Our Research
We have a wide range of projects that are broadly grouped in the following two themes:
Regulation of DSB repair pathways and the impact on cancer therapy
Damage to DNA is induced daily from metabolites and occurs upon exposure to ionizing radiation. One type of DNA damage, double-strand DNA breaks (DSBs), is particularly toxic to cells—a single DSB is enough to cause cellular lethality if left unrepaired. Cells have evolved factors that act in two major mechanistically distinct pathways to repair DSBs. One pathway uses the sister chromatid available during S phase as template to synthesize DNA to replace the damage, termed homologous recombination (HR). Another pathway, termed non-homologous end joining (NHEJ), molecularly joins the two ends of the DNA break during G1, when sister chromatid is not available. The relative contribution of these two competing pathways differ in different cell types and in different phases of the cell cycle. The balance of pathway choices between these two pathways is therefore critical for maintaining genomic stability.
Two key DNA repair molecules, 53BP1 and BRCA1, compete to facilitate NHEJ and HR, respectively. A decisive factor in the choice between DSB repair pathways is in the competition between DNA end protection that is necessary for NHEJ, and DNA end resection that is necessary for HR. DSB end resection that facilities HR pathway must be appropriately restricted to S/G2 phases of the cell cycle, when the intact sister chromatid is available. Depletion of NHEJ promoting factors such as 53BP1 allows DNA end resection in the G1 phase, thereby impairing DSB repair and causing genomic instability. Conversely, loss of the HR protein, BRCA1 (critical for initiating end resection) allows the error-prone NHEJ pathway to dominate throughout the cell cycle potentially also contributing to mutations/deletions.
Inhibitors of the DNA repair protein poly ADP-ribose polymerase (PARP) have been an effective therapy for cancer patients with BRCA1-deficiency, since PARP inhibition induces DSB that requires BRCA1-mediated HR to repair the damage. However, over time these cancers acquire resistance to PARP inhibitor, by losing 53BP1 or pro-NHEJ associated factors (RIF1, PTIP) that consequently restore resection and the activation of HR pathway.
Our aim is to identify and understand biomolecules that modulate the activities of 53BP1 and BRCA1, thereby influencing the cellular choice between HR and NHEJ repair pathway under various physiological conditions.
(a) MicroRNAs: We have conducted functional screens to discover microRNA (miRNA)s that down-modulate DSB repair proteins, and influence specific repair pathways. Ectopic expression of NHEJ-regulating miRNAs in BRCA1-deficient ovarian causes PARP inhibitor resistance. In turn these patients have poor prognosis. We now utilize focused sequencing of miRNAs from BRCA deficient tumors that are resistant to therapy to identify miRNAs that influence the DNA repair machinery. Importantly these miRNAs have significant clinical relevance as therapeutic targets or predictive biomarkers of response.
(b) Non-coding RNAs (ncRNAs) and RNA binding proteins: NHEJ-promoting activities of 53BP1 is dependent on its recruitment to DSBs via the interaction of its tandem Tudor domain with dimethylated lysine 20 of histone H4 (H4K20me2) on the damage chromatin. We have identified a novel RNA binding protein, TIRR (Tudor Interacting Repair Regulator) that directly binds the tandem Tudor domain of 53BP1, and masks its H4K20me2 binding motif. Over-expression of TIRR impedes 53BP1 function by blocking its localization to damage site on the chromatin, while depletion of TIRR destabilizes 53BP1 in the nuclear soluble fraction and alters the DSB-induced protein complex centering 53BP1 to broadly impede repair. Since TIRR is an RNA binding protein, we are uncovering novel non-coding RNAs that associate with TIRR. Our aims are to define the precise mechanism by which these ncRNAs associate with TIRR and functionally impact DSB repair.
