Abstract PD4-07: PET imaging of PARP-1 expression in breast cancer

18F-FluorThanatrace ([18F]-FTT) is a novel radiotracer shown to quantify Poly [ADP-ribose] polymerase 1 (PARP-1) expression in vitro and in vivo through a receptor-ligand interaction. A recent study at the University of Pennsylvania in women with ovarian cancer demonstrated in vivo visualization of PARP-1 expression in tumors using this radiotracer that closely correlated with an in vitro assay of PARP-1 in tumor tissue (Makvandi, M. J. Clin. Invest. 128:2116, 2018). A radioligand with PARP-1 specificity, [125I]-KX1, was also developed as a companion tool for ex vivo evaluation of PARP-1 expression and PARP inhibitor (PARPi) drug occupancy by radioligand binding assay (Makvandi, M. Cancer Res. 76:4516, 2016). As the first step in validating this biomarker in breast cancer, we performed a prospective clinical trial comparing in vivo [18F]-FTTuptake and ex vivo PARP-1 expression in women with primary breast cancer.

Methods: 24 patients with Stage I-IV primary breast cancer were imaged with [18F]-FTT prior to any therapy including surgery. We correlated in vivo uptake with ex vivo immunohistochemistry (IHC) for PARP-1 and [125I]-KX1 autoradiography in untreated surgical specimens. Tumors were analyzed for alterations in DNA repair genes, copy number-based as well as mutational signatures indicative of homologous recombination deficiency (HRD) and mutational burden, using our established protocol (Maxwell, KN, Nature Commun. 8:319, 2017).

Results: [18F]-FTT uptake was visualized above background in all primary breast tumors and known metastases. Two areas of unexpected uptake revealed an unknown contralateral breast cancer and an ovarian carcinoid, respectively. We expected that uptake might be highest in triple negative breast cancer (TNBC), where PARPi have been most heavily studied. However, a range of tracer uptake was observed in tumors independent of breast cancer subtype (hormone receptor positive/HER2 negative, TNBC, HER2+) and BRCA status. Uptake ratios (SUVmax tumor/SUV max opposite breast) ranged from 1.2-10.5 with a median 4.0. Ex vivo[125I]-KX1 autoradiography was performed on a subset of untreated primary tumors (n=5) and compared with IHC staining for PARP-1 on sequential sections. This revealed a close spatial correspondence between elevated PARP-1 expression by IHC and regions of elevated [125I]-KX1 binding radiographically. There was also a strong positive correlation between in vivo [18F]-FTT uptake and ex vivo quantitative [125I]-KX1 autoradiography (r=0.78). Genomic analysis of HRD in all tumors is pending and will be reported.

Conclusion: Initial analyses support the ability of [18F]-FTT to visualize and measure PARP-1 expression in breast cancer. This is the first step toward developing an imaging companion diagnostic to help guide PARP inhibitor treatment in breast cancer. Ongoing studies are expanding upon these results, testing the extent to which expression of PARP-1 by [18F]-FTT can predict response to PARP inhibitors and measure target engagement during therapy.

Citation Format: McDonald ES, Carlin S, Maxwell KN, Nayak A, Doot RK, Pantel AR, Farwell MD, Pryma DA, Clark AS, Shah P, DeMichele AM, Ziober A, Schubert EK, Palmer K, Lee HS, Matro J, de la Cruz L, Tchou J, Anderson DN, Feldman MD, Sheffer RE, Knollman H, Schnall MD, Makvandi M, Domchek S, Hubbard RA, Mach RH, Mankoff DA. PET imaging of PARP-1 expression in breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr PD4-07.

Longitudinal optical monitoring of blood flow in breast tumors during neoadjuvant chemotherapy

We measure tissue blood flow markers in breast tumors during neoadjuvant chemotherapy and investigate their correlation to pathologic complete response in a pilot longitudinal patient study (n  =  4). Tumor blood flow is quantified optically by diffuse correlation spectroscopy (DCS), and tissue optical properties, blood oxygen saturation, and total hemoglobin concentration are derived from concurrent diffuse optical spectroscopic imaging (DOSI). The study represents the first longitudinal DCS measurement of neoadjuvant chemotherapy in humans over the entire course of treatment; it therefore offers a first correlation between DCS flow indices and pathologic complete response. The use of absolute optical properties measured by DOSI facilitates significant improvement of DCS blood flow calculation, which typically assumes optical properties based on literature values. Additionally, the combination of the DCS blood flow index and the tissue oxygen saturation from DOSI permits investigation of tissue oxygen metabolism. Pilot results from four patients suggest that lower blood flow in the lesion-bearing breast is correlated with pathologic complete response. Both absolute lesion blood flow and lesion flow relative to the contralateral breast exhibit potential for characterization of pathological response. This initial demonstration of the combined optical approach for chemotherapy monitoring provides incentive for more comprehensive studies in the future and can help power those investigations.

