 and MART1, most prominently after 72–96 h when many of the cells had already undergone apoptosis. To determine whether these effects could be seen in vivo, the immunocompromised SCID mouse model was used. T-oligo dramatically reduced tumor burden when MM-AN cells were treated either prior to tail vein injection, in a protocol that yielded numerous metastases in control animals, or following intralesional or intraperitoneal injection in animals with MM-AN cells implanted either subcutaneously or intraperitoneally (
). Small residual tumors in the T-oligo-treated animals contained cells undergoing apoptosis and/or showing evidence of differentiation by immunostaining, whereas control tumors did not. As in the cancer prevention studies, there was no evidence of T-oligo toxicity following either systemic or local injection daily for up to 26 d. These data suggest that T-oligos may provide a safe and effective novel means of treating advanced melanoma.
Summary and Conclusions
Work in many laboratories over the past decade has established a central role for the telomere in maintaining genomic integrity. Available data may be interpreted to indicate that telomere disruption, whether due to acute DNA damage or progressive telomere shortening, is the initial event that triggers multiple DNA damage responses. The specific initiating event is likely exposure of the otherwise concealed single-stranded 3′ overhang, tandem repeats of TTAGGG, a signal that can be provided to cells in the absence of DNA damage by exogenously provided T-oligos. The ability of T-oligo treatment to trigger SOS-like responses and/or to cause selective apoptosis of already malignantly transformed cells may provide an important new means of cancer prevention and treatment.
This work was supported in part by NIH (grant.R01CA1C5156) and the Carter Family, Foundation.