Abstract P2-08-42: Use of breast cancer index to analyze tumor proliferation and endocrine responsiveness in genomic intermediate risk patients

Background: Breast Cancer Index (BCI) is a genomic assay that stratifies patients (pts) for cumulative 10-year and late (post–5-year) risk of distant recurrence and predicts the likelihood of extended endocrine therapy (EET) benefit based on the algorithmic analysis of gene expression from two functional gene cassettes: 1) The Molecular Grade Index (MGI), which contains 5 proliferation genes and 2) The HoxB13/IL17BR (H/I) ratio, an endocrine response biomarker. BCI more precisely resolves genomic intermediate risk patients into low- and high-risk groups (Sestak, Clin Cancer Res 2016) and adds significant prognostic data beyond clinical features (ie, clinical treatment score; Sestak, JAMA Oncol 2017) by interrogating different aspects of tumor biology such as proliferation and endocrine response. The objective of this study was to measure tumor proliferation and endocrine responsiveness using MGI and H/I respectively and evaluate their correlation with age in genomic intermediate risk pts.

Methods: This study utilized a subset of cases from the BCI Clinical Database for Correlative Studies, an IRB-approved de-identified database that contains clinicopathologic and molecular variables from clinical cases submitted for BCI testing. Genomic intermediate risk cases were defined as LN-, HER2- (or HER2 status unknown) with 21-gene recurrence scores (RS) of 11 to 25 using cut-points from the TAILORx study. Quantitative scores for MGI and H/I were derived by algorithmic analysis of BCI gene expression. Age groups (<50y and ≥51y) were determined using the date of diagnosis. MGI and H/I were evaluated in two different genomic intermediate risk groups: RS 11-15 and RS 16-25. Pearson correlation coefficients were used to determine the correlation between MGI, BCI, H/I, and genomic intermediate risk scores for patients <50y and ≥51y.

Results: Of the 441 pts with RS and BCI results, 303 (69%) with genomic intermediate risk were analyzed. The median MGI score in pts ≥51y was higher in the RS 16-25 group compared to RS 11-15, but there was no difference in MGI score between the two genomic intermediate RS risk groups for pts <50y. In contrast to MGI, median H/I was higher in the RS 16-25 group irrespective of age, with 37% of pts <50y and 41% of pts ≥51y having tumors predicted as more likely to benefit from EET using validated cut-points for H/I. There was no significant correlation between tumor proliferation (MGI: r=0.166) or endocrine responsiveness (H/I: r=0.244) with genomic intermediate-risk RS 11-25 group.

Conclusion: These data, which show variations in tumor proliferation and endocrine signaling based on age and genomic intermediate risk group, highlight the importance of measuring different features of tumor biology for risk classification and prediction of therapy response. The absence of correlations between tumor proliferation and genomic intermediate risk and age, and between estrogen signaling and genomic intermediate risk and age, suggests that assays such as BCI that combine distinct aspects of tumor biology for prognosticating risk of recurrence and prediction of benefit from endocrine therapy provide additional value for individualizing the management of patients with early-stage ER+ breast cancer.

Citation Format: Yan F, Master AK, Israel MA, Liu J, Zhang Y, Schnabel CA, Gadi VK. Use of breast cancer index to analyze tumor proliferation and endocrine responsiveness in genomic intermediate risk patients [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 P2-08-42.

Secretion-mediated STAT3 activation promotes self-renewal of glioma stem-like cells during hypoxia

High-grade gliomas (HGGs) include the most common and the most aggressive primary brain tumor of adults and children. Despite multimodality treatment, most high-grade gliomas eventually recur and are ultimately incurable. Several studies suggest that the initiation, progression, and recurrence of gliomas are driven, at least partly, by cancer stem-like cells. A defining characteristic of these cancer stem-like cells is their capacity to self-renew. We have identified a hypoxia-induced pathway that utilizes the Hypoxia Inducible Factor 1α (HIF-1α) transcription factor and the JAK1/2-STAT3 (Janus Kinase 1/2 - Signal Transducer and Activator of Transcription 3) axis to enhance the self-renewal of glioma stem-like cells. Hypoxia is a commonly found pathologic feature of HGGs. Under hypoxic conditions, HIF-1α levels are greatly increased in glioma stem-like cells. Increased HIF-1α activates the JAK1/2-STAT3 axis and enhances tumor stem-like cell self-renewal. Our data further demonstrate the importance of Vascular Endothelial Growth Factor (VEGF) secretion for this pathway of hypoxia-mediated self-renewal. Brefeldin A and EHT-1864, agents that significantly inhibit VEGF secretion, decreased stem cell self-renewal, inhibited tumor growth, and increased the survival of mice allografted with S100β-v-erbB/p53-/- glioma stem-like cells. These agents also inhibit the expression of a hypoxia gene expression signature that is associated with decreased survival of HGG patients. These findings suggest that targeting the secretion of extracellular, autocrine/paracrine mediators of glioma stem-like cell self-renewal could potentially contribute to the treatment of HGGs.

Abstract P1-06-10: Correlative analysis of breast cancer index (BCI) restratification of 21-gene recurrence score (RS) in patients with hormone receptor-positive (HR+), node-negative breast cancer

Background: In a cross-stratification analysis performed within the TransATAC cohort, Breast Cancer Index (BCI) and 21-gene Recurrence Score (RS) had a concordance of 58.2% (Sestak et al., Clin Cancer Res, 2016). Restratification by BCI of the low and intermediate RS risk groups led to subgroups with significantly different rates (P < 0.001 and P = 0.003, respectively); in contrast, restratified subgroups created by RS of BCI risk groups did not differ significantly. The objective of this study was to analyze the concordance of BCI and RS test results in HR+, node-negative (LN-) patients tested in the real-world setting and to investigate molecular, clinical, and pathologic correlates within discrepant cases.

Methods: This study utilized a subset of cases from the BCI Clinical Database for Correlative Studies, an IRB-approved de-identified database which contains clinicopathologic and molecular variables from clinical cases submitted for BCI testing. Clinicopathologic variables, abstracted from pathology reports, were available for a subset of cases. This analysis evaluated cases from LN- patients with available RS data. Concordance was evaluated between BCI Prognostic risk groups (Low, High) and RS risk groups (Low, Intermediate, High based on TAILOR Rx cutpoints [0-10, 11-25, and 26+]). Fisher's Exact tests were used to compare molecular (HoxB13/IL17BR [H/I] endocrine response biomarker and Molecular Grade Index [MGI] proliferation marker) and clinicopathologic (age, grade, size, HER2, Ki67) data in discrepant risk groups.

Results: There were 456 LN- patients included. Median age was 58.0y (range 27.2-84.0y; 73.9% ≥50y); 33.1%/50.1%/16.8% were grade 1/2/3; and 24.0%/59.5%/15.3% were T1ab/T1c/T2. BCI classified 47.8% (n=218) of patients as Low Risk vs 52.2% (n=238) as High Risk. RS classified 17.1% (n=78), 67.1% (n=306), and 15.8% (n=72) of patients as Low, Intermediate, and High Risk, respectively. BCI restratified RS-Low patients as high risk in 17.1% of cases, restratified RS-Intermediate as Low Risk in 48.4% and High Risk in 51.6%, and restratified RS-High as Low risk in 20.8% of cases. In RS-Low patients, only H/I (P=0.0004) and MGI (P=0.047) were significantly correlated with restratification to BCI-High Risk. In RS-Intermediate patients, H/I (P<0.0001), MGI (P<0.0001), grade (P<0.0001), and Ki67 >20% (P=0.0003) were significantly correlated with restratification by BCI to High or Low Risk. In RS-High patients, H/I (P=0.0008), MGI (P<0.0001), grade (P=0.016) were significantly correlated with restratification to Low Risk.

Conclusion: BCI restratified a substantial proportion of patients in each RS risk group. Based on previous studies demonstrating that BCI has improved prognostic ability for assessment of risk of late distant recurrence (Sgroi et al., Lancet Oncol, 2013), these results highlight the clinical utility of BCI within all RS risk groups. The estrogen signaling pathway biomarker H/I and proliferative biomarkers (MGI, grade, Ki67) were associated with restratification by BCI, while age, HER2 status, and tumor size were not.

Citation Format: Yan F, Master AK, Israel MA, Liu J, Schnabel CA, Hurvitz S, Gadi VK. Correlative analysis of breast cancer index (BCI) restratification of 21-gene recurrence score (RS) in patients with hormone receptor-positive (HR+), node-negative breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-06-10.