(c) Phosphatases: While 53BP1-promoting NHEJ pathway repairs DSBs mainly in G1, and BRCA1-promoting HR repairs DSB mainly in S/G2, intriguingly during mitosis, the DSB are recognized but not repaired. Recent phosphoproteomic studies in mitotic cells have revealed most proteins, including DDR factors, are hyperphosphorylated in mitosis, thereby inhibiting their activities. Concordantly, we have observed that phosphorylation of 53BP1 in its chromatin-binding region during mitosis impedes its recruitment to chromatin. Only through dephosphorylation mediated by the serine/threonine protein phosphatase complex PP4C/R3Beta at these sites can 53BP1 activities be restored in G1. Ectopic recruitment of 53BP1, thus activation of NHEJ pathway during mitosis intriguingly causes genome instability. In collaboration with David Pellman’s laboratory at Harvard Medical School, we are using a combination of cytological tools and single-cell sequencing to elucidate how dysregulated phosphodynamics lead to the ectopic activation of DSB repair during mitosis that ultimately result in genome instability. Furthermore, we have developed phospho-proteomic strategies to systematically investigate how phosphatases facilitates DSB repair.
(d) Systematic identification of factors that influence the BRCA-pathway: In collaboration with the Broad Institute, we employed a whole genome CRISPR library to comprehensively identify factors that restore HR-mediated DSB repair in BRCA1-deficient ovarian cancer cell lines that render these cells resistant to PARP inhibitors. Characterizing the factors identified from the screen will provide insight to how HR cross-talk with other signaling pathways and elucidate the mechanism of chemo-resistance.
II. Clinical applicability of serum miRNAs as biomarkers
Of the many classes of circulating nucleic acids, miRNAs are remarkably stable against various physiological and physical conditions, and are readily detectable and quantifiable by rapid and sensitive PCR approach in a high throughput manner. These characteristics of miRNAs rendered them ideal noninvasive biomarkers for various pathological conditions. We are focusing on identifying and establishing miRNA expression signatures as potential diagnostic tools for the following clinical challenges:
(a) Radiation response: The exposure of human populations to radiation, either accidentally (e.g. nuclear plant incident) or intentionally (e.g. terrorism) poses a significant threat to public health. One of the most important steps in the medical management of a nuclear disaster is to readily identify individuals whose exposure to dose of radiation that has caused significant injury to internal organs and tissues. Existing biodosimetry techniques and devices do not predict the severity of injury sustained by specific organs and tissues, and thus do not allow for the prompt organ- and tissue-targeted medical treatment. We have identified a set of serum miRNAs whose expression in response to radiation predicts long-term radiation-induced hematopoietic injury in mice. We have demonstrated in non-human primates (NHPs) this response mechanism is evolutionarily conserved and thus that serum miRNA signatures can potentially serve biomarkers to predict radiation-induced fatality in a human population. Genomic analysis of these radiation-responsive conserved miRNAs revealed that a common transcriptional network regulates these miRNAs in human, mice, and NHPs. From the clinical perspective, peripheral blood cells from radiation-therapy patients also show radiation-induced changes in miRNA expression. We also observed that differential expression of serum miRNAs may reflect biological tumor response (or resistance) to thoracic radiation in non-small-cell-lung cancer (NSCLC) patients.
(b) Predicting radiation-induced mortality: In animal models we have evidence that distinct sets of serum miRNAs correlates with radiation dose, and also the time-frame after radiation exposure (that is days versus weeks). Overall the goal is to build on our preliminary studies and establish miRNA signatures that predict radiation injury to allow for timely and appropriate treatment of radiation victims.
(c) Radiation therapy-induced secondary malignancy: Although radiation has long been an effective and popular mode of cancer therapeutics, it can damage normal tissues and often cause subsequent tumors, especially among pediatric patients. To improve life quality of these patients and the effectiveness of radiation therapy regimen, we need to identify potential noninvasive biomarkers to detect secondary malignancy early in the treatment process. We are currently identifying serum miRNAs drawn from the blood of patients who have recently undergone irradiation therapy, that will allow us to predict patient susceptibility to long-term radiation-induced cancers.
(d) Early detection of ovarian cancer: Ovarian cancer is one of the most lethal tumors due to difficulty in catching the signs of cancer during early stage. The five-year survival rate for all stages of ovarian cancer is 47%. However, for cases where a diagnosis of the disease is made early, when the cancer is still confined to the primary site, the five-year survival rate is 92.7%. Patient survival is therefore critically dependent on availability of non-invasive diagnostic tools. We have found a set of serum miRNAs whose targets serve as highly accurate predictor for invasive ovarian cancer. From these miRNA signature, we can distinguish borderline malignant tumors from invasive tumors. We are currently examining the exact mechanism by which these miRNA signals ovarian cancer malignancy.