Abstract P6-10-08: Preliminary results of a first in human Phase 1 clinical trial to demonstrate safety and feasibility of chimeric antigen receptor T (CART) cells directed against c-Met in the treatment of breast cancer

This abstract was withdrawn by the authors.

Abstract PD1-04: Association of aspirin and clinical outcomes in patients with invasive breast cancer

Long-term low-dose aspirin use has been observed to reduce the risk of colorectal, breast and other cancers. The most prominent effect has been in colorectal cancer, in which large-scale meta-analyses have shown that there is an approximately 20% relative risk reduction in participants who took aspirin for four or more years. The role of long-term NSAID use in breast cancer risk is less clear although preliminary observational case-control studies suggest an association between aspirin use and reduced incidence of hormone receptor-positive breast cancers though no clear evidence exists to support a clear mortality benefit among patients with a history of prior NSAID use as opposed to those who do not.

To investigate whether a history of aspirin use is associated with improved clinical outcome in breast cancer, we examined the pattern of aspirin use, cancer pathology and overall survival of over 1000 patients diagnosed with and treated for invasive breast cancer at our institution, for whom long-term follow up was available. A history of aspirin use for at least a period of 30 days prior to breast cancer diagnosis was reported in nearly 14% of individuals. Aspirin use was associated with being older than the age of 50 at diagnosis (79.8% vs 66.5%; Fisher's Exact Test (P < 3.2x10-3) and being of African American race (49.1% vs 28.7%; P < 3.4x10-2), when compared to those who have not used aspirin.

Aspirin use correlated with prognostic factors that are known to be associated with poor outcomes. They include axilla node positive disease (44.5% vs 27.0%, p< 0.032), evidence of lymphovascular invasion (24.7% vs 15.4%, p< 0.049), Her2-neu positive disease (<0.0083). In contrast to prior retrospective case-control studies, no significant association between aspirin use and hormone receptor positive disease was noted for either ER (p=0.19) or PR(+) receptor status (p=0.12). Finally, we examined if aspirin use prior to breast cancer diagnosis has any impact on disease outcome. Over a median follow up of 60.0 months, univariate analysis using cox proportional hazard modeling demonstrated that the use of low-dose aspirin prior to the diagnosis of breast cancer was associated with an increased all-cause mortality when compared to patients without aspirin use prior to cancer diagnosis (HR=3.084, 95% CI=1.961 to 4.848). On multivariate analysis, we found that recent history of aspirin use was significantly associated with a worse overall survival (HR 2.65; 95%CI 1.37 -5.12, P < 3.77 x 10-3), when controlled for other prognostic factors including receptor status, tumor size, tumor grade, number of positive regional lymph nodes, positive margins, as well as race and age at diagnosis.

This is the first study to report on the association of aspirin use with breast cancer outcomes in a large patient cohort treated at a single institution. Although aspirin in breast and cancers has been associated with reduced cancer incidence, a history of aspirin use prior to breast cancer diagnosis does not appear to be protective or associated with improve clinical outcomes or survival among breast cancer patients. Ongoing efforts are examining the mechanism underlying this association.

Citation Format: Li YR, Steel L, Carrigan E, Tchou J. Association of aspirin and clinical outcomes in patients with invasive breast cancer. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr PD1-04.

Abstract P3-06-01: Association between metformin use and improved survival in breast cancer patients: A population-based study

Background: Long-term use of AMP-kinase agonist Metformin, the first-line oral hypoglycemic agent, has recently been shown to be associated with both a decreased incidence of and improved survival following diagnosis of breast and other solid tumors. However, the sample size of most reported studies has been limited and has not been able to clearly delineate if there are molecular subtype-dependent effects.