Abstract P4-09-09: Correlation of breast cancer index (BCI) predictive (HoxB13/IL17BR) results to nodal status and hormone receptor expression in early stage HR+ breast cancer

Background: The HoxB13/IL17BR (H/I) ratio, the endocrine response component of the Breast Cancer Index (BCI), was initially discovered in patients with LN- breast cancer, and validated as a predictive biomarker of benefit from extended endocrine therapy (EET) in a cohort from MA.17 that included both LN- and LN+ patients. Here, correlative analyses were performed to further characterize BCI Predictive (H/I) results with nodal status and quantitative hormone receptor expression.

Methods: Analyses were performed using data from the BCI Clinical Database for Correlative Studies, an IRB-approved de-identified database that contains >50 clinicopathologic and molecular variables from cases submitted for BCI in clinical practice (N=19,126). Clinicopathologic variables were abstracted from pathology reports, and were available for a subset of these cases. Cases from patients with confirmed nodal status were analyzed. Chi-squared tests and ANOVA were used to compare results between subgroups.

Results: Analyses included 13,114 patients (median age 58.9y; range 23-92y); 9562 were LN- and 3552 LN+. The distribution of individual H/I scores and proportion of patients classified as High H/I were generally similar in LN- and LN+ patients, though a modestly greater proportion of LN+ patients (46.2%) were classified as High H/I compared to LN- patients (42.0%; P<0.01). In both LN- and LN+ patients, median quantitative ER (qER) expression was slightly higher in patients with Low H/I than with High H/I; qPR showed the same trend but with a larger magnitude (P<0.0001 for both). Similar results were observed for percent positive staining by IHC (P<0.0001). In both LN- and LN+ patients, H/I showed a weak negative correlation with qER (LN-, 0.227; LN+, 0.192) and qPR expression (LN-, 0.311; LN+, 0.311).

Conclusion: In this study to evaluate potential biological correlates of BCI, results showed that H/I biomarker activity did not appear to be dependent on nodal status. Secondly, although ER expression is an established biomarker for endocrine sensitivity, High H/I status did not correlate with increased quantitative ER and PR expression. The H/I ratio may be an independent new marker for endocrine sensitivity independent of the strength of ER expression.

Citation Format: Elias A, Israel MA, Zhang Y, Schnabel CA, Mayordomo J. Correlation of breast cancer index (BCI) predictive (HoxB13/IL17BR) results to nodal status and hormone receptor expression in early stage HR+ breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P4-09-09.

Abstract P1-07-18: Clinicopathologic and molecular correlates of breast cancer index (BCI) results in patients with HR+, LN- breast cancer that are high risk of late distant recurrence (DR) / low likelihood of benefit from extended endocrine therapy (EET)

Background: BCI is a gene expression assay for patients with early stage HR+ breast cancer that provides 2 results: BCI Predictive, based on the HoxB13/IL17BR (H/I) ratio, reports a prediction of high vs low likelihood of benefit from EET; BCI Prognostic, based on the algorithmic combination of H/I and a set of proliferation-based genes, reports the risk of late distant recurrence (DR). Clinical actionability is distinct based on the 4 possible combinations of prognostic and predictive results. To better characterize patients classified by BCI as having a high risk of late DR but a low likelihood of benefit from EET, we assessed clinicopathologic and molecular correlates in this subset.

Methods: The BCI Clinical Database for Correlative Studies is a de-identified database containing >50 clinicopathologic and molecular variables from cases submitted for BCI in clinical practice (N=19,126). Clinicopathologic variables abstracted from pathology reports were available for a subset of these cases. Molecular proliferation status (molecular grade index [MGI]) and clinicopathologic parameters were examined in the 4 possible BCI result categories of BCI Prognostic (High vs Low risk) and BCI Predictive (High vs Low H/I). Chi-squared tests and ANOVA were used to compare BCI results within subsets.

Results: Analyses included 3843 LN- pts with clinicopathologic data: Median age was 59.1y (range 26-89y; 74% ≥50y); 30.9%, 51.7%, and 17.4% were Grade 1, 2, and 3, respectively; 27.8%, 48.9%, 21.7%, and 1.6% were T1a/b, T1c, T2, and T3, respectively. BCI categorized 41.4% of pts as having Low risk/Low likelihood of benefit, 31.3% with High risk/High benefit, 18.0% with High risk/Low benefit, and 9.3% with Low risk/High benefit. Patients with High Risk/Low Benefit had increased median proliferation scores (MGI), and a greater proportion of pts with grade 2/3 tumors and high Ki67 scores compared to pts with Low Risk/Low Benefit (P<.0001 for all). In contrast, there were only modest differences in clinicopathologic parameters between patients with High Risk/ Low Benefit and those with High Risk/High Benefit.

Conclusion: In characterizing the molecular and clinical correlates in BCI cases with a high risk of late DR but low likelihood of benefit from EET, we found that higher proliferative status was associated with classification of high risk of DR. Future studies might investigate whether patients with this molecular pattern might benefit from combinatorial therapy (e.g., CDK 4/6 inhibitors) with EET. This study highlights the importance of predictive biomarkers for individualized EET therapy recommendation.

Citation Format: Royce M, Poage G, Israel MA, Schnabel CA, Holmes FA. Clinicopathologic and molecular correlates of breast cancer index (BCI) results in patients with HR+, LN- breast cancer that are high risk of late distant recurrence (DR) / low likelihood of benefit from extended endocrine therapy (EET) [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-18.

Abstract P1-07-12: Correlation of breast cancer index (BCI) prognostic and predictive results to clinicopathologic risk groups in early stage HR+ breast cancer

Background: Both clinicopathologic factors and genomic tests have been shown to be prognostic for risk of late distant recurrence (DR); however, few studies have characterized differential patient stratification.Breast Cancer Index (BCI) is a validated gene expression assay for patients with early-stage HR+ breast cancer that provides a prognostic result for high vs low risk of late distant recurrence and a separate predictive result (based on the HoxB13/IL17BR [H/I] ratio) for high vs low likelihood of benefit from extended endocrine therapy. Thus four categories of results are possible based on a patient's tumor biology. To better understand how patient stratification is affected by a combination of clinicopathologic and genomic factors, this study examined BCI assay results within clinicopathologic risk categories based on tumor size and grade.

Methods: This study utilized data from the BCI Clinical Database for Correlative Studies, an IRB-approved de-identified database which contains clinicopathologic and molecular variables from 19,126 clinical cases submitted for BCI testing. Clinicopathologic variables, abstracted from pathology reports, were available for a subset of these cases. This analysis evaluated cases from LN- patients with available clinicopathologic data. Chi-squared tests were used to compare BCI results between tumor size and grade subgroups.

Results: Analyses included 3843 LN- patients (median age 59.1y; range 26-89y; 74% ≥50y), of which 31%, 52%, 17% were Grade 1, 2, and 3, respectively, and 5%, 22.7%, 48.9%, 21.7%, and 1.6% were T1mi/a, T1b, T1c, T2, and T3, respectively. In analysis based on tumor size, there was a wide distribution of individual BCI Prognostic scores in all tumor size subsets; however, the proportion of patients classified as high risk increased with larger tumor size (T1a/b 39.0%, T1c 50.1%, T2 61.0%; p<.0001). In contrast, BCI Predictive (H/I) was not as strongly correlated with size, with a modestly larger proportion of patients classified as High H/I with larger tumor size (T1a/b 37.2%, T1c 40.5%, and T2 45.3%; p=.005). Within each tumor size category, the proportion of patients classified as BCI High Risk and High H/I increased with tumor grade (p<.0001). However, there was a wide distribution of individual risk assessments by BCI Prognostic and stratification by BCI Predictive (H/I) in all size + grade subsets. In patients with the most favorable clinicopathologic risk profile (T1a/b, G1), BCI classified 20% as high risk, 68% of whom also had High H/I.

Conclusion: While BCI results correlated with tumor size and grade, BCI identified substantial proportions of patients with favorable clinicopathologic features as high risk for late DR and apparent high likelihood of benefit from EET; conversely, BCI also identified patients with high risk clinicopathologic features as low risk for late distant recurrence and apparent low likelihood of benefit from EET. These findings help to differentiate between genomic-based and clinicopathologic-based risk/benefit assessment for patients considering EET.

Citation Format: Mayordomo J, Falkson C, Kepes J, Israel MA, Schroeder BE, Schnabel CA, Elias A. Correlation of breast cancer index (BCI) prognostic and predictive results to clinicopathologic risk groups in early stage HR+ breast cancer [abstract]. In: Proceedings of the 2017 San Antonio Breast Cancer Symposium; 2017 Dec 5-9; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2018;78(4 Suppl):Abstract nr P1-07-12.