Regulation of DSB repair pathways and the impact on cancer therapy
Damage to DNA is induced daily from metabolites and occurs upon exposure to ionizing radiation. One type of DNA damage, double-strand DNA breaks (DSBs), is particularly toxic to cells—a single DSB is enough to cause cellular lethality if left unrepaired. Cells have evolved factors that act in two major mechanistically distinct pathways to repair DSBs. One pathway uses the sister chromatid available during S phase as template to synthesize DNA to replace the damage, termed homologous recombination (HR). Another pathway, termed non-homologous end joining (NHEJ), molecularly joins the two ends of the DNA break during G1, when sister chromatid is not available. The relative contribution of these two competing pathways differ in different cell types and in different phases of the cell cycle. The balance of pathway choices between these two pathways is therefore critical for maintaining genomic stability.
Two key DNA repair molecules, 53BP1 and BRCA1, compete to facilitate NHEJ and HR, respectively. A decisive factor in the choice between DSB repair pathways is in the competition between DNA end protection that is necessary for NHEJ, and DNA end resection that is necessary for HR. DSB end resection that facilities HR pathway must be appropriately restricted to S/G2 phases of the cell cycle, when the intact sister chromatid is available. Depletion of NHEJ promoting factors such as 53BP1 allows DNA end resection in the G1 phase, thereby impairing DSB repair and causing genomic instability. Conversely, loss of the HR protein, BRCA1 (critical for initiating end resection) allows the error-prone NHEJ pathway to dominate throughout the cell cycle potentially also contributing to mutations/deletions.
Inhibitors of the DNA repair protein poly ADP-ribose polymerase (PARP) have been an effective therapy for cancer patients with BRCA1-deficiency, since PARP inhibition induces DSB that requires BRCA1-mediated HR to repair the damage. However, over time these cancers acquire resistance to PARP inhibitor, by losing 53BP1 or pro-NHEJ associated factors (RIF1, PTIP) that consequently restore resection and the activation of HR pathway.
Our aim is to identify and understand biomolecules that modulate the activities of 53BP1 and BRCA1, thereby influencing the cellular choice between HR and NHEJ repair pathway under various physiological conditions.
(a) MicroRNAs: We have conducted functional screens to discover microRNA (miRNA)s that down-modulate DSB repair proteins, and influence specific repair pathways. Ectopic expression of NHEJ-regulating miRNAs in BRCA1-deficient ovarian causes PARP inhibitor resistance. In turn these patients have poor prognosis. We now utilize focused sequencing of miRNAs from BRCA deficient tumors that are resistant to therapy to identify miRNAs that influence the DNA repair machinery. Importantly these miRNAs have significant clinical relevance as therapeutic targets or predictive biomarkers of response.
(b) Non-coding RNAs (ncRNAs) and RNA binding proteins: NHEJ-promoting activities of 53BP1 is dependent on its recruitment to DSBs via the interaction of its tandem Tudor domain with dimethylated lysine 20 of histone H4 (H4K20me2) on the damage chromatin. We have identified a novel RNA binding protein, TIRR (Tudor Interacting Repair Regulator) that directly binds the tandem Tudor domain of 53BP1, and masks its H4K20me2 binding motif. Over-expression of TIRR impedes 53BP1 function by blocking its localization to damage site on the chromatin, while depletion of TIRR destabilizes 53BP1 in the nuclear soluble fraction and alters the DSB-induced protein complex centering 53BP1 to broadly impede repair. Since TIRR is an RNA binding protein, we are uncovering novel non-coding RNAs that associate with TIRR. Our aims are to define the precise mechanism by which these ncRNAs associate with TIRR and functionally impact DSB repair.