Method: We implemented a population-based, retrospective case-control analysis of the effect of metformin on breast cancer outcomes. We comprehensively searched the electronic medical records of all female patients who were cared for within the University of Pennsylvania Health System between the years 1997 and 2013. To ensure the accuracy of patient selection, we focused our preliminary analysis on only patients (n = 20,684) who received an ICD9 code for breast cancer (BC) on at least 2 separate in-person visits. We assessed the use of metformin by identifying those patients who have at least a record of a metformin prescription both before the first breast cancer diagnosis ICD9 code and after the first breast cancer diagnosis visit. Overall survival was determined by assessing a combination of electronic medical record of patient expiration and the SSN Death Certificate Index. Statistical tests, including chi-sq and cox-hazard modeling were implemented in R, Graphpad Prism or EXCEL.

Results: In the preliminary analysis, we identified a total of 29,251 female patients either diagnosed or reported a history of BC within the UPHS EMRs. Stratifying by per patient visits, we limited the subsequent analysis to the 20,684 BC patients with at least 2 in-person visits where BC was recorded as a diagnosis code. Of these, there were 534 patients who had a record of metformin prescription prior to and after the first BC diagnosis code and 20,150 BC patients who did not. We found that with a mean follow up of (6.40 years), there were a total of 8 deaths (1.28%) in metformin users and 526 deaths (2.62%) among the no metformin users. Using a two-tailed chi-sq test, we found that metformin use was significantly associated with improved overall survival rates (98.50 vs 96.94%, p<0.05). We are currently assessing the role of age, ethnicity, the presence of diabetes, health metrics related to insulin resistance, and the use of other metabolic regulators (insulin, sulfonylurea) using a cox-hazard proportional risk model.

Conclusion: Our preliminary analyses suggest that metformin use before and following diagnosis may be protective against overall mortality among women diagnosed with breast cancer. A more refined Cox proportional hazard modelling is ongoing, which will shed light on how factors such as the use of other metabolic agents, age, and breast cancer subtype may interact with metformin's effect on overall and disease-free BC survival.

Citation Information: Cancer Res 2013;73(24 Suppl): Abstract nr P3-06-01.

Imaging the redox states of human breast cancer core biopsies

Currently, the gold standard to establish benign vs. malignant breast tissue diagnosis requires an invasive biopsy followed by tissue fixation for subsequent histopathological examination. This process takes at least 24 h resulting in tissues that are less suitable for molecular, functional, or metabolic analysis. We have recently conducted redox scanning (cryogenic NADH/flavoprotein fluorescence imaging) on snap-frozen breast tissue biopsy samples obtained from human breast cancer patients at the time of their breast cancer surgery. The redox state was readily determined by the redox scanner at liquid nitrogen temperature with extraordinary sensitivity, giving oxidized flavoproteins (Fp) an up to tenfold discrimination of cancer to non-cancer of breast in our preliminary data. Our finding suggests that the identified metabolic parameters could discriminate between cancer and non-cancer breast tissues without subjecting tissues to fixatives. The remainder of the frozen tissue is available for additional analysis such as molecular analysis and conventional histopathology. We propose that this novel redox scanning procedure may assist in tissue diagnosis in ex vivo tissues.

P1-03-09: Significance of FAP, SMA and CD31 Expression in the Stroma of Breast Cancer

Breast cancer stroma heterogeneity has been demonstrated in various gene expression profile analyses. Whether there is any association between stroma heterogeneity and the molecular phenotype of breast cancer has yet to be established. Therefore, we performed immunohistochemical analyses (IHC) to evaluate the expression of the following stromal cell markers (fibroblast activation protein (FAP), smooth muscle actin (SMA), and CD31, an endothelial cell marker) in tumor tissues from a contemporary cohort of 52 patients comprising of all four molecular subtypes (luminal A (n=25); luminal B (n=2); Her2-neu (+) (n=5); and basal (n=20)). We hypothesize that stroma heterogeneity as reflected by the proportion of stromal cells staining (+) for FAP and SMA may correlate with their molecular epithelial phenotype. Furthermore, studying the distribution of these stromal cell markers in IHC sections may evaluate their spatial relationship with tumor cells, immune cells, and tumor microvasculature which may have strategic significance within the tumor/microenvironment.