Inhibitor of DNA binding-4 promotes angiogenesis and growth of glioblastoma multiforme by elevating matrix GLA levels

Inhibitor of differentiation-4 is highly expressed in glioblastoma multiforme (GBM). We report a novel pro-angiogenic function for inhibitor of differentiation-4 in the growth of glioblastoma xenografts. Tumor-derived cell cultures expressing elevated levels of ID4 produced enlarged xenografts in immunosuppressed mice that were better vascularized than corresponding control tumors and expressed elevated matrix GLA protein (MGP) that mediated enhanced tumor angiogenesis. Inhibition of MGP resulted in smaller and less vascularized xenografts. Our finding shows a novel function for ID4 in tumor angiogenesis, and identifies ID4 and MGP as possible therapeutic targets for GBM.

EGFR overexpression and radiation response in glioblastoma multiforme

PURPOSE

Recent studies have suggested relative radioresistance in glioblastoma multiforme (GM) tumors in older patients, consistent with their shorter survival. Two common molecular genetic abnormalities in GM are age related: epidermal growth factor receptor (EGFR) overexpression in older patients and p53 mutations in younger patients. We tested whether these abnormalities correlated with clinical heterogeneity in GM response to radiation treatment.

METHODS AND MATERIALS

Radiographically assessed radiation response (5-level scale) was correlated with EGFR immunoreactivity, p53 immunoreactivity, and p53 exon 5-8 mutation status in 170 GM patients treated using 2 prospective clinical protocols. Spearman rank correlation and proportional-odds ordinal regression were used for univariate and multivariate analysis.

RESULTS

Positive EGFR immunoreactivity predicted poor radiographically assessed radiation response (p = 0.046). Thirty-three percent of tumors with no EGFR immunoreactivity had good radiation responses (>50% reduction in tumor size by CT or MRI), compared to 18% of tumors with intermediate EGFR staining and 9% of tumors with strong staining. There was no significant relationship between p53 immunoreactivity or mutation status and radiation response. Significant relationships were noted between EGFR score and older age and between p53 score or mutation status and younger age.

CONCLUSION

The observed relative radioresistance of some GMs is associated with overexpression of EGFR.

(E)-5-(2-bromovinyl)-2'-deoxyuridine potentiates ganciclovir-mediated cytotoxicity on herpes simplex virus-thymidine kinase--expressing cells

Tumor cells expressing the thymidine kinase gene of the herpes simplex virus (HSV-tk) are rendered highly susceptible to the cytotoxic effects of different antiherpes drugs. In an attempt to enhance cytotoxicity of this therapeutic approach in glioma and other tumor cell lines transduced with the HSV-tk gene, we evaluated tumor cell killing following co-administration of two different prodrugs metabolized by HSV-tk, (E)-5-(2-bromovinyl)-2'-deoxyuridine (BVDU), and ganciclovir (GCV). In 8 of 12 cell lines investigated, addition of BVDU in concentrations showing no cytotoxic effect or only limited cytotoxicity could enhance GCV-mediated cell killing by as much as one order of magnitude. In co-cultures consisting of HSV-tk(+) (9L STK) and HSV-tk(-) (9L wild-type) cells, we also observed potentiation of GCV-mediated cytotoxicity in the presence of BVDU, suggesting strongly enhanced bystander cell killing. BVDU is thought to exert its cytotoxic effect through inhibition of thymidylate synthase activity or by incorporation into replicating DNA. Both effects could be observed in all HSV-tk--expressing cells investigated, including cell lines which did not exhibit cytotoxicity after incubation with BVDU. These findings argue against current concepts of BVDU-mediated cytotoxicity in HSV-tk--expressing cells. Taken together, our data suggest that gene therapy utilizing prodrug activating enzymes may be rendered more effective by simultaneous treatment with two different prodrugs metabolized by the same enzyme.

The gene for the axonal cell adhesion molecule TAX-1 is amplified and aberrantly expressed in malignant gliomas

The human TAX-1 gene encodes a Mr 135,000 glycoprotein that is transiently expressed on the surface of a subset of neurons during development and is involved in neurite outgrowth. The TAX-1 gene has been mapped to a region on chromosome 1 that has been implicated in microcephaly and the Van der Woude syndrome. Using restriction landmark genome scanning to search for amplified genes in gliomas, we found TAX-1 to be amplified in 2 high-grade gliomas among a group of 26 gliomas investigated. Real-time reverse transcription-quantitative PCR analysis detected high levels of TAX-1 mRNA in glial tumors, even in the absence of TAX-1 gene amplification. Immunohistochemical analysis revealed abundant levels of TAX-1 in neoplastic glial cells of glioblastoma multiforme tumors. Because glial tumors are highly invasive and in view of the role of TAX-1 in neurite outgrowth, we investigated the potential role of TAX-1 in glioma cell migration. Using an in vitro assay, we found that the migration of glioma tumor cells is profoundly reduced in the presence of either an anti-TAX-1 antibody or a TAX-1 antisense oligonucleotide. Our findings suggest that TAX-1 plays a role in glial tumorigenesis and may provide a potential target for therapeutic intervention.

Protease pretreatment increases the efficacy of adenovirus-mediated gene therapy for the treatment of an experimental glioblastoma model

Effective virus-mediated gene therapy for cancer will be facilitated by procedures that enhance the low level of gene transfer mediated by replication-deficient, recombinant viral vectors. We found recently that protease pretreatment of solid tumors is a useful strategy for enhancing virus-mediated gene transduction in vivo. In this study, we examined the potential of protease pretreatment to improve the efficacy of a gene therapy strategy for prodrug activation that depends on infection with a recombinant adenovirus encoding herpes simplex virus thymidine kinase (Ad-HSV-tk). Trypsin or a dissolved mixture of collagenase/dispase was inoculated into xenografts derived from the human glioblastoma multiforme-derived cell lines, U87 or U251. Ad-HSV-tk was administered 24 h after protease pretreatment, and animals were then treated for 10 days with ganciclovir (GCV). We found that protease pretreatment increased the efficacy of adenovirus mediated HSV-tk/GCV gene therapy in these experimental tumor models. Mice receiving Ad-HSV-tk/GCV after protease pretreatment demonstrated a significantly greater regression of tumors compared with those treated with Ad-HSV-tk/GCV alone. No adverse effects of protease pretreatment were observed. No signs of metastasis were seen either by histological inspection of lymph nodes or by a PCR-based analysis of selected mouse tissues to detect human tumor cells. Our findings indicate that protease pretreatment may be a useful strategy to enhance the efficacy of virus-mediated cancer gene therapy.

Analysis of complex relationships between age, p53, epidermal growth factor receptor, and survival in glioblastoma patients

Glioblastoma multiforme (GBM) carries a dismal prognosis. However, a range of survival times exists, and parameters that define prognostic groups may help to optimize treatment. To identify such prognostic groups, we analyzed tumor tissue from 110 cases of newly diagnosed GBM from two clinical protocols. Similar to other studies, we found no association of epidermal growth factor receptor (EGFR) overexpression (as assessed by immunohistochemistry), p53 immunopositivity, or p53 mutation with survival in the entire sample. However, EGFR overexpression showed trends toward worse prognosis in patients younger than the median age, but better prognosis in patients older than the median age. This interaction of EGFR with age group was statistically significant and led us to focus our further analyses on the younger patients. In this group, a statistically significant association of EGFR overexpression with worse survival was identified in the p53-negative but not p53-positive tumors. We found a similar result after screening these cases for mutations in p53: EGFR overexpression was negatively associated with survival only in the p53 wild-type cases. To confirm this unexpected result, this finding was reproduced in a validation sample of an additional 42 tumors from younger patients on the same two clinical protocols. This complex relationship between EGFR and p53 in younger patients remained in a multivariate analysis that incorporated additional prognostic variables. The results suggest that analysis of prognostic markers in GBM is complex, and maximal information may require analysis of subgroups based on age and the status of specific markers such as p53. In addition, they suggest a specific group of patients on which to focus promising therapies targeting EGFR.