(c) Phosphatases: While 53BP1-promoting NHEJ pathway repairs DSBs mainly in G1, and BRCA1-promoting HR repairs DSB mainly in S/G2, intriguingly during mitosis, the DSB are recognized but not repaired. Recent phosphoproteomic studies in mitotic cells have revealed most proteins, including DDR factors, are hyperphosphorylated in mitosis, thereby inhibiting their activities. Concordantly, we have observed that phosphorylation of 53BP1 in its chromatin-binding region during mitosis impedes its recruitment to chromatin. Only through dephosphorylation mediated by the serine/threonine protein phosphatase complex PP4C/R3Beta at these sites can 53BP1 activities be restored in G1. Ectopic recruitment of 53BP1, thus activation of NHEJ pathway during mitosis intriguingly causes genome instability. In collaboration with David Pellman’s laboratory at Harvard Medical School, we are using a combination of cytological tools and single-cell sequencing to elucidate how dysregulated phosphodynamics lead to the ectopic activation of DSB repair during mitosis that ultimately result in genome instability. Furthermore, we have developed phospho-proteomic strategies to systematically investigate how phosphatases facilitates DSB repair.
(d) Systematic identification of factors that influence the BRCA-pathway: In collaboration with the Broad Institute, we employed a whole genome CRISPR library to comprehensively identify factors that restore HR-mediated DSB repair in BRCA1-deficient ovarian cancer cell lines that render these cells resistant to PARP inhibitors. Characterizing the factors identified from the screen will provide insight to how HR cross-talk with other signaling pathways and elucidate the mechanism of chemo-resistance.
II. Clinical applicability of serum miRNAs as biomarkers
Of the many classes of circulating nucleic acids, miRNAs are remarkably stable against various physiological and physical conditions, and are readily detectable and quantifiable by rapid and sensitive PCR approach in a high throughput manner. These characteristics of miRNAs rendered them ideal noninvasive biomarkers for various pathological conditions. We are focusing on identifying and establishing miRNA expression signatures as potential diagnostic tools for the following clinical challenges:
(a) Radiation response: The exposure of human populations to radiation, either accidentally (e.g. nuclear plant incident) or intentionally (e.g. terrorism) poses a significant threat to public health. One of the most important steps in the medical management of a nuclear disaster is to readily identify individuals whose exposure to dose of radiation that has caused significant injury to internal organs and tissues. Existing biodosimetry techniques and devices do not predict the severity of injury sustained by specific organs and tissues, and thus do not allow for the prompt organ- and tissue-targeted medical treatment. We have identified a set of serum miRNAs whose expression in response to radiation predicts long-term radiation-induced hematopoietic injury in mice. We have demonstrated in non-human primates (NHPs) this response mechanism is evolutionarily conserved and thus that serum miRNA signatures can potentially serve biomarkers to predict radiation-induced fatality in a human population. Genomic analysis of these radiation-responsive conserved miRNAs revealed that a common transcriptional network regulates these miRNAs in human, mice, and NHPs. From the clinical perspective, peripheral blood cells from radiation-therapy patients also show radiation-induced changes in miRNA expression. We also observed that differential expression of serum miRNAs may reflect biological tumor response (or resistance) to thoracic radiation in non-small-cell-lung cancer (NSCLC) patients.
(b) Predicting radiation-induced mortality: In animal models we have evidence that distinct sets of serum miRNAs correlates with radiation dose, and also the time-frame after radiation exposure (that is days versus weeks). Overall the goal is to build on our preliminary studies and establish miRNA signatures that predict radiation injury to allow for timely and appropriate treatment of radiation victims.
(c) Radiation therapy-induced secondary malignancy: Although radiation has long been an effective and popular mode of cancer therapeutics, it can damage normal tissues and often cause subsequent tumors, especially among pediatric patients. To improve life quality of these patients and the effectiveness of radiation therapy regimen, we need to identify potential noninvasive biomarkers to detect secondary malignancy early in the treatment process. We are currently identifying serum miRNAs drawn from the blood of patients who have recently undergone irradiation therapy, that will allow us to predict patient susceptibility to long-term radiation-induced cancers.
(d) Early detection of ovarian cancer: Ovarian cancer is one of the most lethal tumors due to difficulty in catching the signs of cancer during early stage. The five-year survival rate for all stages of ovarian cancer is 47%. However, for cases where a diagnosis of the disease is made early, when the cancer is still confined to the primary site, the five-year survival rate is 92.7%. Patient survival is therefore critically dependent on availability of non-invasive diagnostic tools. We have found a set of serum miRNAs whose targets serve as highly accurate predictor for invasive ovarian cancer. From these miRNA signature, we can distinguish borderline malignant tumors from invasive tumors. We are currently examining the exact mechanism by which these miRNA signals ovarian cancer malignancy.