Differentiation of benign and malignant breast lesions by in-vivo three-dimensional diffuse optical tomography

Abstract #805

Background: Diffuse optical tomography (DOT) uses near-infrared light to non-invasively image total hemoglobin concentration and blood oxygen saturation in the human breast. Given its low cost, ease of use, and possibility of repeated measured over time, DOT is a promising adjunctive imaging modality for screening, diagnosis and monitoring of neoadjuvant therapy. In this study we explored the performance of DOT to differentiate benign and malignant breast lesions.
 Method and Materials: Forty-seven women with clinical or mammographic abnormalities were prospectively recruited for DOT. Most patients underwent gadolinium-enhanced MRI examination. Three-dimensional oxy-, deoxy-hemoglobin, total hemoglobin concentration, blood oxygen saturation and scattering coefficient images of each breast were reconstructed. Tumor-to-normal (T/N) ratios of these parameters were computed by defining tumor regions with guidance from MRI and radiology reports. In addition, optical index was constructed based on these parameters to maximize the T/N contrast. Only the biopsy-proven lesions were selected (51 breast lesions) and classified into three groups: benign lesions (N=10), malignant lesions where DOT preceded core biopsy (N=20) and malignant lesions where DOT was performed after core-biopsy (N=21). We fit a mixed effects model that estimated the mean optical T/N ratios and optical index for each group, and using the resulting standard errors developed 95% confidence intervals and tested the hypothesis that each optical contrast parameter was unity.
 Results: Malignant cancers showed statistically significant higher total hemoglobin concentration, scattering, oxy-hemoglobin concentration and optical index (P=0.01-0.04) compared to normal tissue. Furthermore, malignant lesions exhibited a two-fold average increase in an optical index derived from the endogenous optical parameters (95% CI: 1.4 - 2.4). To test whether bleeding due to core biopsy influence DOT results, we compared if there was statistically significant differences between two groups measured before or after core-biopsy. There were no statistically significant differences in these groups, suggesting that post biopsy hemorrhage did not alter the DOT results. Benign tumors did not show statistical significance in all of the T/N ratios. AUC of total hemoglobin concentration, scattering, oxy-hemoglobin and optical index suggested good discriminatory power with values between 0.90 and 0.99.
 Discussion: The data demonstrates the feasibility of differentiating benign and malignant lesions by quantitative three-dimensional DOT when the tumor location information is provided by other imaging modality. The main drawback of this study is the small number of benign lesions, which warrants further study. DOT technology is still at its developing stage and needs more investigation to find its niche in breast imaging.

Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 805.

Contralateral prophylactic mastectomy in affected carriers of deleterious BRCA1 or BRCA2 mutations: Does timing of genetic testing matter?

10512

Introduction: Prophylactic mastectomy (PM) reduces breast cancer risk by >90% in women who are carriers of deleterious BRCA1 or BRCA2 mutations. Genetic testing prior to breast cancer surgery has been reported to affect PM decision in high-risk patients. The ideal timing of genetic testing as well as the clinical characteristics of affected mutation carriers electing PM remain unclear. This is a pilot study to identify significant clinical characteristics associated with affected carriers who had undergone PM. Methods: Retrospective chart review was performed on 103 breast cancer affected BRCA1 or BRCA2 mutation carriers that were seen at our high risk clinic who had undergone genetic testing between 1995 and 2005. Clinical characteristic, initial surgery treatment modalities, and dates of genetic testing and prophylactic mastectomy were collected and analyzed using the chi-square or Fisher exact tests. Results: Of the 103 affected mutation carriers, 30 (29%) underwent prophylactic mastectomy (PM) where as 73 (71%) did not (no PM). Ethnicity, age of diagnosis, tumor size, nodal status, family history and initial breast cancer surgery types (BCT vs. mastectomy) were not significantly different between the two groups. Of the 30 women who underwent PM, 19 (63%) vs. 9 (30%) underwent PM before and after their genetic testing respectively. Of the 19 women who had PM before their genetic testing, 4 (21%) had BCT as initial treatment whereas 15 (78.9%) had mastectomy as initial treatment and 9 of 15 (60%) had synchronous contralateral prophylactic mastectomy. For the 9 women who had PM after their genetic testing, 5 (56%) had BCT vs. 4 (44%) had mastectomy as their initial surgical treatment (p < 0.001). Conclusion: In this study, we found a significant correlation between the initial breast cancer surgical modality of mastectomy with women undergoing prophylactic mastectomy. The impact of family history of breast cancer does not appear to be significant. Physician recommendations or patient preference are unknown. Our data suggest that high risk women who elect mastectomy as their initial surgical management to treat their breast cancer are more likely to undergo prophylactic mastectomy regardless of knowledge of their mutation status.