Pretreatment with protease is a useful experimental strategy for enhancing adenovirus-mediated cancer gene therapy

A key impediment to the development of effective virus-mediated gene therapy for cancer is the low level of gene transfer that occurs after the administration of recombinant viral vectors. Improving in vivo infection and transduction efficiency is an important goal for gene therapy. The limited distribution of gene delivery is particularly problematic when large vectors such as recombinant adenoviruses and retroviruses are used to mediate transgene delivery to solid tumors. To facilitate the spread of virus, we have investigated the potential of administering proteases prior to the intratumoral inoculation of recombinant replication deficient adenovirus. For these studies, we chose proteases that are active against collagen and the other extracellular matrix proteins found in primary brain tumor tissue, but are not widely expressed in normal brain. Various concentrations of a mixture of collagenase/dispase or trypsin were inoculated into xenografts of human glioblastoma multiforme-derived brain tumor cell lines U87, U251, and SF767. Subsequently, recombinant adenovirus encoding the beta-galactosidase gene was administered and tumor tissue was examined for evidence of virus infection. Both collagenase/dispase and trypsin enhanced virus infection, indicating that protease pretreatment may be a useful strategy for enhancing virus-mediated gene transduction for many in vivo applications.

The basic helix-loop-helix transcription factors myogenin and Id2 mediate specific induction of caveolin-3 gene expression during embryonic development

Caveolin-3 protein is the only member of the caveolin family that shows a unique muscle-specific expression pattern, and loss of its functional activity causes muscular dystrophy. Caveolin-3 mRNA levels are dramatically increased during the formation of myotubes in the C2C12 cell line. In this study, we characterized the human caveolin-3 5'-flanking region. Promoter analyses demonstrate that the proximal E box element serves as a myogenin binding site and is both necessary and sufficient to control caveolin-3 gene transcription. Transient transfection assays indicated that overexpression of myogenin activates caveolin-3 reporter gene expression, whereas Id2 overexpression inhibited caveolin-3 promoter activation by myogenin. A mutant Id2 protein lacking the HLH domain was not capable of suppressing myogenin-mediated activation. Determination of caveolin-3 transcript distribution patterns in vivo revealed that mRNA was first detectable at day 10 of gestation in the developing somites and heart. Caveolin-3 protein in myoblasts and myotubes was expressed in both the plasma membrane and vesicular structures. During skeletal myogenesis the level of Id2, an inhibitor of differentiation, decreases, allowing the induced basic helix-loop-helix transcription factor myogenin to form transcriptionally active heterodimers that bind to the caveolin-3 promoter and thereby mediate its transcription.

Id genes in nervous system development

Id genes encode helix-loop-helix proteins that function to mediate processes important for normal development including cellular differentiation, proliferation and apoptosis. Id proteins act as negative regulators of other transcription factors, which are essential for cell determination and differentiation in diverse cell types, and interact with proteins important for cell cycle regulation. Studies of Id gene expression in the nervous system and in neural cells in culture indicate that Id proteins contribute to the regulation of mammalian nervous system development. Also, recognition of a wide variety of proteins with which Id transcription factors are capable of interacting suggests that it will be possible to understand more precisely their specific functions and importantly how these are integrated.

Id-2 regulates critical aspects of human cytotrophoblast differentiation, invasion and migration

During early human placental development, the conceptus attaches itself to the uterus through cytotrophoblast invasion. Invasive cytotrophoblast cells differentiate from precursor villous cytotrophoblasts, but the essential regulating factors in this process are unknown. Basic helix-loop-helix (bHLH) transcription factor dimers are essential regulators of mouse trophoblast development. We therefore examined the importance of this family of factors in the human placenta. In many cell lineages, bHLH factors are sequestered by members of the Id family, HLH proteins that lack the basic DNA binding domain (Inhibitor of DNA binding proteins (Id-1 to Id-4)). During differentiation of some tissues, Id expression declines, allowing bHLH factors to dimerize, bind DNA and trans-activate lineage-specific genes. To begin to study the role of bHLH transcription factors in human placental development, we first characterized Id expression in cytotrophoblast cells. The cells expressed Id-3 constitutively; Id-2 was downregulated, at the mRNA and protein levels, as the cells differentiated in culture and in situ, respectively. In cases when cytotrophoblast differentiation was compromised (in placentas from women with preeclampsia, or in cells grown under hypoxic conditions in culture), Id-2 expression was maintained. To assess the functional relevance of these correlations, we used an adenovirus vector to maintain Id-2 protein expression in cultured cytotrophoblasts. Compared to control (lacZ-expressing) cells, cytotrophoblasts transduced to constitutively express Id-2 retained characteristics of undifferentiated cells: (alpha)1 integrin expression was low and cyclin B expression was retained. Furthermore, invasion through Matrigel was partially inhibited and migration was strikingly enhanced in Id-2-expressing cells. These results suggest that Id-2 and the bHLH factors that it partners play important roles in human cytotrophoblast development.

Overexpression of E2F1 in glioma-derived cell lines induces a p53-independent apoptosis that is further enhanced by ionizing radiation

Glioma cell lines show variable responses to radiation in a manner influenced by their p53 status. Irradiation of glioma cell lines does not generally induce apoptosis. When wild-type p53 is present, these cells undergo a G1 arrest that is closely associated with increased radiosensitivity as measured by clonogenic survival. Previously, others have shown that dysregulated overexpression of E2F1 induces apoptosis in cell lines with either functional or inactivated p53. We found that regardless of p53 status, apoptosis induced by overexpression of E2F1 in glioma cell lines was further enhanced by treatment with ionizing radiation. BAX induction did not follow E2F1 overexpression or irradiation in the glioma cell lines tested. Thus, the apoptotic response of glioma-derived cells to irradiation can be enhanced by E2F1 by a mechanism that does not involve the induction of BAX.

Id-1 and Id-2 are overexpressed in pancreatic cancer and in dysplastic lesions in chronic pancreatitis

Id proteins antagonize basic helix-loop-helix proteins, inhibit differentiation, and enhance cell proliferation. In this study we compared the expression of Id-1, Id-2, and Id-3 in the normal pancreas, in pancreatic cancer, and in chronic pancreatitis (CP). Northern blot analysis demonstrated that all three Id mRNA species were expressed at high levels in pancreatic cancer samples by comparison with normal or CP samples. Pancreatic cancer cell lines frequently coexpressed all three Ids, exhibiting a good correlation between Id mRNA and protein levels, as determined by immunoblotting with highly specific anti-Id antibodies. Immunohistochemistry using these antibodies demonstrated the presence of faint Id-1 and Id-2 immunostaining in pancreatic ductal cells in the normal pancreas, whereas Id-3 immunoreactivity ranged from weak to strong. In the cancer tissues, many of the cancer cells exhibited abundant Id-1, Id-2, and Id-3 immunoreactivity. Scoring on the basis of percentage of positive cells and intensity of immunostaining indicated that Id-1 and Id-2 were increased significantly in the cancer cells by comparison with the respective controls. Mild to moderate Id immunoreactivity was also seen in the ductal cells in the CP-like areas adjacent to these cells and in the ductal cells of small and interlobular ducts in CP. In contrast, in dysplastic and atypical papillary ducts in CP, Id-1 and Id-2 immunoreactivity was as significantly elevated as in the cancer cells. These findings suggest that increased Id expression may be associated with enhanced proliferative potential of pancreatic cancer cells and of proliferating or dysplastic ductal cells in CP.

Immortalization of primary human keratinocytes by the helix-loop-helix protein, Id-1

Basic helix-loop-helix (bHLH) DNA-binding proteins have been demonstrated to regulate tissue-specific transcription within multiple cell lineages. The Id family of helix-loop-helix proteins does not possess a basic DNA-binding domain and functions as a negative regulator of bHLH proteins. Overexpression of Id proteins within a variety of cell types has been shown to inhibit their ability to differentiate under appropriate conditions. We demonstrate that ectopic expression of Id-1 leads to activation of telomerase activity and immortalization of primary human keratinocytes. These immortalized cells have a decreased capacity to differentiate as well as activate phosphorylation of the retinoblastoma protein. Additionally, these cells acquire an impaired p53-mediated DNA-damage response as a late event in immortalization. We conclude that bHLH proteins play a pivotal role in regulating normal keratinocyte growth and differentiation, which can be disrupted by the immortalizing functions of Id-1 through activation of telomerase activity and inactivation of the retinoblastoma protein.