No significant financial relationships to disclose.

A mammographic dilemma: calcification or haemosiderin as a cause of opacities? Validation of a new digital diagnostic tool

Core biopsies of an area of microcalcification demonstrated large collections of macrophages containing haemosiderin, with evidence of minimal microcalcification on H&E staining. Algorithms were developed that were capable of differentiating with high accuracy those signs due to calcification, using quantitative measurements such as the apparent volume composition of calcium. Using the linear attenuation coefficients of calcification and assuming an ellipsoid model for the 3-dimensional shape of calcification, we computed the relative calcification volume for each region of interest. The difference in the linear attenuation coefficients of iron and calcification allowed the two to be differentiated on a mammogram based on this measure of relative calcification volume.

The catalytic mechanism of Fpg protein. Evidence for a Schiff base intermediate and amino terminus localization of the catalytic site

Our recent structure-activity analysis of Fpg protein of Escherichia coli, using oligodeoxynucleotides containing various 8-oxopurine derivatives, has allowed us to postulate an enzyme mechanism involving protonation of 8-oxoguanine at O-6 and nucleophilic attack of the deoxyribose moiety at C-1' leading to the formation of an enzyme-substrate Schiff base intermediate (Tchou, J., Bodepudi, V., Shibutani, S., Antoshechkin, I., Miller, J., Grollman, A. P., and Johnson, F. (1994) J. Biol. Chem. 269, 15318-15324). In this paper, sodium cyanoborohydride has been used to convert the transient intermediate to a covalent enzyme-DNA complex. The location of the active site of Fpg protein is further delineated using two approaches. 1) A radiolabeled DNA substrate is used to tag the active site of Fpg protein, using sodium cyanoborohydride. The active site is mapped to the first 73 amino acid residue fragment by cyanogen bromide cleavage analysis. 2) A maltose-binding protein fusion system is used to generate amino-terminal modifications of Fpg protein to explore the role of the amino-terminal region in DNA binding and catalysis. Results support the conclusion that the active site of Fpg protein is located at or near the amino terminus. Thus, Fpg protein may act in a similar fashion as T4 endonuclease V, a DNA repair enzyme that uses its amino-terminal alpha-amino group of threonine to carry out catalysis via Schiff base formation (Dodson et al., 1993).

Substrate specificity of Fpg protein. Recognition and cleavage of oxidatively damaged DNA

The 8-oxoguanine-DNA glycosylase of Escherichia coli, also known as formamidopyrimidine-DNA glycosylase (Fpg protein), has N-glycosylase and AP-lyase activities. This enzyme repairs oxidative DNA damage by efficiently removing formamidopyrimidine lesions and 8-oxoguanine residues from DNA. Defined oligodeoxynucleotides containing various 8-oxopurines were used to examine the substrate specificity of Fpg protein and to establish the role of functional groups in DNA on damage recognition and catalysis. Binding affinities of Fpg protein were established for duplex oligodeoxynucleotides containing 8-oxo-2'-deoxyguanine, 8-oxo-2'-deoxyadenine, 8-oxo-2'-deoxynebularine, 8-oxo-2'-deoxyinosine, abasic sites, and a ring-open adduct of C8-aminofluorene guanine. The C8 keto group of 8-oxodG:dC presents in the major groove and is correlated with tight binding (Kd = 8.9 nM). Binding is much weaker when the C8 keto functional group is in the minor groove, as in 8-oxodG:dA (Kd = 340 nM). Km and Vmax were determined for the cleavage reaction. Specificity constants (Kcat/Km) are consistently higher for oligodeoxynucleotide duplexes containing 8-oxopurines with C6 and C8 keto groups, as in 8-oxodG:dC and 8-oxodI:dC, where Kcat/Km are 9.3 and 18 min-1 nM x 10(-3), respectively. 8-oxodN:dC lacks the C6 keto group; the specificity constant is 0.024 min-1 nM x 10(-3). Taken together, our data suggest that the C8 keto group of 8-oxodeoxyguanine and the carbonyl moiety of formamidopyrimidine enable Fpg protein to recognize and bind duplex DNA containing these modified bases. An enzyme-catalyzed reaction involving the C6 keto group of the substrate leads to removal of these lesions. A mechanism involving protonation at O-6 of 8-oxoguanine is proposed to account for the N-glycosylase activity of this enzyme.