A repertoire of differentially expressed transcription factors that offers insight into mechanisms of human cytotrophoblast differentiation

During human placental development, specialized cells allocated to the extraembryonic lineage (cytotrophoblasts) invade the uterus, anchoring the conceptus to the decidua and tapping a supply of maternal blood. This unusual behavior requires cytotrophoblasts to assume highly specialized characteristics; some are commonly associated with tumor cells, while others are typical of endothelia. Here we investigated the transcriptional mechanisms that control cytotrophoblast differentiation/invasion. Specifically, we examined the cells' expression of a number of transcription factors, at the RNA level, as they differentiated along the invasive pathway in vitro. Since basic helix-loop-helix (bHLH) proteins play important roles in murine trophoblast differentiation, we first examined their expression by cytotrophoblasts. As in murine placental development, expression of the human homologue of Mash-2 was confined to progenitor cells. But expression of Hand-1, which promotes differentiation of murine trophoblast giant cells, was not detected. We also found that cytotrophoblasts upregulated the expression of bHLH/PAS factors that function in adaptive responses to hypoxia, including hEPAS-1, which is expressed primarily in endothelial cells. Quite unexpectedly, we discovered that cytotrophoblasts express high levels of mRNA encoding the human homologue of the Drosophila neuronal fate gene, glial cells missing-1 (gcm-1). We also found evidence of crosstalk between the bHLH and GCM-1 regulatory networks. Together, these results offer insights into the transcriptional mechanisms that govern cytotrophoblast differentiation/invasion. Interestingly, these mechanisms suggest analogies with those that govern differentiation of murine stem cells allocated to both the intra- and extraembryonic lineages.

Id gene expression as a key mediator of tumor cell biology

Id genes encode members of the helix-loop-helix (HLH) family of transcription factors that inhibit transcription by forming inactive heterodimers with basic HLH (bHLH) proteins. There are four members of the Id gene family recognized in mammals, and the proteins they encode share homology primarily in their HLH domain. bHLH proteins typically form heterodimers with other bHLH proteins, and their basic domain binds to a DNA sequence element, the E-box, activating transcription. Products of Id genes lack the basic DNA binding domain of the bHLH transcription factors, and when they heterodimerize with bHLH proteins, the complexes are inactive. Generally, high levels of Id mRNA are detected in proliferative undifferentiated, embryonal cells and lower levels are detected in well-differentiated, mature, adult tissues. In vitro, these genes are generally expressed at lower levels in cells after the induction of differentiation. Recently, high levels of expression of Id genes have been identified in cell lines derived from a wide variety of different tumors and in tumor tissues as well. These findings suggest that not only the inappropriate proliferation of tumors but also the anaplastic characteristics that contribute to their malignant behavior may be regulated by Id gene expression.

Molecular genetics of brain tumors

As many as 40000 patients are newly diagnosed each year as having brain tumors. About half of these are metastatic foci of tumors originating outside the central nervous system, while the other half are primary tumors of central nervous system tissues. These are a diverse group of neoplasms. Currently, primary brain tumors are classified in a manner that reflects their histological appearance and location. The identification of cancer as a disorder of genes, however, has opened the possibility of classifying tumors according to the genetic alterations that underlie their pathogenesis and that regulate their malignant behavior. Two major classes of genes critical for the development of all types of cancer, including brain tumors, are now recognized: tumor suppressor genes, which encode genes that function to inhibit cell proliferation and tumor development, and oncogenes, which encode proteins that stimulate proliferation and mediate biological activities important for invasion, neoangiogenesis, immune escape, and other characteristics of malignancy. While in most cases the specific pathways regulating tumor characteristics such as tumor neoangiogenesis and tissue invasion remain to be defined, recognition of the genetic changes characteristic of individual tumor types should provide opportunities to develop more effective, less toxic therapies.

Id4 expression induces apoptosis in astrocytic cultures and is down-regulated by activation of the cAMP-dependent signal transduction pathway

The Id family of helix-loop-helix transcription factors has been implicated in the regulation of cellular differentiation in several different lineages. We have explored the potential regulatory role of the cyclic AMP-dependent signaling pathway on Id gene expression in astroglial primary cultures. We found that primary cultures of mouse forebrain astrocytes constitutively expressed the four known members of the Id gene family, Id1, Id2, Id3, and Id4. During culture in presence of serum for 4 weeks, the expression of Id4 was up-regulated. In these same cultures, treatment with dibutyryl-cyclic AMP, a cyclic AMP analogue known to promote astrocyte differentiation, dramatically and selectively decreased Id4 gene expression. This effect was detectable after short-term treatment and was maintained during long-term treatment. Forskolin and pentoxifylline, two other agents known to elevate intracellular cyclic AMP through different mechanisms, also potently decreased Id4 gene expression. Furthermore, overexpression of Id4 in an astrocyte-derived cell line induced cells to round up and die by apoptosis. These results indicate that the cyclic AMP pathway acts as an inhibitor of Id4 gene expression in astrocytes, identify a new function for Id4, and suggest that Id4 is strategically positioned in the chain of molecular events regulating astrocyte differentiation and apoptosis.

Fas (APO-1/CD95) signaling pathway is intact in radioresistant human glioma cells

Radiation-induced apoptosis can be mediated through pathways initiated by either DNA damage or ceramide-induced Fas signaling. Glioblastoma multiforme is a primary brain tumor that is highly resistant to irradiation, and U-87 MG, SF126, and T98G are glioblastoma-derived cell lines that mimic this characteristic. We found that these radioresistant glioma cells are susceptible to Fas-mediated cell death induced by treatment with either anti-Fas antibody or exogenous ceramide. Fas-mediated cell death in these cell lines is p53-independent. These data demonstrate that apoptosis can be induced by ceramide and mediated through the Fas pathway in glioma cells, although high-dose ionizing radiation fails to trigger this pathway.

p53 function influences the effect of fractionated radiotherapy on glioblastoma tumors

PURPOSE

Glioblastoma multiforme brain tumors (GM) are treated with a spectrum of fractionation regimens based on the clinical and anatomical characteristics of the tumor but rarely based on the molecular characteristics of the individual neoplasm. This study tests the hypothesis that the response of cell lines derived from GM to fractionated radiotherapy depends on the function of wild-type p53 (wt p53), a tumor suppressor gene frequently mutated in GM tumors.

METHODS & MATERIALS

Isogenic derivatives of glioblastoma cells differing only in p53 function were prepared using a retroviral vector expressing a dominant negative mutant of p53 (mt p53). Radiation survival in vitro was quantitated using linear quadratic and repair-saturation mathematical models. Apoptosis was assayed by a terminal deoxynucleotide transferase-labeling technique and chromatin morphology.

RESULTS

We have previously reported the generation of isogenic GM cell lines differing only in p53 function. U87-175.4, lacking wt p53 function, had a significantly lower alpha/beta value than U87-LUX.8, expressing functional wt p53, leading us to hypothesize that fractionated irradiation would preferentially spare GM cells harboring mt p53 compared with those expressing functional, wt p53. Survival curves following either 2.0 Gy or 3.5 Gy/fraction demonstrated that lack of functional wt p53 was associated with resistance to fractionated irradiation. Radiation-induced apoptosis could not account for the observed differences in clonogenic survival. Rather, our data suggested that a deficit in the G1-checkpoint contributed to increased resistance to fractionated irradiation of cells expressing mutant p53.

CONCLUSIONS

The effect of fractionated radiotherapy in GM may depend on the function of the tumor suppressor gene p53. A potential clinical consequence of these findings is that hyperfractionation regimens may provide a therapeutic advantage specifically for tumors expressing wt p53 whereas a radiotherapy course of fewer, larger fractions may be appropriate for the treatment of tumors carrying p53 mutations. Further studies are needed to confirm our proposal that the p53 status of GM tumors can be used to guide our choice of fractionation schemes.

Injury selectively down-regulates the gene encoding for the Id4 transcription factor in primary cultures of forebrain astrocytes

Astrogliosis is an important component of the response to injury of the central nervous system (CNS). The Id family of helix-loop-helix (HLH) transcription factors has been implicated in the regulation of cellular differentiation in several different lineages and may contribute to the regulation of astrogliosis. We examined the expression of Id genes in primary cultures of mouse forebrain astrocytes under experimental conditions in which astrogliosis was elicited by mechanical injury. Astrocyte cultures expressed the four known members of the Id gene family, Id1, Id2, Id3, and Id4. After injury, at a time when astrocytes developed the characteristic phenotypic changes of astrogliosis, Id4 expression decreased dramatically. Id1, Id2, and Id3 mRNA levels did not change. These results identify Id4 as a candidate marker of astroglial activation in culture and suggest that Id4 expression plays a role in the process of astrogliosis.

The intrinsic radioresistance of glioblastoma-derived cell lines is associated with a failure of p53 to induce p21(BAX) expression

Radiation is the primary modality of therapy for all commonly occurring malignant brain tumors, including medulloblastoma and glioblastoma. These two brain tumors, however, have a distinctly different response to radiation therapy. Medulloblastoma is very sensitive to radiation therapy, whereas glioblastoma is highly resistant, and the long-term survival of medulloblastoma patients exceeds 50%, while there are few long-term survivors among glioblastoma patients. p53-mediated apoptosis is thought to be an important mechanism mediating the cytotoxic response of tumors to radiotherapy. In this study, we compared the response to radiation of five cell lines that have wild-type p53: three derived from glioblastoma and two derived from medulloblastoma. We found that the medulloblastoma-derived cell lines underwent extensive radiation-induced apoptotic cell death, while those from glioblastomas did not exhibit significant radiation-induced apoptosis. p53-mediated induction of p21(BAX) is thought to be a key component of the pathway mediating apoptosis after the exposure of cells to cytotoxins, and the expression of mRNA encoding p21(BAX) was correlated with these cell lines undergoing radiation-induced apoptosis. The failure of p53 to induce p21(BAX) expression in glioblastoma-derived cell lines is likely to be of biologic significance, since inhibition of p21(BAX) induction in medulloblastoma resulted in a loss of radiation-induced apoptosis, while forced expression of p21(BAX) in glioblastoma was sufficient to induce apoptosis. The failure of p53 to induce p21(BAX) in glioblastoma-derived cell lines suggests a distinct mechanism of radioresistance and may represent a critical factor in determining therapeutic responsiveness to radiation in glioblastomas.

The mouse Id2 and Id4 genes: structural organization and chromosomal localization

The Id proteins belong to a family of nuclear HLH proteins lacking a basic region and thought to function as dominant-negative regulators of bHLH proteins during cell growth and differentiation. In this paper, we report the genomic organization of the mouse Id2 and Id4 genes. These genes each span approximately 3 kb of the mouse genome and are each organized as three exons with recognizable splice donor and acceptor consensus sequences. Their genomic organization is very similar, consistent with their having evolved from a common, ancestral Id-like gene. Using FISH analysis, we have localized the mouse Id2 and Id4 genes to mouse chromosome 12 and 13, respectively.

Tumor cell apoptosis present at diagnosis may predict treatment outcome for patients with medulloblastoma

PURPOSE

To determine if the degree of tumor cell apoptosis at diagnosis predicts outcome, tissue sections of medulloblastoma were examined and the amount of apoptosis and progression-free survival were correlated.

PATIENTS AND METHODS

The study cohort consisted of 43 children in whom medulloblastoma was diagnosed between 1984 and 1995: 29 patients at high risk (HR) treated with radiation and chemotherapy, and 14 children at low risk (LR) treated with radiation alone. A terminal deoxynucleotidyl transferase (TdT) end-labeling assay was used to detect apoptosis in paraffin-embedded tissue sections prepared at diagnosis.

RESULTS

Progression-free survival was examined in cohorts of children whose tumors were divided into quartiles based on the apoptotic index (AI) of their pretreatment tumor specimens. A comparison of these four groups of children revealed an association between AI and outcome (p = 0.03); patients with tumors in the highest AI quartile had substantially improved outcome compared to all other patients combined (p = 0.02). In this cohort of patients treated with different therapies, assignment at the time of diagnosis to LR and HR groups based on widely-accepted clinical criteria was not closely associated with outcome (p = 0.47).

CONCLUSION

AI is a strong indicator of treatment outcome for children with medulloblastoma after treatment with cytotoxic therapy, independent of risk group. Because HR and LR patients included in this study received different modalities of cytotoxic therapy, it is possible that AI predicts outcome independent of the precise antineoplastic therapy a patient receives.

Id2 promotes apoptosis by a novel mechanism independent of dimerization to basic helix-loop-helix factors

Members of the helix-loop-helix (HLH) family of Id proteins have demonstrated roles in the regulation of differentiation and cell proliferation. Id proteins inhibit differentiation by HLH-mediated heterodimerization with basic HLH transcription factors. This blocks their sequence-specific binding to DNA and activation of target genes that are often expressed in a tissue-specific manner. Id proteins can also act as positive regulators of cell proliferation. The different mechanisms proposed for Id-mediated promotion of entry into S phase also involve HLH-mediated interactions affecting regulators of the G1/S transition. We have found that Id2 augments apoptosis in both interleukin-3 (IL-3)-dependent 32D.3 myeloid progenitors and U2OS osteosarcoma cells. We could not detect a similar activity for Id3. In contrast to the effects of Id2 on differentiation and cell proliferation, Id2-mediated apoptosis is independent of HLH-mediated dimerization. The ability of Id2 to promote cell death resides in its N-terminal region and is associated with the enhanced expression of a known component of the programmed cell death pathway, the proapoptotic gene BAX.

The helix-loop-helix protein Id2 is overexpressed in human pancreatic cancer

Id2 belongs to the Id family of helix-loop-helix (HLH) proteins, which upon heterodimerization with basic HLH proteins prevent basic HLH proteins from DNA binding. Proteins of the Id family act as negative regulatory transcriptional factors, and their expression correlates with cell proliferation and arrested differentiation in many cell lineages. In this study, we characterized the expression of Id2 in normal and cancerous pancreatic tissues. Pancreatic cancers markedly overexpressed Id2 mRNA in comparison to the normal pancreas. Furthermore, there was abundant Id2 immunoreactivity in the cancer cells within the pancreatic tumor mass. In PANC-1 human pancreatic cancer cells, steady-state Id2 mRNA levels increased upon serum addition and decreased after induction of differentiation with either sodium butyrate or 12-O-tetradecanoylphorbol-13-acetate. Inhibition of Id2 expression with Id2 antisense oligonucleotides inhibited the growth of these cells, whereas random and sense oligonucleotides were without effect. These findings suggest that Id2 may have a role in human pancreatic cancer.

Ionizing radiation inhibits chemotherapy-induced apoptosis in cultured glioma cells: implications for combined modality therapy

Surgical resection followed by radiation therapy is the mainstay of treatment for glioblastoma multiforme (GBM), the most aggressive of the malignant gliomas. The poor clinical response of GBM and the intrinsic radiation resistance of this tumor type have prompted clinical investigations seeking to define the role of chemotherapy in the treatment of GBM. In this study, we examined the cytotoxic response of GBM-derived cell lines to treatment with both radiation and chemotherapy. We observed that the sensitivity of glioma cells to cisplatin- and FAS-induced apoptosis was diminished by prior treatment with ionizing radiation. Radiation conferred resistance to cisplatin and FAS cytotoxicity in a dose- and time-dependent manner. Radiation diminished the cisplatin-induced cytotoxicity of malignant glioma cells but failed to alter the cisplatin susceptibility of normal primary human astrocytes. Given the role of p53 in the response of cells to irradiation, we evaluated whether p53 function affects the observed radiation-induced resistance to cisplatin. By examining isogenic cell lines differing only in p53 function, we demonstrated that radiation conferred resistance to cisplatin independently of p53. Current clinical strategies in the treatment of astrocytic tumors, which include combined modality therapy, have been empirically derived from limited clinical experience. Further understanding of the molecular determinants of apoptosis associated with combined modality therapy may guide the design of more efficacious multimodality protocols.

Cloning of human ENC-1 and evaluation of its expression and regulation in nervous system tumors

We recently identified and characterized a novel murine gene, ENC-1, that is expressed primarily in the nervous system and encodes an actin-binding protein. To gain insight into a potential role for ENC-1 gene in the processes of cell differentiation and malignant transformation in the human nervous system, we first cloned and characterized the human homologue of ENC-1. The human ENC-1 gene appeared to be highly expressed in adult brain and spinal cord, and in a number of cell lines derived from nervous system tumors we detected low steady-state levels of ENC-1 mRNA. We used a neuroblastoma differentiation model, the retinoic acid-induced neuronal differentiation of SMS-KCNR cells, to study the regulation of the ENC-1 gene during neural crest cell differentiation. We found that the expression of ENC-1 increased dramatically in the differentiated SMS-KCNR cells as compared to control undifferentiated cells. These results suggest that ENC-1 expression plays a role during differentiation of neural crest cells and may be down regulated in neuroblastoma tumors.