Function of the zinc finger in Escherichia coli Fpg protein

Fpg protein of Escherichia coli cleaves duplex DNA containing the oxidatively damaged base 8-oxo-7,8-dihydroguanine (Tchou, J., Kasai, H., Shibutani, S., Chung, M.-H., Laval, J., Grollman, A. P., and Nishimura, S. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 4690-4694). This DNA repair enzyme contains one zinc atom/protein molecule (Boiteux, S., O'Connor, T. R., Lederer, F., Gougette, A., and Laval, J. (1990) J. Biol. Chem. 265, 3916-3922); its N-glycosylase and apurinic/apyrimidinic lyase activities are physically associated. Amino acid sequence analysis reveals a putative single zinc finger motif of the CC/CC type located near the carboxyl terminus. A gel mobility shift assay was used to assay binding of Fpg protein to a noncleavable substrate analog, namely an oligodeoxynucleotide duplex containing a single tetrahydrofuran residue. High resolution hydroxyl radical DNA footprinting showed protection centered around the tetrahydrofuran residue. No footprint was observed on the complementary strand. To establish the role of COOH-terminal zinc finger in DNA binding and/or DNA cleavage, amino acid substitutions and an amber mutation were introduced at Cys-244 (C244S, C244H, C244A, and C244amber). In addition, a double amino acid substitution was generated at Cys-244 and Cys-247 (C244S/C247S). These mutant Fpg proteins lack DNA binding or cleavage activity, as tested in crude lysates of Escherichia coli. Wild type Fpg protein contains one zinc/protein molecule, whereas the mutant Fpg protein (C244S/C247S) lacks zinc, as measured by atomic absorption spectroscopy. This mutation did not significantly alter secondary structure, as assessed by circular dichroism spectroscopy. Our results suggest that Fpg protein utilizes its single COOH-terminal zinc finger motif in DNA binding.

A repair system for 8-oxo-7,8-dihydrodeoxyguanine

Active oxygen species can damage DNA and may play a role in aging and carcinogenesis. We have tested MutY glycosylase for activity on undamaged mispairs as well as mispairs formed with the oxidatively damaged substrates, 8-oxo-7,8-dihydrodeoxyguanine (GO) or 8-oxo-7,8-dihydrodeoxyadenine (AO). MutY acts as a glycosylase on four of the heteroduplexes tested, A/G, A/GO, A/C, and A/AO, removing the undamaged adenine from each substrate. Genetic data suggest that the primary substrate for MutY glycosylase in vivo is the A/GO mispair. We present biochemical evidence demonstrating that MutY glycosylase is an important part of a repair system that includes the MutM and MutT proteins. The GO repair system is dedicated to the repair of the oxidatively damaged guanine and the mutations it can induce.

8-oxoguanine (8-hydroxyguanine) DNA glycosylase and its substrate specificity

Substrate specificities of FPG protein (also known as formamidopyrimidine DNA glycosylase) and 8-hydroxyguanine endonuclease were compared by using defined duplex oligodeoxynucleotides containing single residues of 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-oxodG), 8-oxo-7,8-dihydro-2'-deoxyadenosine (8-oxodA), and 2,6-diamino-4-hydroxy-5-(N-methyl)formamidopyrimidine (Me-Fapy). Duplexes containing 8-oxodG positioned opposite dC, dG, or dT were cleaved, whereas single-stranded DNA and duplexes containing 8-oxodG.dA or 8-oxodA positioned opposite any of the four DNA bases were relatively resistant. Both enzymes cut duplexes containing 8-oxoG.dC 3' and 5' to the modified base but failed to cleave duplex DNA containing synthetic abasic sites, mismatches containing dG, or unmodified DNA. 8-Oxoguanine, identified by HPLC-electrochemical detection techniques, was released during the enzymatic reaction. Apparent Km values for FPG protein acting on duplex substrates containing a single Me-Fapy or 8-oxodG residue positioned opposite dC were 41 and 8 nM, respectively, and those for 8-hydroxyguanine endonuclease were 30 and 13 nM, respectively. Comparison of the properties of the two enzyme activities suggest that they are identical. In view of the widespread distribution of 8-oxodG in cellular DNA, the demonstrated miscoding and mutagenic properties of this lesion, and the existence of a bacterial gene coding for FPG protein, we propose that 8-oxodG DNA is the primary physiological substrate for a constituent glycosylase found in bacteria and mammalian cells.