Comparison of gadopentetate dimeglumine and albumin-(Gd-DTPA)30 for microvessel characterization in an intracranial glioma model

To compare the performance of macromolecular albumin gadolinium-diethylenetriamine pentaacetic acid (Gd-DTPA)30 and low molecular weight gadopentetate dimeglumine for microvessel characterization, we examined an intracranial 9L glioma model in which increased angiogenesis, hypervascularity, and hyperpermeability mimic characteristics of clinical malignant brain tumors. Dynamic MRI data were analyzed using a bidirectional, two-compartment kinetic model to extract quantitative estimates for fractional blood volume (fBV) and permeability surface area product (PS). Three criteria were used for comparison of contrast agent performance: (a) tumor conspicuity, defined as the contrast-to-noise ratio (CNR); (b) dynamic range of differential permeability estimates between tumor and normal brain; (c) reasonableness of blood volume estimates. Gadopentetate was superior to macromolecular albumin-(Gd-DTPA)30 for detection of 9L brain gliomas and for measurements of hyperpermeability.

ENC-1: a novel mammalian kelch-related gene specifically expressed in the nervous system encodes an actin-binding protein

We have identified and characterized a novel murine gene, Ectoderm-Neural Cortex-1 (ENC-1), that is an early and highly specific marker of neural induction in vertebrates. ENC-1, which encodes a kelch family related protein, is expressed during early gastrulation in the prospective neuroectodermal region of the epiblast and later in development throughout the nervous system (NS). ENC-1 expression is highly dynamic and, after neurulation, preferentially defines prospective cortical areas. The only apparent expression of ENC-1 outside the NS is restricted to the rostral-most somitomere of the presomitic mesoderm, at the times corresponding to the epithelialization that precedes somite formation. Cellular expression of epitope-tagged ENC-1 shows extensive co-localization of ENC-1 with the actin cytoskeleton, and immunoprecipitation studies demonstrate a physical association between ENC-1 and actin. ENC-1 functions as an actin-binding protein that may be important in the organization of the actin cytoskeleton during neural fate specification and development of the NS.

Id genes encoding inhibitors of transcription are expressed during in vitro astrocyte differentiation and in cell lines derived from astrocytic tumors

Id proteins belong to a class of nuclear transcription factors known as helix-loop-helix proteins. It has been reported that Id genes function as negative regulators of differentiation, and Id gene expression is down-regulated during cell differentiation. We examined the regulation of Id genes during astrocyte differentiation in a murine nervous system precursor cell line, NSEHip2-28, which is able to differentiate along the astroglial lineage, as well as in human astroglial tumor cell lines. Upon induction of NSEHip2-28 differentiation, at a time when glial fibrillary acidic protein expression became detectable, the expression of all four Id family members initially increased dramatically, and subsequently decreased. Furthermore, varying levels of Id gene expression were found in astroglial tumor cell lines displaying variable degrees of lineage-specific differentiation. These results suggest that the expression of Id family members may play an important role in the control of astrocyte differentiation.

P16 deletion and mutation analysis in human brain tumors

We screened human primary and recurrent malignant glioma, juvenile pilocytic astrocytoma, medulloblastoma, and meningioma tissue specimens for alterations in p16 gene structure. Single strand conformation polymorphism (SSCP) analysis was used to screen for point mutations, and a quantitative polymerase chain reaction-based assay was used to screen for homozygous gene deletions. In malignant glioma specimens, homozygous p16 gene deletions were significantly more common in high-grade tumors than in low-grade gliomas. Point mutations causing alteration in predicted protein structure were not detected. Medulloblastomas showed rare homozygous deletions and no point mutations. No mutations were detected in meningiomas.

RTVP-1, a novel human gene with sequence similarity to genes of diverse species, is expressed in tumor cell lines of glial but not neuronal origin

A novel gene, RTVP-1, which shows significant sequence identity to the mammalian testis-specific proteins, a family of plant pathogenesis-related proteins and the vespid venom allergen, antigen-5, has been isolated from a cDNA library of the human glioblastoma brain tumor cell line, U-251 MG. The highest degree of sequence identity was with the human testis-specific protein, TPX1 (38.7% over 119 amino acids). Northern hybridization analysis revealed that in fetal tissue RTVP-1 RNA was detected only in the kidney, but its expression was ubiquitous in adult tissues including brain. Multiple mRNAs encoded by RTVP-1 were highly expressed in a panel of cell lines from nervous system tumors arising from glia, although expression was low or absent in nonglial-derived nervous system tumor cell lines.

Adaptation of a simple flow cytometric assay to identify different stages during apoptosis

Apoptosis occurs under many different physiological and pathological conditions, and it reflects a genetically encoded suicide program that can be triggered by different stimuli in susceptible cells. We adapted a flow cytometric assay for the detection of apoptosis based on differential staining of viable cells with two different DNA binding dyes, propidium iodide (PI) and Hoechst 33342 (Ho342). Apoptosis was induced in different cell lines by gamma irradiation, an anti-FAS monoclonal antibody, or ganciclovir in Herpes simplex virus-thymidine kinase-expressing cells. We could identify three different populations that appeared sequentially after the induction of apoptosis. Cells corresponding to these populations were sorted and assessed for evidence of apoptosis as determined by alterations of nuclear morphology and detection of endonucleolytic activity. This analysis revealed a PI- population with subtle apoptotic changes and increased Ho342 fluorescence compared with untreated cells. Extensive apoptotic alterations were observed in a PI+ population that increased over time following the induction of apoptosis. A third population was characterized by an intermediate intensity of PI fluorescence and decreased Ho342 fluorescence compared with the other populations. This population appeared late after treatment and consisted of apoptotic bodies. Taken together, these data suggest that distinct stages of apoptosis can be identified by differential staining of cells with Ho342 and PI. This assay should be useful for the detection and further characterization of cells at different stages in the apoptotic process.

p53-dependent G1 arrest and p53-independent apoptosis influence the radiobiologic response of glioblastoma

PURPOSE

Loss of the p53 tumor suppressor gene has been associated with tumor progression, disease relapse, poor response to antineoplastic therapy, and poor prognosis in many malignancies. We have investigated the contribution of p53-mediated radiation-induced apoptosis and G1 arrest to the well described radiation resistance of glioblastoma multiforme (GM) cells.

METHODS AND MATERIALS

Radiation survival in vitro was quantitated using linear quadratic and repair-saturation mathematical models. Isogenic derivatives of glioblastoma cells differing only in their p53 status were generated using a retroviral vector expressing a dominant negative mutant of p53. Radiation-induced apoptosis was assayed by Fluorescence-activated cell sorter (FACS) analysis, terminal deoxynucleotide transferase labeling technique, and chromatin morphology. Cells were synchronized in early G1 phase and mitotic and labeling indices were measured.

RESULTS

Radiation-induced apoptosis of GM cells was independent of functional wild-type p53 (wt p53). Decreased susceptibility to radiation-induced apoptosis was associated with lower alpha values characterizing the shoulder of the clonogenic radiation survival curve. Using isogenic GM cells differing only in their p53 activity, we found that a p53-mediated function, radiation-induced G1 arrest, could also influence the value of alpha and clonogenic radiation resistance. Inactivation of wt p53 function by a dominant negative mutant of p53 resulted in a significantly diminished alpha value with no alteration in cellular susceptibility to radiation-induced apoptosis. The clonal derivative U87-LUX.8 expressing a functional wt p53 had an alpha (Gy-1) value of 0.609, whereas the isogenic clonal derivative U87-175.4 lacking wt p53 function had an alpha (Gy-1) value of 0.175.

CONCLUSION

We conclude that two distinct cellular responses to radiation, p53-independent apoptosis and p53-dependent G1-arrest, influence radiobiological parameters that characterize the radiation response of glioblastoma cells. Further understanding of the molecular basis of GM radiation resistance will lead to improvement in existing therapeutic modalities and to the development of novel treatment approaches.

Molecular cloning of the cDNA encoding a helix-loop-helix protein, mouse ID1B: tissue-specific expression of ID1A and ID1B genes

A cDNA encoding a 168 amino acid mouse ID1B helix-loop-helix protein, the longest among the ID family of proteins so far identified, was cloned and its nucleotide sequence determined. Mouse ID1B mRNA is distinguishable from the mRNA encoding ID1A at its 3' end, and the relative level of expression of these two different mRNAs is similar in most tissues, with exception of skeletal muscle and kidney. Comparison of the most carboxyl terminal predicted amino acid sequences of ID1 proteins reveals 100% identity for ID1A, but the predicted Id1B proteins of several species are different.

Manganese superoxide dismutase expression in human central nervous system tumors

Manganese superoxide dismutase (MnSOD) is a superoxide anion scavenger located in mitochondria. Increased expression of MnSOD can diminish oxygen radical-mediated injuries and the cytotoxic effects of tumor necrosis factor alpha, ionizing radiation, and certain chemotherapeutic agents. We used immunohistochemical staining to analyze 42 specimens of human brain tumors and 3 normal brain controls with a polyclonal antibody recognizing human MnSOD. We measured MnSOD in cerebrospinal fluid (CSF) from 14 patients with brain tumors and 7 control patients using an ELISA. Although MnSOD is not readily detected in normal brain, malignant central nervous system tumors, including tumors metastatic to the brain, displayed marked immunoreactivity to MnSOD intracellularly, in the extracellular matrix and in the tumor endothelial cells. Grade IV astrocytomas (glioblastomas), Grade III astrocytomas, and medulloblastomas were strongly immunoreactive, whereas Grade II astrocytomas had much less immunoreactivity. ELISA analysis of CSF samples from patients with malignant tumors also revealed high levels of MnSOD protein, up to 45-fold greater than the level of control CSF samples.

Herpes simplex virus thymidine kinase/ganciclovir-mediated apoptotic death of bystander cells

An emerging strategy for cancer gene therapy involves the transfer of the herpes simplex virus thymidine kinase (HSV-tk) gene into tumor cells, rendering them susceptible to the cytotoxic effects of ganciclovir. The observation that HSV-tk-expressing cells can also induce cell death in neighboring cells, which do not express HSV-tk, has been called the bystander effect. Gap junction-mediated transfer of cytotoxic molecules to bystander cells may be an important mechanism of bystander cell death, although others have suggested a role for phagocytosis. In this study, we evaluated the mode of cell death in bystander cells. We detected apoptosis in bystander cells and found that bystander cell death could be inhibited by BCL2 expression. We determined that ganciclovir incubations for 10 h were sufficient to induce cell death in most bystander cells cocultured with HSV-tk-expressing cells. During this period, no phagocytosis was detected, although it was obvious at later stages.

Id2 expression increases with differentiation of human myeloid cells

Id proteins are helix-loop-helix (HLH) transcriptional factors that lack the basic DNA binding domain. The Id proteins have been reported generally to function as inhibitors of cell differentiation, and their gene expression is often downregulated during cell differentiation. We examined the expression of human Id mRNAs by Northern hybridization in 11 human myeloid cell lines, several myeloid cell lines induced to differentiate, fresh myeloid leukemia samples, and normal human myeloid cells. Id2 mRNA was expressed in myelomonoblastic and monoblastic leukemic cells (PLB-985, THP-1, and U-937) but was weakly expressed in myeloblastic leukemic cells (KG-1 and HL-60). Id2 mRNA levels markedly increased with induction of differentiation of myeloid blasts (HL-60, PLB-985, THP-1, and U-937) toward either granulocytes or macrophages. Examination of fresh acute myeloid leukemic samples from 22 individuals also showed prominent Id2 mRNA expression in those samples having more differentiated blasts. Using the French-American-British classification, only 2 of 8 M0/M1 samples expressed Id2 mRNA; however, 10 of 13 M2/M3/M4 samples expressed it. In normal human myeloid cells, Id2 mRNA was expressed in cultured macrophages from bone marrow and in mature granulocytes and monocytes from peripheral blood. The half-life of Id2 mRNA was short (1 hour), and its expression was inducible by cessation of protein synthesis. Id3 mRNA was moderately expressed in monoblastic cell lines (THP-1 and U-937), and levels decreased with their differentiation. Almost no Id3 expression was detectable in either other myeloid leukemia lines, fresh leukemic samples, or normal human myeloid cells by Northern analyses. Id1 mRNA was not detected by polymerase chain reaction in either leukemic or normal myeloid cells except in K562 myeloid/erythroid cells. These results showed that Id2 mRNA was constitutively expressed in more mature myeloid blast cells and level markedly increased with terminal myeloid differentiation, suggesting that Id2 protein may inhibit an HLH transcriptional complex that normally represses myeloid differentiation.

Id2 specifically alters regulation of the cell cycle by tumor suppressor proteins

Cells which are highly proliferative typically lack expression of differentiated, lineage-specific characteristics. Id2, a member of the helix-loop-helix (HLH) protein family known to inhibit cell differentiation, binds to the retinoblastoma protein (pRb) and abolishes its growth-suppressing activity. We found that Id2 but not Id1 or Id3 was able to bind in vitro not only pRb but also the related proteins p107 and p130. Also, an association between Id2 and p107 or p130 was observed in vivo in transiently transfected Saos-2 cells. In agreement with these results, expression of Id1 or Id3 did not affect the block of cell cycle progression mediated by pRb. Conversely, expression of Id2 specifically reversed the cell cycle arrest induced by each of the three members of the pRb family. Furthermore, the growth-suppressive activities of cyclin-dependent kinase inhibitors p16 and p21 were efficiently antagonized by high levels of Id2 but not by Id1 Id3. Consistent with the role of p16 as a selective inhibitor of pRb and pRb-related protein kinase activity, p16-imposed cell cycle arrest was completely abolished by Id2. Only a partial reversal of p21-induced growth suppression was observed, which correlated with the presence of a functional pRb. We also documented decreased levels of cyclin D1 protein and mRNA and the loss of cyclin D1-cdk4 complexes in cells constitutively expressing Id2. These data provide evidence for important Id2-mediated alterations in cell cycle components normally involved in the regulatory events of cell cycle progression, and they highlight a specific role for Id2 as an antagonist of multiple tumor suppressor proteins.

Cytogenetic damage and the radiation-induced G1-phase checkpoint

It is proposed that genomic integrity is preserved after DNA damage in a variety of ways. X irradiation induces a p53-dependent G1-phase cell cycle checkpoint which putatively allows time for repair of DNA damage. The p53 protein is also involved in the initiation of apoptosis after radiation-induced DNA damage, presumably leading to the elimination of lethally damaged cells from the irradiated population. To test the hypothesis that repair occurs in the additional time provided by the activation of the G1-phase checkpoint, we investigated whether the presence of a G1-phase arrest modified the frequency and type of chromosomal rearrangements at the first mitosis after irradiation. Isogenic cell lines derived from the same human glioma cell line, but differing in p53 status, were used. Purified G1-phase cells, isolated by centrifugal elutriation and X-irradiated, were studied. The wild-type p53 cell line demonstrated a dose-dependent arrest during G1 phase, as determined by flow cytometry. These cells remained in G1-phase as long as 48 h after irradiation. Cells expressing a dominant-negative p53 mutation accumulated to a much lesser extent in G1 phase after irradiation. Cells lacking the G1-phase checkpoint showed increased survival at all radiation doses. There were no significant differences in the type or frequency of total chromosomal aberrations in mitotic cells from either cell line after 1,2,4 or 6 Gy X rays, as measured by conventional cytogenetic analysis. There was an increase, however, in the number of reciprocal translocations in mitotic cells with mutant p53 (lacking a G1-phase checkpoint), as measured by fluorescence in situ hybridization with a chromosome 4-specific DNA library, but only after 6 Gy. The results suggest that the presence of a well-defined p53-dependent G1-phase arrest does not reduce chromosomal aberrations caused by low doses of ionizing radiation markedly, but may reduce the overall degree of survival by triggering other G1-phase events.

Cell cycle synchrony unmasks the influence of p53 function on radiosensitivity of human glioblastoma cells

Although ionizing radiation causes DNA damage that can play a role in tumorigenesis, such irradiation is also an important modality of cancer therapy. We studied the radiation response of the U-87 MG human glioblastoma cell line and transfected derivatives in which p53 function had been inactivated. Although little effect of p53 on the radiation sensitivity of asynchronously growing cultures could be detected, inactivation of p53 resulted in a large increase in clonogenic survival when cells synchronized by mitotic selection were irradiated in early G1. The radiation dose sufficient to reduce cellular clonogenicity by 1 log in cells expressing functional p53 was 3.26 +/- 0.12 Gy, whereas a much higher dose (7.41 +/- 0.44 Gy) was required to achieve the same killing effect in cells in which p53 was inactivated. Apoptosis was excluded as a probable mechanism contributing to the radiosensitivity of these cells. Fluorescence-activated cell sorter analysis, continuous labeling with tritiated thymidine, and time-lapse videomicroscopy documented the first example of a prolonged p53-dependent G1 arrest induced by ionizing radiation during the first postirradiation cell cycle of tumor cells, suggesting a role for G1 arrest in determining the sensitivity of these cells to irradiation.